Latest revision as of 02:48, 21 February 2024

Hello, World! ^[1]

Table 1

Estimation Method			MLR	MLR	MLR	MLR	MLR	PCA	MLR	PCA	PCA	PCA	PCA	PCA	PCA	MLR	PCA	PCA	PCA	KDM	MLR	PCA	KDM	MLR	PCA	MLR	PCA	PCA	KDM	Deep learning	KDM	KDM	MLR	PCA	PCA	PCA	KDM	KDM	Deep learning	Deep learning	KDM	MLR	PCA	KDM	KDM	MLR	PCA	KDM	KDM	KDM	KDM	Deep learning
Year			1965	1976	1982	1983	1984	1988	1988	1989	1990	1996	2003	2003	2007	2008	2008	2009	2010	2010	2010	2012	2013	2013	2013	2013	2014	2014	2015	2016	2016	2017	2017	2017	2017	2017	2018	2018	2018	2019	2019	2019	2019	2019	2020	2020	2020	2020	2020	2020	2021	2021
Biomarker Count			7	7	4	8	3	10	10	7	9	8	9	5	5	16	5	11	7	10	10	7	10	10	9	33	6	6	10	10	10	7	8	8	5	5	8	7	19	4	5	6	6	12	10	10	10	11	10	12	7	33
System	Biomarker	Count	^[2]	^[3]	^[4]	^[5]	^[6]	^[7]	^[7]	^[8]	^[9]	^[10]	^[11]	^[12]	^[13]	^[14]	^[15]	^[16]	^[1]	^[17]	^[17]	^[18]	^[19]	^[19]	^[19]	^[20]	^[21]	^[22]	^[23]	^[24]	^[25]	^[26]	^[26]	^[26]	^[27]	^[28]	^[29]	^[30]	^[31]	^[32]	^[33]	^[33]	^[33]	^[34]	^[35]	^[35]	^[35]	^[36]	^[37]	^[38]	^[39]	^[40]
Cardiovascular System	Systolic Blood Pressure (SBP)	32	x	x		x		x	x	x	x	x	x	x	x	x	x	x				x	x	x	x	x			x			x	x	x			x	x			x	x	x	x				x		x	x
Cardiovascular System	Diastolic Blood Pressure (DBP)	6														x										x					x														x	x	x
Cardiovascular System	Pulse Pressure	4																	x							x	x									x
Cardiovascular System	Mean Arterial Pressure	1																																	x
Cardiovascular System	Pulse	3						x	x	x
Cardiovascular System	Pulse Wave Velocity	1				x
Cardiovascular System	Heart Rate	1									x
Cardiovascular System	Intima-Media Thickness	2																	x									x
Cardiovascular System	Minimum Intima-Media Thickness	2																									x									x
Cardiovascular System	End Diastolic Velocity	1																	x
Cardiovascular System	Mitral Valve E/A Peak	2																	x																	x
Cardiovascular System	MVEL, MVES, MVEA	3																	x								x	x
Cardiovascular System	Atherosclerosis Index	2								x	x
Cardiovascular System	NT-proBNP	1																																																		x
Cardiovascular System	Cardiac Troponin I	1																																																		x
Cardiovascular System	Creatine Phosphokinase	1																								x
Cardiovascular System	Homocysteine	1																								x
Respiratory System	Forced Vital Capacity (FVC)	10						x	x	x	x	x	x			x				x	x					x
Respiratory System	Forced Expiratory Volume in 1 Second (FEV1)	25		x	x								x	x	x	x	x	x				x	x	x	x	x			x			x	x	x			x	x			x	x	x		x	x	x
Respiratory System	Vital Capacity	2	x			x
Respiratory System	Maximal Midexpiratory Flow Rate 75/25	1																									x
Respiratory System	VO2 Max	2																x				x
Nervous System	Mini-Mental State Examination (MMSE)	3																																											x	x	x
Nervous System	Digital Symbol Test	2				x																						x
Nervous System	Numeric Memory	2																		x	x
Nervous System	Associated Memory	2																		x	x
Nervous System	Topological Memory	2																		x	x
Nervous System	Short-Time Memory	1					x
Nervous System	Concentration	2																		x	x
Nervous System	Intellectuality - Mental Defect	1				x
Nervous System	Trail Making Test	2																									x									x
Endocrine Metabolic System	Glucose	10										x		x		x										x				x					x			x	x	x												x
Endocrine Metabolic System	HBA1C	13																x					x	x	x	x			x								x			x				x				x	x	x	x
Endocrine Metabolic System	C-peptide	1																																																		x
Endocrine Metabolic System	Insulin	1																																																		x
Endocrine Metabolic System	Triglyceride	8		x							x					x										x									x				x											x		x
Endocrine Metabolic System	Total Cholesterol (TC)	22	x	x	x			x	x			x				x							x	x	x	x			x	x		x	x	x			x	x	x									x		x	x
Endocrine Metabolic System	High-Density Lipoprotein (HDL)	4														x																			x				x													x
Endocrine Metabolic System	Low-Density Lipoprotein (LDL)	4																x								x													x													x
Endocrine Metabolic System	Apolipoprotein A1 and B	1																																																		x
Endocrine Metabolic System	Thyroid-Stimulating Hormone (TSH)	2														x															x
Endocrine Metabolic System	Testosterone	2														x																																				x
Endocrine Metabolic System	Vitamin D	4																																							x	x	x									x
Endocrine Metabolic System	Calcium	2																													x								x
Endocrine Metabolic System	Potassium	1																																					x
Endocrine Metabolic System	Sodium	1																																					x
Endocrine Metabolic System	Inorganic Phosphorus	1																													x
Endocrine Metabolic System	Urea	26		x				x	x	x	x	x	x		x		x	x					x	x	x	x			x	x	x	x	x	x					x	x				x				x	x	x
Endocrine Metabolic System	Creatinine	18														x							x	x		x			x		x		x	x			x		x			x	x	x				x	x	x	x	x
Endocrine Metabolic System	Estimated Glomerular Filtration Rate (eGFR)	3																																											x	x	x
Endocrine Metabolic System	Uric Acid	3																																										x					x			x
Endocrine Metabolic System	Cystatin C	5																	x								x	x								x																x
Endocrine Metabolic System	Creatinine Clearance	1																								x
Endocrine Metabolic System	Urine Specific Gravity	1																								x
Endocrine Metabolic System	Urine pH	1																								x
Digestive System	Alanine Aminotransferase (ALT)	3										x														x																										x
Digestive System	Aspartate Aminotransferase (AST)	9						x	x	x	x															x						x	x	x																		x
Digestive System	Alkaline Phosphatase (ALP)	12		x																			x	x	x	x			x	x	x						x							x					x		x
Digestive System	Total Protein	3																								x					x								x
Digestive System	Albumin	21						x	x				x		x		x	x					x	x	x	x			x	x	x						x		x					x				x	x	x	x	x
Digestive System	Albumin/Globulin Ratio (A/G)	5						x	x				x	x												x
Digestive System	Total Bilirubin	2																								x													x
Digestive System	Direct Bilirubin	1																								x
Digestive System	Amylase	1																								x
Digestive System	Lactate Dehydrogenase	4			x						x	x														x
Digestive System	Alpha 2 Globulin	1																												x
Digestive System	Gamma Glutamyl Transpeptidase	1																								x
Hematologic System	Red Blood Cell	9											x				x													x									x		x	x	x							x		x
Hematologic System	Red Blood Cell Volume Distribution Width	3																												x														x								x
Hematologic System	Hematocrit	4											x		x															x									x
Hematologic System	Mean Corpuscular Volume	5																																					x					x				x	x			x
Hematologic System	Mean Corpuscular Hemoglobin	1												x
Hematologic System	Mean Corpuscular Hemoglobin Concentration	2																																					x													x
Hematologic System	Hemoglobin	12						x	x	x	x	x	x																		x								x						x	x	x					x
Hematologic System	White Blood Cell	4																																										x				x	x			x
Hematologic System	Granulocytes	1																																																		x
Hematologic System	Neutrophils	1																																																		x
Hematologic System	Basophils, Eosinophils	1																																																		x
Hematologic System	Lymphocytes	5																												x														x				x	x			x
Hematologic System	Monocytes	4																																											x	x	x					x
Hematologic System	Platelet	3																																					x											x		x
Hematologic System	Mean Platelet Volume	1																																																		x
Hematologic System	Platelet Distribution Width	1																																																		x
Hematologic System	Erythrocyte Sedimentation Rate	3		x												x		x
Hematologic System	D-dimer, Fibrinogen	3																	x									x																								x
Hematologic System	Ferritin	5																														x	x	x						x										x
Hematologic System	Transferrin	1																																																x
Sensory System	Visual Accommodation	3				x														x	x
Sensory System	Visual Reaction Time	2																		x	x
Sensory System	Visual Acuity	3	x					x	x
Sensory System	Hearing	6	x		x	x												x		x	x
Sensory System	Vibrotactile	4	x				x													x	x
Inflammatory System	C-Reactive Protein (CRP)	12																					x	x	x				x								x	x						x				x	x	x	x	x
Inflammatory System	Cytomegalovirus Optical Density	5																					x	x	x				x																			x
Inflammatory System	Interleukin-6	1																																				x
Inflammatory System	P-selectin	1																																				x
Motion Index	Grip Strength	8	x				x													x	x	x																							x	x	x
Motion Index	Vertical Jump	1																				x
Motion Index	Timed Up and Go Test	3																																											x	x	x
Motion Index	Chair Rise Time	3																																											x	x	x
Body Morphology Index	Waist Circumference (WC)	7																x				x				x						x	x	x	x
Body Morphology Index	Waist-to-Hip Ratio	2														x										x
Body Morphology Index	Waist-to-Height Ratio	3																																							x	x	x
Body Morphology Index	Body Mass Index (BMI)	2														x										x
Body Morphology Index	Weight	1				x
Body Morphology Index	Height	3																																											x	x	x
Body Morphology Index	Body Fat	3														x		x								x
Body Morphology Index	Lean Body Mass	2														x										x
Body Morphology Index	Soft Lean Mass	1																				x
Genetic Index	Terminal Telomere Restriction Fragment	1																										x

Table 3

Biomarker Key	MLR	PCA	KDM	Deep learning
SBP	23213031 5841151 3226152 950448 6667707 17889950 23642770 28110151 30899733	23213031 3226152 2737197 2282902 8803500 12634284 12672981 17921421 18840798 18597867 22433233 28110151 30899733	23213031 26150497 28110151 28958059 28977464 30999227 30899733 31566204 32946548 34038024
DBP	17889950 23642770 31179487	31179487	27216811 31179487
PP	23642770	29188884 24522464 19940465
MAP		28203066
Pulse	3226152	3226152 2737197
PWV	6667707
HR		2282902
IMT		24659482 19940465
MinIMT		29188884 24522464
EDV		19940465
MVEAP		29188884 19940465
MVEL_MVES_MVEA		MVEL 19940465 MVES 24522464 MVEA 24659482
AI		2737197 2282902
NTproBNP				34453631
CTnI				34453631
CPK	23642770
Homocysteine	23642770
FVC	3226152 17889950 23642770 20005245	3226152 2737197 2282902 8803500 12634284	20005245
FEV1	23213031 950448 7162237 17889950 23642770 28110151 30899733 31179487	23213031 12634284 12672981 17921421 18840798 18597867 22433233 28110151 30899733 31179487	23213031 26150497 28110151 28958059 28977464 30899733 31179487
VC	5841151 6667707
MMFR		24522464
VO2Max		18597867 22433233
CR				33744131
MMSE	31179487	31179487	31179487
DST	6667707	24659482
NM	20005245		20005245
AM	20005245		20005245
TM	20005245		20005245
STM	6610563
Concentration	20005245		20005245
IMD	6667707
TMT		29188884 24522464
Glucose	17889950 23642770	8803500 12672981 28203066	28977464	27191382 29340580 30644411 34453631
HBA1C	23213031 23642770	23213031 18597867	23213031 26150497 28958059 30999227 31566204 32946548 31693736 34038024	30644411
CPeptide				34453631
Insulin				34453631
TG	950448 17889950 23642770	2282902 28203066	32946548	29340580 34453631
TC	23213031 5841151 3226152 950448 7162237 17889950 23642770 28110151	23213031 3226152 8803500 28110151	23213031 26150497 28110151 28958059 28977464 31566204 32946548 34038024	27191382 29340580
HDL	17889950	28203066		29340580 34453631
LDL	23642770	18597867		29340580 34453631
ApoA1B				34453631
TSH	17889950		27216811
Testosterone	17889950			34453631
VitaminD	30899733	30899733	30899733	34453631
Calcium			27216811	29340580
Potassium				29340580
Sodium				29340580
InorganicPhosphorus			27216811
Urea	23213031 3226152 950448 23642770 28110151	23213031 3226152 2737197 2282902 8803500 12634284 17921421 18840798 18597867 28110151	23213031 26150497 27216811 28110151 30999227 31566204 32946548 31693736	27191382 29340580 30644411
Creatinine	23213031 17889950 23642770 28110151 30899733	28110151 30899733	23213031 26150497 27216811 28958059 30999227 31566204 32946548 31693736 34038024	29340580 34453631
eGFR	31179487	31179487	31179487
UricAcid			30999227 31693736	34453631
CystatinC		29188884 24659482 24522464 19940465		34453631
CreatinineClearance	23642770
UrineSG	23642770
UrinepH	23642770
ALT	23642770	8803500		34453631
AST	3226152 23642770 28110151	3226152 2737197 2282902 28110151	28110151	34453631
ALP	23213031 950448 23642770	23213031	23213031 26150497 27216811 28958059 30999227 31693736 34038024	27191382
TotalProtein	23642770		27216811	29340580
Albumin	23213031 3226152 23642770	23213031 3226152 12634284 17921421 18840798 18597867	23213031 26150497 27216811 28958059 30999227 31566204 32946548 31693736 34038024	27191382 29340580 34453631
AGRatio	3226152 23642770	3226152 12634284 12672981
TotalBilirubin	23642770			29340580
DirectBilirubin	23642770
Amylase	23642770
LDH	7162237 23642770	2282902 8803500
Alpha2Globulin				27191382
GGT	23642770
RBC	30899733	12634284 18840798 30899733	30899733 32946548	27191382 29340580 34453631
RDW			30999227	27191382 34453631
Hematocrit		12634284 17921421		27191382 29340580
MCV			30999227 31566204 31693736	29340580 34453631
MCH		12672981
MCHC				29340580 34453631
Hemoglobin	3226152 31179487	3226152 2737197 2282902 8803500 12634284 31179487	27216811 31179487	29340580 34453631
WBC			30999227 31566204 31693736	34453631
Granulocytes				34453631
Neutrophils				34453631
BasophilsEosinophils				34453631 34453631
Lymphocytes			30999227 31566204 31693736	27191382 34453631
Monocytes	31179487	31179487	31179487	34453631
Platelet			32946548	29340580 34453631
MPV				34453631
PDW				34453631
ESR	950448 17889950	18597867
DdimerFibrinogen		24659482 19940465		34453631
Ferritin	28110151	28110151	28110151 32946548	30644411
Transferrin			32946548
VisualAcc	6667707 20005245		20005245
VisualReactTime	20005245		20005245
VisualAcuity	5841151 3226152	3226152
Hearing	5841151 7162237 6667707 20005245	18597867	20005245
Vibrotactile	5841151 6610563 20005245		20005245
RetinalPhotos
CRP	23213031	23213031	23213031 26150497 28958059 28977464 30999227 31566204 32946548 31693736 34038024	34453631
CMVOptDensity	23213031	23213031	23213031 26150497 31566204
IL6			28977464
Pselectin			28977464
GripStrength	5841151 6610563 31179487 20005245	22433233 31179487	31179487 20005245
VerticalJump		22433233
TUGTest	31179487	31179487	31179487
ChairRiseTime	31179487	31179487	31179487
PhysicalActivityWeek				29581467 31388024
WC	23642770 28110151	18597867 22433233 28110151 28203066	28110151
WaistHipRatio	17889950 23642770
WaistHeightRatio	30899733	30899733	30899733
BMI	17889950 23642770
Weight	6667707
Height	31179487	31179487	31179487
BodyFat	17889950 23642770	18597867
LeanBodyMass	17889950 23642770
SoftLeanMass		22433233
TelomereLength		24659482

Table 2

Assessment methods	Researchers	Year	Country	Sample size	Age range	Population	Aging biomarkers (Candidate → Final)
MLR	Hollingsworth JW et al.:1965, Correlations between tests of aging in Hiroshima subjects--an attempt to define "physiologic age" ^[2]	1965	Japan	169 Males 268 Females	10–70+ years	General population	17 → 9
MLR	Webster IW & Logie AR:1976, A relationship between functional age and health status in female subjects ^[3]	1976	Australia	1,080 Females	21–83 years	General population	37 → 7
MLR	Takeda H et al.:1982, Evaluation of biological age and physical age by multiple regression analysis ^[4]	1982	Japan	200 Males	20–69 years	Healthy population	10 → 5
MLR	Voitenko VP & Tokar AV:1983, The assessment of biological age and sex differences of human aging ^[5]	1983	Soviet Union	88 Males 109 Females	19–73 years	General population	122 → 11
MLR	Dubina TL et al.:1984, Biological age and its estimation. III. Introduction of a correction to the multiple regression model of biological age in cross-sectional and longitudinal studies ^[6]	1984	Soviet Union	100 Males 63 Females	60–100 years	Healthy population	21 → 3
MLR /PCA	Nakamura E et al.:1988, Assessment of biological age by principal component analysis ^[7]	1988	Japan	462 Males	30–80 years	Healthy population	30 → 11
PCA	Nakamura E et al.:1989, Biological age versus physical fitness age ^[8]	1989	Japan	69 Males	Average 42.6 ± 9.4 years	Healthy population	18 → 7
PCA	Nakamura E et al.:1990, Biological age versus physical fitness age in women ^[9]	1990	Japan	65 Females	20–64 years	Healthy population	18 → 9
PCA	Nakamura E et al.:1996, Effects of habitual physical exercise on physiological age in men aged 20-85 years as estimated using principal component analysis ^[10]	1996	Japan	221 Males	20–85 years	Healthy population	17 → 8
PCA	Nakamura E & Miyao K:2003, Further evaluation of the basic nature of the human biological aging process based on a factor analysis of age-related physiological variables ^[11]	2003	Japan	86 Males	31–77 years	Healthy population (including some early functional decline or disease)	25 → 9
PCA	Ueno LM et al.:2003, Biomarkers of aging in women and the rate of longitudinal changes ^[12]	2003	Japan	981 Females (cross-sectional study) 110 Females (longitudinal study)	28–80 years	Healthy population	31 → 5
PCA	Nakamura E & Miyao K:2007, A method for identifying biomarkers of aging and constructing an index of biological age in humans ^[13]	2007	Japan	86 Males	31–77 years	Healthy population (including some early functional decline or disease)	29 → 5
MLR	Bae CY et al.:2008, Development of models for predicting biological age (BA) with physical, biochemical, and hormonal parameters ^[14]	2008	Korea	1,302 Males 2,273 Females	40–88 years	General population	80 → 25
PCA	Nakamura E & Miyao K:2008, Sex differences in human biological aging ^[15]	2008	Japan	86 Males 93 Females	31–77 years	Healthy population (including some early functional decline or disease)	29 → 5
PCA	Park J et al.:2009, Developing a biological age assessment equation using principal component analysis and clinical biomarkers of aging in Korean men ^[16]	2009	Korea	1,588 Males	30–77 years	Healthy population (including some early functional decline or disease)	11
PCA	Bai X et al.:2010, Evaluation of biological aging process - a population-based study of healthy people in China ^[1]	2010	China	392 Males 460 Females	30–98 years	Healthy population (including some early functional decline or disease)	108 → 8
MLR/PCA/KDM	Cho IH et al.:2010, An empirical comparative study on biological age estimation algorithms with an application of Work Ability Index (WAI) ^[17]	2010	Korea	200 Males	30–70 years	General population	16 → 11/3 principal components
PCA	Jee H et al.:2012, Development and application of biological age prediction models with physical fitness and physiological components in Korean adults ^[18]	2012	Korea	1,604 Males 760 Females	30–85 years	Healthy population	14 → 8
MLR	Bae CY et al.:2013, Models for estimating the biological age of five organs using clinical biomarkers that are commonly measured in clinical practice settings ^[20]	2013	Korea	66,168 Males 55,021 Females	20–89 years	General population	34
MLR/PCA/KDM	Levine ME:2013, Modeling the rate of senescence: can estimated biological age predict mortality more accurately than chronological age? ^[19]	2013	United States	9,389 People	30–75 years	NHANES (1988–1994)	21 → 10
PCA	Zhang WG et al.:2014, Construction of an integral formula of biological age for a healthy Chinese population using principle component analysis ^[21]	2014	China	505 People	35–91 years	Healthy population	114 → 7
PCA	Zhang WG et al.:2014, Select aging biomarkers based on telomere length and chronological age to build a biological age equation ^[22]	2014	China	69 Males 70 Females	35–91 years	Healthy population	105 → 6
KDM	Belsky DW et al.:2015, Quantification of biological aging in young adults ^[23]	2015	New Zealand	954 People	38 years	The Dunedin Study (1972–1973)	10
KDM	Mitnitski A et al.:2017, Heterogeneity of Human Aging and Its Assessment ^[25]	2016	Canada	1,013 People (61.6% Females)	Average 80.8 ± 7.2 years	Canadian Study of Health and Aging (1991–1992)	22 → 10
DNN	Putin E et al.:2016, Deep biomarkers of human aging: Application of deep neural networks to biomarker development ^[24]	2016	Russia	62,419 People	0–100 years	Anonymous population	41
MLR/PCA/KDM	Jee H & Park J:2017, Selection of an optimal set of biomarkers and comparative analyses of biological age estimation models in Korean females ^[26]	2017	Korea	912 Females	30–80 years	Healthy population	31 → 8
PCA	Kang YG et al.:2017, Models for estimating the metabolic syndrome biological age as the new index for evaluation and management of metabolic syndrome ^[27]	2017	Korea	165,395 Males 98,433 Females	Average 44.2 ± 10.6 years	Healthy population (including some early functional decline or disease)	5
PCA	Zhang W et al.:2017, Model Construction for Biological Age Based on a Cross-Sectional Study of a Healthy Chinese Han population ^[28]	2017	China	581 Males 792 Females	19–93 years	Healthy population	74 → 5
KDM	Brown PJ et al.:2018, Biological Age, Not Chronological Age, Is Associated with Late-Life Depression ^[29]	2018	United States	1,356 Males 1,420 Females	70–79 years	The Health ABC Study (2013.11)	8
DNN	Mamoshina P et al.:2018, Population Specific Biomarkers of Human Aging: A Big Data Study Using South Korean, Canadian, and Eastern European Patient Populations ^[31]	2018	Korea, Canada, Eastern Europe	142,379 People	≥20 years	Anonymous population	19
KDM	Murabito JM et al.:2018, Measures of Biologic Age in a Community Sample Predict Mortality and Age-Related Disease: The Framingham Offspring Study ^[30]	2018	United States	2,532–3,417 People	Average 45/62/67 years (Exam 2/7/8)	The Framingham Heart Study Exam 2 (1979–1983) Exam 7 (1998–2001) Exam 8 (2005–2008)	clinical BA:6 inflammatory BA:9
CNN	Pyrkov TV et al.:2018, Extracting biological age from biomedical data via deep learning: too much of a good thing? ^[41]	2018	United States	7,454 People (51% Females)	6–84 years	NHANES (2003–2006)	1-Week Activity Data
KDM	Hastings WJ et al.:2019, Comparability of biological aging measures in the National Health and Nutrition Examination Study, 1999-2002 ^[34]	2019	United States	6,731 People (52% Males)	20–84 years	NHANES (1999–2002)	12
MLR/PCA/KDM	Jee H:2019, Selection of a set of biomarkers and comparisons of biological age estimation models for Korean men ^[33]	2019	Korea	940 Males	30–80 years	Healthy population	32 → 6
DNN	Mamoshina P et al.:2019, Blood Biochemistry Analysis to Detect Smoking Status and Quantify Accelerated Aging in Smokers ^[32]	2019	Canada	149,000 People	Average 55 years	Anonymous population	18/20/23(three DNN models)
ConvLSTM	Rahman SA & Adjeroh DA:2019, Deep Learning using Convolutional LSTM estimates Biological Age from Physical Activity ^[42]	2019	United States	7,104 People	18–84 years	NHANES (2003–2006)	1-Week Activity Data
KDM	Gaydosh L et al.:2020, Testing Proposed Quantifications of Biological Aging in Taiwanese Older Adults ^[36]	2020	China Taiwan	951 People	Average 67.7 ± 8.3 years	Social Environment and Biomarkers of Aging Study (2000)	11
KDM	Liu Z:2021, Development and Validation of 2 Composite Aging Measures Using Routine Clinical Biomarkers in the Chinese Population: Analyses From 2 Prospective Cohort Studies ^[38]	2020	China	8,119 People (53.5% Females)	20–79 years	China Nutrition and Health Survey (2009)	27 → 12
KDM	Parker DC et al.:2020, Association of Blood Chemistry Quantifications of Biological Aging With Disability and Mortality in Older Adults ^[37]	2020	United States	1,374 People (35% Males)	71–102 years	Duke Established Populations for Epidemiologic Studies of the Elderly (1991–1992)	10
MLR/PCA/KDM	Zhong X et al.:2020, Estimating Biological Age in the Singapore Longitudinal Aging Study ^[35]	2020	Singapore	2,844 People	55–94 years	Singapore Longitudinal Aging Studies (2008.03–2013.11)	68 → 8/10(Males/Females)
PCA/KDM	Chan et al.	2021	UK	141,254 People	40–70 years	Healthy population	110 → 51 principal components
DNN	Gialluisi A et al.:2022, Exploring domains, clinical implications and environmental associations of a deep learning marker of biological ageing ^[40]	2021	Italy	23,858 People (51.7% Females)	Average 55.9 ± 12.0 years	The Moli-Sani Study (2005.03–2010.04)	36
KDM	Kuo CL et al.:2021, Genetic associations for two biological age measures point to distinct aging phenotypes ^[39]	2021	UK	294,293 People	Average 56.7 ± 8.0 years	UK Biobank (2006–2010)	7
CNN	Raghu VK et al.:2021, Deep Learning to Estimate Biological Age From Chest Radiographs ^[43]	2021	United States	116,035 People	40–100 years	General population	Chest X-ray dataset
MLR/KDM	Bahour et al.	2022	United States	2,459 People	20–80 years	Diabetes, pre-diabetes, and NHANES (2017–2018) population	8
Deep learning	Nusinovici et al.	2022	Korea	40,480 People	≥65 years	Korean Health Screening study	retinal photos

↑ ^1.0 ^1.1 ^1.2 Bai X et al.: Evaluation of biological aging process - a population-based study of healthy people in China. Gerontology 2010. (PMID 19940465) [PubMed] [DOI] BACKGROUND: Although there have been cross-sectional and longitudinal studies examining biological age (BA) with chronological age (CA)-related changes in physical, physiological, biochemical, and hormonal variables, few studies have performed echocardiographic evaluation of the cardiovascular system and inflammatory biomarkers. Furthermore, little is known about biomarkers of aging and BA score (BAS) for healthy people in China. OBJECTIVES: The purpose of this study was to identify the biomarkers of healthy aging and to establish BAS for healthy people in China. METHODS: We examined 2,876 men and women aged 30-98 years old in three Chinese cities, and 852 healthy subjects were assessed with 108 physical, morphological, physiological and biochemical variables. After excluding binary variables, variables that had a correlation coefficient with CA of < or =0.25 and redundant variables, eight variables including CA, arterial pulse pressure (PP), intima-media thickness (IMT), end diastolic velocity (EDV), ratio of peak velocity of early filling to atrial filling (E/A), mitral valve annulus lateral wall of peak velocity of early filling (MVEL), cystatin C (CYSC), and fibrinogen (FIB) were selected as candidate biomarkers of aging based on a factor-weighted BAS composite for predicting BA. RESULTS: The BAS equation was 0.248 (CA) + 0.195 (IMT) - 0.196 (EDV) - 0.167 (E/A) - 0.166 (MVEL) + 0.188 (PP) + 0.182 (FIB) + 0.193 (CYSC). Individual BAS were significantly correlated with CA (r = 0.893, p < 0.001). Biological aging rate predicted by BAS was accelerated with increases in CA, and peaked when healthy men and women reached > or =75 years of age. CONCLUSIONS: Our data suggest that BAS is superior to CA in assessing the rate of aging in healthy Chinese people. The cardiovascular variables play a crucial role in the evaluation of biological aging. Biological aging rate appears to be age specific.
↑ ^2.0 ^2.1 Hollingsworth JW et al.: Correlations between tests of aging in Hiroshima subjects--an attempt to define "physiologic age". Yale J Biol Med 1965. (PMID 5841151) [PubMed] [Full text]
↑ ^3.0 ^3.1 Webster IW & Logie AR: A relationship between functional age and health status in female subjects. J Gerontol 1976. (PMID 950448) [PubMed] [DOI] A multiple regression equation was used to predict age from seven clinical variables in 1080 apparently well female subjects. A multiple correlation coefficient of R = 0.77 was achieved by five of the variables: timed forced expiratory volume, systolic blood pressure, plasma urea nitrogen, cholesterol, and alkaline phosphatase. On the basis of selection by medical questionnaire responses and other objective criteria, 9% of the subjects were nonsmokers and healthier than the rest. These selected subjects showed a significant reduction in preducted age. Within this group, subjective perception of health was associated with differences in predicted age: poor health with an increase and good health with a decrease in functional age. This study of functional age was based on the healthiest segment of the population in order to minimize the effect of overt pathological processes on the aging rate. An association has been demonstrated between health impairment and predicted age as a measure of the aging rate.
↑ ^4.0 ^4.1 Takeda H et al.: Evaluation of biological age and physical age by multiple regression analysis. Med Inform (Lond) 1982. (PMID 7162237) [PubMed] [DOI]
↑ ^5.0 ^5.1 Voitenko VP & Tokar AV: The assessment of biological age and sex differences of human aging. Exp Aging Res 1983. (PMID 6667707) [PubMed] [DOI] Biological age can be assessed by means of clinical parameters. Some parameters are more suitable than others; criteria for a rational selection of such parameters are discussed in detail. The main tool used in this essay is multiple linear regression. The data reveal a characteristic sex difference in human and, presumably, mammalian aging.
↑ ^6.0 ^6.1 Dubina TL et al.: Biological age and its estimation. III. Introduction of a correction to the multiple regression model of biological age in cross-sectional and longitudinal studies. Exp Gerontol 1984. (PMID 6610563) [PubMed] [DOI] In order to eliminate systematic error inherent in assessment of an individual biological age (BA), a correction is suggested for the multiple regression model of BA. The larger the difference between an individual's chronological age and middle age of the sample, the greater the correction value. BA estimates of apparently healthy people of 60 to 100 years were assessed using measurements of three physiological indices: hand grip strength, short-time memory, and vibrotactile sensitivity. Introduction of the correction allowed comparison of BA estimates of the same individual obtained from repeated observations. Preliminary results show there are at least four types of the BA changes with advancing age.
↑ ^7.0 ^7.1 ^7.2 Nakamura E et al.: Assessment of biological age by principal component analysis. Mech Ageing Dev 1988. (PMID 3226152) [PubMed] [DOI] A method of assessing biological age by the application of principal component analysis is reported. Healthy individuals (462) randomly selected from about 6000 men who had taken a 2-day health examination were studied. Out of the 30 physiological variables examined in routine check-ups, 11 variables were selected as suitable for the assessment of biological age based on the results of factor analysis and the physiological meaning of each test. This variable set was then submitted to principal component analysis, and the 1st principal component obtained from this analysis was used as an equation for assessing one's biological age. However, the biological age calculated from this equation is expressed as a score, so the estimated score was transformed to years (biological age) using the T-score idea. The biological age estimated by this method is practically useful and theoretically valid in contrast with the multiple regression model, because this approach eliminates and overcomes the following 2 big problems of the multiple regression model: (1) the distortion of the individual biological age at the regression edges; and (2) a theoretical contradiction in that a perfect model will merely be predicting the subject's chronological age, not his biological age.
↑ ^8.0 ^8.1 Nakamura E et al.: Biological age versus physical fitness age. Eur J Appl Physiol Occup Physiol 1989. (PMID 2737197) [PubMed] [DOI] A population of healthy middle-aged (n = 69) and elderly men (n = 12), who participated in a health promotion program, was studied to determine whether really physically fit individuals are in good biological condition, and also whether improvement of physical fitness in the middle-aged and the elderly reduces their "rate of aging". Biological and physical fitness ages of the individuals studied were estimated from the data for 18 physiological function tests and 5 physical fitness tests, respectively, by a principal component model. The correlation coefficient between the estimated biological and physical fitness ages was 0.72 (p less than 0.01). Detailed analyses of the relationship between the estimated biological and physical fitness ages revealed that those who manifested a higher ("older") physical fitness age did not necessarily have a higher biological age, but those who manifested a lower ("younger") physical fitness age were also found to have a lower biological age. These results suggested that there were considerable individual variations in the relationship between biological condition and physical fitness among individuals with an old physical fitness age, but those who were in a state of high physical fitness maintained a relatively good biological condition. The data regarding the elderly men who had maintained a regular exercise program indicated that their estimated biological ages were considerably younger than the expected values. This might suggest that in older individuals regular physical activity may provide physiological improvements which in turn might reduce "the rate of aging".
↑ ^9.0 ^9.1 Nakamura E et al.: Biological age versus physical fitness age in women. Eur J Appl Physiol Occup Physiol 1990. (PMID 2282902) [PubMed] [DOI] The purpose of this study was to determine whether adult women who are in a state of high physical fitness are in a good state biologically, in terms of biological and physical fitness ages as estimated by statistical means. The subjects were 65 healthy Japanese women (aged 20-64 years). Biological and physical fitness ages were estimated from the data for 18 physiological function tests and 5 physical fitness tests, respectively, by a principal component model. The correlation coefficient between biological and physical fitness ages was 0.70 (P less than 0.01), which was generally regarded as a high correlation. Therefore, those who were in a state of high physical fitness were considered to be in good biological condition. This result is in good agreement with the results (r = 0.72) from adult men, on whom we reported previously. A statistical analysis to ascertain the relative importance of each contributory variable associated with the variance in biological age suggested that routine clinical evaluation of blood pressure and lipid metabolism might play an important role in determining not only the presence and severity of vascular disease but also the rate of biological aging in women.
↑ ^10.0 ^10.1 Nakamura E et al.: Effects of habitual physical exercise on physiological age in men aged 20-85 years as estimated using principal component analysis. Eur J Appl Physiol Occup Physiol 1996. (PMID 8803500) [PubMed] [DOI] A population of 221 healthy adult men (aged 20-85 years) was studied to determine whether those who exercised regularly were in good biological condition, and also whether those who were in a state of high physical fitness were in a good state biologically, in terms of physiological age (PA) and physical fitness age (FA) as estimated by principal component analysis. A group of 17 physiological function tests and 5 physical fitness tests were employed to estimate PA and FA, respectively. The results of this study indicated that those who maintained high physical fitness at all age decade groups from 20 to 79 years had a trend towards maintaining a relatively lower PA (physiologically younger). Mean PA and FA of the trained group were younger by 4.7 and 7.3 years, respectively than those of the untrained group. In addition, the slope of regression line of PA on chronological age was more gentle in the trained group than that in the untrained group. These results would suggest that those who are in a state of high physical fitness maintain a relatively good physiological condition, and that regular physical exercise may delay physiological changes normally seen with aging, and consequently may increase the life span.
↑ ^11.0 ^11.1 Nakamura E & Miyao K: Further evaluation of the basic nature of the human biological aging process based on a factor analysis of age-related physiological variables. J Gerontol A Biol Sci Med Sci 2003. (PMID 12634284) [PubMed] [DOI] This study aimed to reexamine whether there exists a primary aging process that controls the rate of aging in a number of different functions. Eighty-six adult males who successively received a 2-day routine health checkup test for 7 years from 1992 to 1998 at the Kyoto Red Cross Hospital were selected as subjects. Nine candidate biomarkers of aging were selected from the 25 physiological variables based on the investigation of age-related changes. A principal factor analysis was applied to the partial correlation matrix for 9 selected biomarkers calculated by controlling for age. Furthermore, a confirmatory factor analysis in testing first- and second-order factor models was applied to the covariance matrix for 9 biomarkers. The results of these factor analyses revealed that there existed one general factor and three system-specific factors. Therefore, biological age changes can be viewed as a time-dependent complex integration of the primary and secondary aging processes.
↑ ^12.0 ^12.1 Ueno LM et al.: Biomarkers of aging in women and the rate of longitudinal changes. J Physiol Anthropol Appl Human Sci 2003. (PMID 12672981) [PubMed] [DOI] The purposes of this study were (1) to estimate biological age score (BAS) in Japanese healthy women based on the 4-7 years longitudinal data for physiological, hematological and biochemical examinations and (2) to examine the rate of aging changes in adult women based on the estimated BAS. The samples consisted of cross-sectional (n=981) and longitudinal (n=110) groups. Out of 31 variables examined, five variables (forced expiratory volume in 1.0 s, systolic blood pressure, mean corpuscular hemoglobin, glucose, albumin/globulin ratio) that met the following criteria: 1) significant cross-sectional correlation with age; 2) significant longitudinal change in the same direction as the cross-sectional correlation; and (3) assessment of redundancy, were selected as candidate biomarkers of aging. This variable set was then submitted into a principal component analysis, and the first principal component obtained from this analysis was used as an equation for assessing one's BAS. Individual BAS showed a high longitudinal stability of age-related changes, suggesting high predictive validity of our newly developed aging measurement equation. However, changes in the aging rate based on the estimated BAS were not constant. The mean slopes of the regression lines of BAS for the three age groups (age<45, 45</=age<65 yrs, 65</=age) were 0.095, 0.065, 0.138, respectively. One-way analysis of variance detected a significant difference (F=5.14, p<0.01) among the three age groups. These results suggest that the rate of aging in adult women is relatively slower until 65 years of age, but after 65, the rate of aging shows a rapid increase. We concluded that the longitudinal method used for selection of variables to compute the BAS was useful and theoretically valid compared to those obtained from cross-sectional data analysis.
↑ ^13.0 ^13.1 Nakamura E & Miyao K: A method for identifying biomarkers of aging and constructing an index of biological age in humans. J Gerontol A Biol Sci Med Sci 2007. (PMID 17921421) [PubMed] [DOI] This study was conducted to identify biomarkers of aging and to construct an index of biological age in humans. Healthy adult men (n = 86) who had received an annual health examination from 1992 through 1998 were studied. From 29 physiological variables, five variables (forced expiratory volume in 1 second, systolic blood pressure, hematocrit, albumin, blood urea nitrogen) were selected as candidate biomarkers of aging. Five candidate biomarkers expressed substantial covariance along one principal component. The first principal component obtained from a principal component analysis was used to calculate biological age scores (BAS). Individual BAS showed high longitudinal stability of age-related changes. Age-related changes of BAS are characterized by three components: age, peak functional capacity, and aging rate. A logistic regression analysis suggested that aging rate was influenced by environmental factors, but peak functional capacity was almost independent of environmental factors.
↑ ^14.0 ^14.1 Bae CY et al.: Development of models for predicting biological age (BA) with physical, biochemical, and hormonal parameters. Arch Gerontol Geriatr 2008. (PMID 17889950) [PubMed] [DOI] Individual differences are the hallmark of aging. Chronological age (CHA) is known that fails to provide an accurate indicator of the aging but biological age (BA) estimates the functional status of an individual in reference to his or her chronological peers on the basis of how well he or she functions in comparison with others of the same CHA. Therefore, we developed models for predicting BA that can be applicable in clinical practice settings. This was a community-based cross-sectional study. Subjects were recruited from the health promotion center in Korea from 2001 to 2005. Among these, data obtained from the 3575 participants (1302 men and 2273 women) was used for clinical evaluation and statistical analysis. For our test battery we selected 25 parameters among the routine tests. For males, the best models were developed using 15, 7, 5, and 4 of the 25 chosen parameters for total, physical, biochemical and hormonal characteristics, respectively (R(2)=0.62, 0.38, 0.33, and 0.36, respectively). Similar to males, for the females, 14, 6, 8, and 3 parameters were developed as the models (R(2)=0.66, 0.40, 0.42, and 0.37, respectively). Our BA prediction models may be used as supplementary tools adding knowledge in the evaluation of aging status.
↑ ^15.0 ^15.1 Nakamura E & Miyao K: Sex differences in human biological aging. J Gerontol A Biol Sci Med Sci 2008. (PMID 18840798) [PubMed] [DOI] This study aims to clarify sex differences in human biological aging and to explore the gender gaps in health and longevity. Eighty-six men and 93 women who received a 2-day routine health checkup for 6-7 years beginning in 1992 at the Kyoto Red Cross Hospital were selected. Five candidate biomarkers of aging (forced expiratory volume in 1.0 second per square of height [FEV(1)/Ht(2)], systolic blood pressure [SBP], red blood cells [RBC], albumin [ALBU], and blood urea nitrogen [BUN]) were selected from 29 physiological variables. Individual biological ages (BAS) were estimated from these five biomarkers by a principal component model. From the investigation of the longitudinal changes of individual BAS, it was suggested that (i) beyond 65 years, the rate of aging showed a rapid increase, and (ii) women had relatively lower functional capabilities compared with men, but the rate of aging was slower than that of men, suggesting that these differences might present both disadvantages and advantages for women with regard to health and longevity.
↑ ^16.0 ^16.1 Park J et al.: Developing a biological age assessment equation using principal component analysis and clinical biomarkers of aging in Korean men. Arch Gerontol Geriatr 2009. (PMID 18597867) [PubMed] [DOI] The purpose of the present study is to find clinically useful candidate biomarkers of aging, and using these to develop an equation measuring biological age (BA) in Korean men, then to validate the clinical usefulness of it. Among 4288 men who received medical health examinations, we selected 1588 men who met the normality criteria of each variable. We assumed that chronological ages (CA) of healthy persons represent the BA of them. Variables showing significant correlations with CA were selected. Redundant variables were excluded. We selected 11 variables: VO(2)max, percent body fat (%BF), waist circumference (WC), forced expiratory volume in 1 s (FEV1), systolic blood pressure (SBP), low density cholesterol (LDLCH), blood urea nitrogen (BUN), serum albumin (SA), erythrocyte sedimentation rate(ESR) hearing threshold (HT), and glycosylated hemoglobin (HBA1C). These 11 variables were then submitted into principal component analysis (PCA) and standardized BA scores were obtained. Using them and T-scale idea, the following equation to assess BA was developed: BA=-28.7+0.83(%BF)+0.48(WC)+0.13(SBP)-0.27(VO(2)max)+0.19(HT)-3.1(FEV1)+0.32(BUN)+0.06(LDLCH)-3.0(SA)+0.34(ESR)+4.6(HBA1C). We compared the BA of 3122 men by their fasting glucose and age level. The BA of the higher glucose level group was significantly higher than that of others at all CA levels. The selected 11 biomarkers encompassed known clinically important factors of adult diseases and functional disabilities. This BA assessment equation can be used in the general Korean male population and it proved to be clinically useful.
↑ ^17.0 ^17.1 ^17.2 Cho IH et al.: An empirical comparative study on biological age estimation algorithms with an application of Work Ability Index (WAI). Mech Ageing Dev 2010. (PMID 20005245) [PubMed] [DOI] In this study, we described the characteristics of five different biological age (BA) estimation algorithms, including (i) multiple linear regression, (ii) principal component analysis, and somewhat unique methods developed by (iii) Hochschild, (iv) Klemera and Doubal, and (v) a variant of Klemera and Doubal's method. The objective of this study is to find the most appropriate method of BA estimation by examining the association between Work Ability Index (WAI) and the differences of each algorithm's estimates from chronological age (CA). The WAI was found to be a measure that reflects an individual's current health status rather than the deterioration caused by a serious dependency with the age. Experiments were conducted on 200 Korean male participants using a BA estimation system developed principally under the concept of non-invasive, simple to operate and human function-based. Using the empirical data, BA estimation as well as various analyses including correlation analysis and discriminant function analysis was performed. As a result, it had been confirmed by the empirical data that Klemera and Doubal's method with uncorrelated variables from principal component analysis produces relatively reliable and acceptable BA estimates.
↑ ^18.0 ^18.1 Jee H et al.: Development and application of biological age prediction models with physical fitness and physiological components in Korean adults. Gerontology 2012. (PMID 22433233) [PubMed] [DOI] BACKGROUND: Several biological age (BA) prediction models have been suggested with a variety of biomarkers. Valid models should be able to measure BA in a relatively short time period and predict subsequent physiological capability. Physiological and physical fitness variables have been shown to be distinctive markers for predicting BA and morbidity. The practical and noninvasive nature of such variables makes them useful as clinical assessment tools in estimating BA for in-depth diagnosis and corresponding intervention. OBJECTIVE: To identify, develop and evaluate biomarkers and BA prediction models and validate their clinical usefulness for the practical diagnosis of functional aging. METHODS: Fourteen variables were measured in 3,112 male and 1,233 female participants aged 30 and older between the years 2004 and 2007. Through a series of parsimonious stepwise elimination processes, two sets of 8 gender-specific variables were selected as candidate biomarkers for 1,604 men and 760 women. Principal component analysis, linear regression analysis and adjustment methods were further applied to obtain two sets of true BA (TBA) prediction models. The TBA models were examined for validity by comparing TBA to the corresponding chronological age (CA) with clinical risk factors. RESULTS: TBA prediction models with r(2) values of 0.638 and 0.672 were developed, each unique to men and women, respectively. The overall mean TBA and CA of the participants were 53.9 and 51.8 years, respectively, with a marginal difference of -2.1 and -1.3 years. The regression slopes or rates of TBA as a function of CA were 1.00 and 1.28 for men and women with r values of 0.799 and 0.820 (p < 0.001), respectively. In comparing TBA to CA rates between healthy and clinical risk groups, both sarcopenic and obese groups showed significant increases in TBA. CONCLUSIONS: The selected biomarkers encompass various complex physiopathological factors related to intrinsic and extrinsic physiological and functional aging. The BA prediction models based on the selected biomarkers could be practical in assessing BA for Korean adults.
↑ ^19.0 ^19.1 ^19.2 ^19.3 Levine ME: Modeling the rate of senescence: can estimated biological age predict mortality more accurately than chronological age?. J Gerontol A Biol Sci Med Sci 2013. (PMID 23213031) [PubMed] [DOI] [Full text] Biological age (BA) is useful for examining differences in aging rates. Nevertheless, little consensus exists regarding optimal methods for calculating BA. The aim of this study is to compare the predictive ability of five BA algorithms. The sample included 9,389 persons, aged 30-75 years, from National Health and Nutrition Examination Survey III. During the 18-year follow-up, 1,843 deaths were counted. Each BA algorithm was compared with chronological age on the basis of predictive sensitivity and strength of association with mortality. Results found that the Klemera and Doubal method was the most reliable predictor of mortality and performed significantly better than chronological age. Furthermore, when included with chronological age in a model, Klemera and Doubal method had more robust predictive ability and caused chronological age to no longer be significantly associated with mortality. Given the potential of BA to highlight heterogeneity, the Klemera and Doubal method algorithm may be useful for studying a number of questions regarding the biology of aging.
↑ ^20.0 ^20.1 Bae CY et al.: Models for estimating the biological age of five organs using clinical biomarkers that are commonly measured in clinical practice settings. Maturitas 2013. (PMID 23642770) [PubMed] [DOI] OBJECTIVES: To date, no worldwide studies have been conducted to estimate the biological age of five organs using clinical biomarkers that are associated with the aging status. Therefore, we conducted this study to develop the models for estimating the biological age of five organs (heart, lung, liver, pancreas, and kidney) using clinical biomarkers which are commonly measured in clinical practice. DESIGN: A cross sectional study. METHODS: Subjects were recruited from the routine health check-up centers in Korea from 2004 through 2010. Data obtained from 121,189 subjects (66,168 men and 55,021 women) were used for clinical evaluation and statistical analysis. We examined the relations between clinical biomarkers associated with five organs and the chronological age and proposed a model for estimating the biological age of five organs. RESULTS: In the models for predicting the biological ages of the heart, lung, liver, pancreas and kidney in men, 12, 2, 8, 3, and 5 parameters were respectively included (R(2)=0.652, 0.427, 0.107, 0.245, and 0.651). In contrast to men, 10, 2, 8, 3, and 5 parameters in women were respectively included (R(2)=0.780, 0.435, 0.140, 0.384, and 0.501). CONCLUSION: We first proposed the models for predicting the biological age of five organs in the current study. We developed those using clinical parameters that can be easily obtained in clinical practice settings. Our biological age prediction models may be used as supplementary tools to assess the aging status of five organs in clinical practice settings.
↑ ^21.0 ^21.1 Zhang WG et al.: Construction of an integral formula of biological age for a healthy Chinese population using principle component analysis. J Nutr Health Aging 2014. (PMID 24522464) [PubMed] [DOI] BACKGROUND: Whereas chronological age (CA) cannot distinguish functional differences among individuals of the same age, the biological age (BA) may be used to reflect the functional state of the body. The purpose of this study was to construct an integral formula of the BA, by using principle component analysis (PCA). METHODS: The vital organ function of 505 healthy individuals of Han origin (age 35-91 years) was examined. A total of 114 indicators of cardiovascular, pulmonary, and brain functions, and clinical, inflammatory, genetic, psychological, and life habit factors were assessed as candidate indicators of aging. Candidate indicators were submitted with CA to correlation and redundancy analyses. The PCA method was used to build an integral formula of the BA for the population. RESULTS: Seven biomarkers were selected in accordance with a certain load standard. These biomarkers included the trail making test (TMT), pulse pressure (PP), mitral valve annulus ventricular septum of the peak velocity of early filling (MVES), minimum carotid artery intimal-medial thickness (IMTmin), maximum internal diameter of the carotid artery (Dmax), maximal midexpiratory flow rate 75/25 (MMEF75/25), and Cystatin C (CysC). The formula for the BA was: BA = 0.0685 (TMT) + 0.267 (PP) - 1.375 (MVES) + 22.443 (IMTmin) + 2.962 (Dmax) - 2.332 (MMEF75/25) + 16.104 (CysC) + 0.137 (CA) + 0.492. CONCLUSION: Several genetic and lifestyle indicators were considered as candidate markers of aging. However, ultimately, only markers reflecting the function of the vital organs were included in the BA formula. This study represents a useful attempt to employ multiple indicators to build a comprehensive BA evaluation formula of aging populations.
↑ ^22.0 ^22.1 Zhang WG et al.: Select aging biomarkers based on telomere length and chronological age to build a biological age equation. Age (Dordr) 2014. (PMID 24659482) [PubMed] [DOI] [Full text] The purpose of this study is to build a biological age (BA) equation combining telomere length with chronological age (CA) and associated aging biomarkers. In total, 139 healthy volunteers were recruited from a Chinese Han cohort in Beijing. A genetic index, renal function indices, cardiovascular function indices, brain function indices, and oxidative stress and inflammation indices (C-reactive protein [CRP]) were measured and analyzed. A BA equation was proposed based on selected parameters, with terminal telomere restriction fragment (TRF) and CA as the two principal components. The selected aging markers included mitral annulus peak E anterior wall (MVEA), intima-media thickness (IMT), cystatin C (CYSC), D-dimer (DD), and digital symbol test (DST). The BA equation was: BA = −2.281TRF + 26.321CYSC + 0.025DD − 104.419MVEA + 34.863IMT − 0.265DST + 0.305CA + 26.346. To conclude, telomere length and CA as double benchmarks may be a new method to build a BA.
↑ ^23.0 ^23.1 Belsky DW et al.: Quantification of biological aging in young adults. Proc Natl Acad Sci U S A 2015. (PMID 26150497) [PubMed] [DOI] [Full text] Antiaging therapies show promise in model organism research. Translation to humans is needed to address the challenges of an aging global population. Interventions to slow human aging will need to be applied to still-young individuals. However, most human aging research examines older adults, many with chronic disease. As a result, little is known about aging in young humans. We studied aging in 954 young humans, the Dunedin Study birth cohort, tracking multiple biomarkers across three time points spanning their third and fourth decades of life. We developed and validated two methods by which aging can be measured in young adults, one cross-sectional and one longitudinal. Our longitudinal measure allows quantification of the pace of coordinated physiological deterioration across multiple organ systems (e.g., pulmonary, periodontal, cardiovascular, renal, hepatic, and immune function). We applied these methods to assess biological aging in young humans who had not yet developed age-related diseases. Young individuals of the same chronological age varied in their "biological aging" (declining integrity of multiple organ systems). Already, before midlife, individuals who were aging more rapidly were less physically able, showed cognitive decline and brain aging, self-reported worse health, and looked older. Measured biological aging in young adults can be used to identify causes of aging and evaluate rejuvenation therapies.
↑ ^24.0 ^24.1 Putin E et al.: Deep biomarkers of human aging: Application of deep neural networks to biomarker development. Aging (Albany NY) 2016. (PMID 27191382) [PubMed] [DOI] [Full text] One of the major impediments in human aging research is the absence of a comprehensive and actionable set of biomarkers that may be targeted and measured to track the effectiveness of therapeutic interventions. In this study, we designed a modular ensemble of 21 deep neural networks (DNNs) of varying depth, structure and optimization to predict human chronological age using a basic blood test. To train the DNNs, we used over 60,000 samples from common blood biochemistry and cell count tests from routine health exams performed by a single laboratory and linked to chronological age and sex. The best performing DNN in the ensemble demonstrated 81.5 % epsilon-accuracy r = 0.90 with R(2) = 0.80 and MAE = 6.07 years in predicting chronological age within a 10 year frame, while the entire ensemble achieved 83.5% epsilon-accuracy r = 0.91 with R(2) = 0.82 and MAE = 5.55 years. The ensemble also identified the 5 most important markers for predicting human chronological age: albumin, glucose, alkaline phosphatase, urea and erythrocytes. To allow for public testing and evaluate real-life performance of the predictor, we developed an online system available at http://www.aging.ai. The ensemble approach may facilitate integration of multi-modal data linked to chronological age and sex that may lead to simple, minimally invasive, and affordable methods of tracking integrated biomarkers of aging in humans and performing cross-species feature importance analysis.
↑ ^25.0 ^25.1 Mitnitski A et al.: Heterogeneity of Human Aging and Its Assessment. J Gerontol A Biol Sci Med Sci 2017. (PMID 27216811) [PubMed] [DOI] [Full text] Understanding the heterogeneity in health of older adults is a compelling question in the biology of aging. We analyzed the performance of five measures of health heterogeneity, judging them by their ability to predict mortality. Using clinical and biomarker data on 1,013 participants of the Canadian Study of Health and Aging who were followed for up to 6 years, we calculated two indices of biological age using the Klemera and Doubal method, which controversially includes using chronological age as a "biomarker," and three frailty indices (FIs) that do not include chronological age: a standard clinical FI, an FI from standard laboratory blood tests and blood pressure, and their combination (FI-combined). Predictive validity was tested using Cox proportional hazards analysis and discriminative ability by the area under the receiver-operating characteristic curves. All five measures showed moderate performance that was improved by combining measures to evaluate larger numbers of items. The greatest addition in explanatory power came from the FI-combined that showed the best mortality prediction in an age-adjusted model. More extensive comparisons across different databases are required, but these results do not support including chronological age as a biomarker.
↑ ^26.0 ^26.1 ^26.2 ^26.3 Jee H & Park J: Selection of an optimal set of biomarkers and comparative analyses of biological age estimation models in Korean females. Arch Gerontol Geriatr 2017. (PMID 28110151) [PubMed] [DOI] To date, an optimal working model which predicts biological age (BA) with a set of working biomarkers has not been devised to represent the Korean female population. Accuracy of prediction and applicability are required of an optimal set of commonly assessed biomarkers to provide information on the health status. The goal of this study was to identify a set of biomarkers that represent the aging process and to develop and compare different BA prediction models to elucidate the most fitting and applicable model for providing information on health status in the Korean female population. Using a series of selection processes, eight clinically assessable variables were selected by analyzing relations between 31 clinical variables and chronologic age in 912 normal, healthy individuals among 3642 female participants with ages ranging from 30 to 80 years. The multiple linear regression (MLR), principal component analysis (PCA), and the Klemera-Doubal (KDM) statistical methods were applied to obtain three different sets of BA prediction models. These three models were assessed by calculating and performing the coefficient determinations (r2), regression slopes, effect sizes, pairwise t-tests, and Bland-Altman plots. The BA models were further compared for the applicability by calculating the BAs of clinical risk groups. MLR showed the narrowing effects at the either ends of the age spectrum with greatest effect sizes. PCA showed the greatest degree of dispersion and deviation from the regression center. These MLR and PCA trends were also exhibited by clinically risk groups. In conclusion, the KDM BA prediction model based on the selected biomarkers was found to provide the most reliable and stable results for the practical assessment of BA.
↑ ^27.0 ^27.1 Kang YG et al.: Models for estimating the metabolic syndrome biological age as the new index for evaluation and management of metabolic syndrome. Clin Interv Aging 2017. (PMID 28203066) [PubMed] [DOI] [Full text] PURPOSE: This study aims to propose a metabolic syndrome (MS) biological age model, through which overall evaluation and management of the health status and aging state in MS can be done easily. Through this model, we hope to provide a novel evaluation and management health index that can be utilized in various health care fields. PATIENT AND METHODS: MS parameters from American Heart Association/National Heart, Lung, and Blood Institute guidelines in 2005 were used as biomarkers for the estimation of MS biological age. MS biological age model development was done by analyzing data of 263,828 participants and clinical application of the developed MS biological age was assessed by analyzing the data of 188,886 subjects. RESULTS: The principal component accounted for 36.1% in male and 38.9% in female of the total variance in the battery of five variables. The correlation coefficient between corrected biological age and chronological age in males and females were 0.711 and 0.737, respectively. Significant difference for mean MS biological age and chronological age between the three groups, normal, at risk and MS, was seen (P<0.001). CONCLUSION: For the comprehensive approach in MS management, MS biological age is expected to be additionally utilized as a novel evaluation and management index along with the traditional MS diagnosis.
↑ ^28.0 ^28.1 Zhang W et al.: Model Construction for Biological Age Based on a Cross-Sectional Study of a Healthy Chinese Han population. J Nutr Health Aging 2017. (PMID 29188884) [PubMed] [DOI] OBJECTIVES: Biological age (BA) has been proposed to evaluate the aging status in an objective way instead of chronological age (CA). The purpose of our study is to construct a more precise formula of BA in the cross-sectional study based on a largest-ever sample of our studies. This formula aims at better evaluation of body function and exploring the disciplines of aging in different genders and age stages. METHODS: A total of 1,373 healthy Chinese Han (age range, 19-93 years) were recruited from five cities in China, including 581 males and 792 females. Physical examination, blood routine, blood chemistry, and other lab tests were performed to obtain a total of 74 clinical variables. Then, the principal component analysis (PCA) was used to select variables and estimate BA. The BA formula was further validated in a population with some diseases (n=266), including cardiovascular diseases, type 2 diabetes, kidney diseases, pulmonary diseases, cancer and disorders in nervous system. RESULTS: The BA formula was constructed as follows: BA = 0.358 (pulse pressure) + 0.258 (trail making test) - 11.552 (mitral valve E/A peak) + 26.383 (minimum intima-media thickness) + 31.965 (Cystatin C) + 0.163 (CA) - 3.902. In validation of the formula, BAs of patients were older than those of healthy persons. The BA accelerates faster in the middle-aged population than in the elderly population (>75 years old). CONCLUSION: This BA formula can reflect health condition changes of aging better than CA in a Chinese Han population.
↑ ^29.0 ^29.1 Brown PJ et al.: Biological Age, Not Chronological Age, Is Associated with Late-Life Depression. J Gerontol A Biol Sci Med Sci 2018. (PMID 28958059) [PubMed] [DOI] [Full text] BACKGROUND: The pathophysiology of late-life depression (LLD) is complex and heterogeneous, with age-related processes implicated in its pathogenesis. This study examined the cross-sectional and longitudinal association between depressive symptoms and a baseline multibiomarker algorithm of biological age (BA) that aggregates indicators of inflammatory, metabolic, cardiovascular, lung, liver, and kidney functioning. METHOD: Data were analyzed from 2,776 men and women from the prospective observational Health Aging and Body Composition Study, who had both evaluable chronological age (CA) and BA. Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression (CES-D) scale. RESULTS: A covariate-adjusted regression model showed that BA (B = 0.03, p = .0471) but not CA (B = -0.01, p = .7185) is associated with baseline CES-D scores. The mean baseline BA for individuals with a CES-D ≥ 10 was 1.28 years greater than in those with a CES-D < 10. Comparatively, there is only a 0.05-year difference in mean CA between the two depression groups. A covariate-adjusted longitudinal model found that baseline BA predicts CES-D score at follow-up (B = 0.04, p = .0058), whereas CA does not (B = 0.03, p = .4125). Additionally, an older BA significantly predicted a CES-D ≥ 10 (B = 0.02, p = .032) over a 10-year period. CONCLUSIONS: A multibiomarker index of an older adult's BA outperformed their CA in predicting subsequent increased and clinically significant depressive symptoms. This result supports the evolving view of LLD as a brain disorder resulting from deleterious age-associated changes across numerous physiological systems.
↑ ^30.0 ^30.1 Murabito JM et al.: Measures of Biologic Age in a Community Sample Predict Mortality and Age-Related Disease: The Framingham Offspring Study. J Gerontol A Biol Sci Med Sci 2018. (PMID 28977464) [PubMed] [DOI] [Full text] BACKGROUND: We tested the association of biologic age (BA) measures constructed from different types of biomarkers with mortality and disease in a community-based sample. METHODS: In Framingham Offspring participants at Exams 7 (1998-2001, mean age 62 ± 10) and 8 (2005-2008, mean age 67 ± 9), we used the Klemera-Doubal method to estimate clinical BA and inflammatory BA and computed the difference (∆age) between BA and CA. Clinical ∆age was computed at Exam 2 (1979-1983, mean age 45 ± 10). At Exam 8, we computed measures of intrinsic and extrinsic epigenetic age. Participants were followed through 2014 for outcomes. Cox proportional hazards models tested the association of each BA estimate with each outcome adjusting for covariates. RESULTS: Sample sizes ranged from 2532 to 3417 participants. In multivariable models, each 1-year increase in clinical ∆age at Exam 2 (hazard ratio [HR] = 1.04-1.06, p < 2 × 10-16) and clinical ∆age and inflammatory ∆age at Exam 7 significantly increased the hazards of mortality and incident cardiovascular disease (HR = 1.01-1.05, p < 2 × 10-7), whereas inflammatory ∆age increased the hazards of cancer (HR = 1.01, p < .05). At Exam 8, increased clinical ∆age, inflammatory ∆age, and extrinsic epigenetic age all significantly increased the hazard of mortality (HR = 1.03-1.05, all p < .05); clinical ∆age and inflammatory ∆age increased cardiovascular disease risk (HR = 1.04-1.05, all p < .01); and clinical ∆age increased cancer risk (HR = 1.03, p < .01) when all three BA measures were included in the model. Intrinsic epigenetic age was not significantly associated with any outcome. CONCLUSIONS: Our findings suggest BA measures may be complementary in predicting risk for mortality and age-related disease.
↑ ^31.0 ^31.1 Mamoshina P et al.: Population Specific Biomarkers of Human Aging: A Big Data Study Using South Korean, Canadian, and Eastern European Patient Populations. J Gerontol A Biol Sci Med Sci 2018. (PMID 29340580) [PubMed] [DOI] [Full text] Accurate and physiologically meaningful biomarkers for human aging are key to assessing antiaging therapies. Given ethnic differences in health, diet, lifestyle, behavior, environmental exposures, and even average rate of biological aging, it stands to reason that aging clocks trained on datasets obtained from specific ethnic populations are more likely to account for these potential confounding factors, resulting in an enhanced capacity to predict chronological age and quantify biological age. Here, we present a deep learning-based hematological aging clock modeled using the large combined dataset of Canadian, South Korean, and Eastern European population blood samples that show increased predictive accuracy in individual populations compared to population specific hematologic aging clocks. The performance of models was also evaluated on publicly available samples of the American population from the National Health and Nutrition Examination Survey (NHANES). In addition, we explored the association between age predicted by both population specific and combined hematological clocks and all-cause mortality. Overall, this study suggests (a) the population specificity of aging patterns and (b) hematologic clocks predicts all-cause mortality. The proposed models were added to the freely-available Aging.AI system expanding the range of tools for analysis of human aging.
↑ ^32.0 ^32.1 Mamoshina P et al.: Blood Biochemistry Analysis to Detect Smoking Status and Quantify Accelerated Aging in Smokers. Sci Rep 2019. (PMID 30644411) [PubMed] [DOI] [Full text] There is an association between smoking and cancer, cardiovascular disease and all-cause mortality. However, currently, there are no affordable and informative tests for assessing the effects of smoking on the rate of biological aging. In this study we demonstrate for the first time that smoking status can be predicted using blood biochemistry and cell count results andthe recent advances in artificial intelligence (AI). By employing age-prediction models developed using supervised deep learning techniques, we found that smokers exhibited higher aging rates than nonsmokers, regardless of their cholesterol ratios and fasting glucose levels. We further used those models to quantify the acceleration of biological aging due to tobacco use. Female smokers were predicted to be twice as old as their chronological age compared to nonsmokers, whereas male smokers were predicted to be one and a half times as old as their chronological age compared to nonsmokers. Our findings suggest that deep learning analysis of routine blood tests could complement or even replace the current error-prone method of self-reporting of smoking status and could be expanded to assess the effect of other lifestyle and environmental factors on aging.
↑ ^33.0 ^33.1 ^33.2 ^33.3 Jee H: Selection of a set of biomarkers and comparisons of biological age estimation models for Korean men. J Exerc Rehabil 2019. (PMID 30899733) [PubMed] [DOI] [Full text] Biological age (BA) represents the rate of the senescence with a set of biomarkers. The BA prediction models have not been compared to obtain an optimal BA prediction model with BA biomarkers for Korean men. The study aims to obtain a set of BA biomarkers and compare three of the reported statistical approaches for an optimal BA prediction model. The Korea National Health and Nutrition Examination Surveys data of 2009 to 2011 were used to select six BA biomarkers from 940 healthy subjects aged between 30 to 80 years. The multiple linear regression (MLR), principal component analysis (PCA), and Klemera and Doubal methods (KDM) were used to obtain three BA prediction models. Correlation coefficients (r) with 95% confidence intervals (CI) and regression slopes were assessed. One of the Euro Quality of Life-5 Dimensions, mobility, was compared for feasibility test of each BA models. KDM showed greatest correlation (r=0.88 [P<0.05]) with smallest 95% CI and regression slope (1.00). PCA also showed strong correlation (r=0.79 [P<0.05]) with small 95% CI and regression slope (0.94). MLR (r=0.68 [P<0.05]) showed over and underestimated BA results at the end of the age spectrum. Estimations of BA were most reliable with KDM. The PCA and MLR approaches were comparatively simple to devise for Korean men.
↑ ^34.0 ^34.1 Hastings WJ et al.: Comparability of biological aging measures in the National Health and Nutrition Examination Study, 1999-2002. Psychoneuroendocrinology 2019. (PMID 30999227) [PubMed] [DOI] [Full text] BACKGROUND: Biological processes of aging are thought to be modifiable causes of many different chronic diseases. Measures of biological aging could provide sensitive endpoints for studies of risk factors hypothesized to shorten healthy lifespan and/or interventions that extend it. But uncertainty remains about how to measure biological aging and if proposed measures assess the same thing. METHOD: We tested four proposed measures of biological aging that could be quantified with available data from the National Health and Nutrition Examination Survey (NHANES), Klemera-Doubal method (KDM) Biological Age, homeostatic dysregulation, Levine Method (LM) Biological Age, and leukocyte telomere length. RESULTS: We analyzed data collected during 1999-2002, when all four biological aging meausres could be taken. Participants' KDM biological ages, homeostatic dysregulation levels, LM biological ages, and telomere length were all correlated with their chronological ages. KDM Biological Age, homeostatic dysregulation, and LM Biological Age were all correlated with one another, but these measures were uncorrelated with telomere length. Participants' with more advanced biological aging performed worse on tests of physical, cognitive, and perceptual functioning and reported more limitations to their daily activities and more pain, and rated themselves as being in worse health. In parallel, participants with risk factors for shorter healthy lifespan exhibited more advanced biological aging. In both sets of analyses, effect-sizes tended to be larger for KDM Biological Age, homeostatic dysregulation, and LM Biological Age as compared to telomere length. DISCUSSION: The cellular-level aging biomarker telomere length may measure different aspects of the aging process as compared to the patient-level physiological composite measures KDM Biological Age, homeostatic dysregulation, and LM Biological Age. Studies aiming to test if risk factors accelerate aging or if interventions may slow aging should not treat proposed measures of aging as interchangeable.
↑ ^35.0 ^35.1 ^35.2 ^35.3 Zhong X et al.: Estimating Biological Age in the Singapore Longitudinal Aging Study. J Gerontol A Biol Sci Med Sci 2020. (PMID 31179487) [PubMed] [DOI] BACKGROUND: Biological age (BA) is a more accurate measure of the rate of human aging than chronological age (CA). However, there is limited consensus regarding measures of BA in life span and healthspan. METHODS: This study investigated measurement sets of 68 physiological biomarkers using data from 2,844 Chinese Singaporeans in two age subgroups (55-70 and 71-94 years) in the Singapore Longitudinal Aging Study (SLAS-2) with 8-year follow-up frailty and mortality data. We computed BA estimate using three commonly used algorithms: Principal Component Analysis (PCA), Multiple Linear Regression (MLR), and Klemera and Doubal (KD) method, and additionally, explored the use of machine learning methods for prediction of mortality and frailty. The most optimal algorithmic estimate of BA compared to CA was evaluated for their associations with risk factors and health outcome. RESULTS: Stepwise selection procedures resulted in the final selection of 8 biomarkers in males and 10 biomarkers in females. The highest-ranking biomarkers were estimated glomerular filtration rate for both genders, and the forced expiratory volume in 1 second in males and females. The BA estimates robustly predicted frailty and mortality and outperformed CA. The best performing KD measure of BA was notably predictive in the younger group (aged 55-70 years). BA estimates obtained using a machine learning train-test method were not more accurate than conventional BA estimates in predicting mortality and frailty in most situations. Biologically older people with the same CA as biologically younger individuals had higher prevalence of frailty and 8-year mortality, and worse health, behavioral, and functional characteristics. CONCLUSIONS: BA is better than CA for measuring life span (mortality) and healthspan (frailty). This measurement set of physiological markers of biological aging among Chinese robustly differentiate biologically old from younger individuals with the same CA.
↑ ^36.0 ^36.1 Gaydosh L et al.: Testing Proposed Quantifications of Biological Aging in Taiwanese Older Adults. J Gerontol A Biol Sci Med Sci 2020. (PMID 31566204) [PubMed] [DOI] [Full text] Quantification of biological aging is of interest in gerontology as a means to surveil aging rates in the population and to evaluate the effects of interventions to increase healthy life span. Analysis of proposed methods to quantify biological aging has focused on samples of midlife or mixed-age adults in the West. Research is needed to test whether quantifications of biological aging can differentiate aging rates among older adults and if quantifications of biological aging developed in Western samples can differentiate aging rates in non-Western populations. We conducted analysis of Klemera-Doubal method (KDM) Biological Age and homeostatic dysregulation measures of biological aging developed in the U.S. NHANES and tested in a sample of older Taiwanese adults in the Social Environment and Biomarkers of Aging Study. We conducted analysis of physical and cognitive function and mortality, comparing quantifications of biological aging to a biomarker index based on norms within our analysis sample and to participants' ratings of their own health. Results showed that quantifications of biological aging (a) predicted differences in physical and cognitive function and in mortality risk among Taiwanese older adults and (b) performed as well as a traditional biomarker index and participant self-rated health for prediction of these outcomes.
↑ ^37.0 ^37.1 Parker DC et al.: Association of Blood Chemistry Quantifications of Biological Aging With Disability and Mortality in Older Adults. J Gerontol A Biol Sci Med Sci 2020. (PMID 31693736) [PubMed] [DOI] [Full text] Quantification of biological aging has been proposed for population surveillance of age-related decline in system integrity and evaluation of geroprotective therapies. However, methods of quantifying biological aging have been little studied in geriatric populations. We analyzed three clinical-biomarker-algorithm methods to quantify biological aging. Klemera-Doubal method Biological Age and homeostatic dysregulation algorithms were parameterized from analysis of U.S. National Health and Nutrition Examination Surveys (NHANES) data (N = 36,207) based on published methods. Levine method Biological Age was adapted from published analysis of NHANES data. Algorithms were applied to biomarker data from the Duke Established Populations for Epidemiologic Studies of the Elderly (Duke-EPESE) cohort of older adults (N = 1,374, aged 71-102 years, 35% male, 52% African American). We tested associations of biological aging measures with participant reported Activities of daily living (ADL), instrumental activities of daily living (IADL) dependencies, and mortality. We evaluated the sensitivity of results to the demographic composition of reference samples and biomarker sets used to develop biological aging algorithms. African American and white Duke-EPESE participants with more advanced biological aging reported dependence in more ADLs and IADLs and were at increased risk of death over follow-up through 2017. Effect sizes were similar across algorithms, but were strongest for Levine method Biological Age (per-quintile increase in ADL incidence rate ratio = 1.25, 95% confidence interval [1.17-1.37], IADL incidence rate ratio = 1.23 [1.15-1.32], mortality hazard ratio = 1.12 [1.08-1.16]). Results were insensitive to demographic composition of reference samples, but modestly sensitive to the biomarker sets used to develop biological aging algorithms. Blood-chemistry-based quantifications of biological aging show promise for evaluating the effectiveness of interventions to extend healthy life span in older adults.
↑ ^38.0 ^38.1 Liu Z: Development and Validation of 2 Composite Aging Measures Using Routine Clinical Biomarkers in the Chinese Population: Analyses From 2 Prospective Cohort Studies. J Gerontol A Biol Sci Med Sci 2021. (PMID 32946548) [PubMed] [DOI] [Full text] BACKGROUND: This study aimed to: (i) develop 2 composite aging measures in the Chinese population using 2 recent advanced algorithms (the Klemera and Doubal method and Mahalanobis distance); and (ii) validate the 2 measures by examining their associations with mortality and disease counts. METHODS: Based on data from the China Nutrition and Health Survey (CHNS) 2009 wave (N = 8119, aged 20-79 years, 53.5% women), a nationwide prospective cohort study of the Chinese population, we developed Klemera and Doubal method-biological age (KDM-BA) and physiological dysregulation (PD, derived from Mahalanobis distance) using 12 biomarkers. For the validation analysis, we used Cox proportional hazard regression models (for mortality) and linear, Poisson, and logistic regression models (for disease counts) to examine the associations. We replicated the validation analysis in the China Health and Retirement Longitudinal Study (CHARLS, N = 9304, aged 45-99 years, 53.4% women). RESULTS: Both aging measures were predictive of mortality after accounting for age and gender (KDM-BA, per 1-year, hazard ratio [HR] = 1.14, 95% confidence interval [CI] = 1.08, 1.19; PD, per 1-SD, HR = 1.50, 95% CI = 1.33, 1.69). With few exceptions, these mortality predictions were robust across stratifications by age, gender, education, and health behaviors. The 2 aging measures were associated with disease counts both cross-sectionally and longitudinally. These results were generally replicable in CHARLS although 4 biomarkers were not available. CONCLUSIONS: We successfully developed and validated 2 composite aging measures-KDM-BA and PD, which have great potentials for applications in early identifications and preventions of aging and aging-related diseases in China.
↑ ^39.0 ^39.1 Kuo CL et al.: Genetic associations for two biological age measures point to distinct aging phenotypes. Aging Cell 2021. (PMID 34038024) [PubMed] [DOI] [Full text] Biological age measures outperform chronological age in predicting various aging outcomes, yet little is known regarding genetic predisposition. We performed genome-wide association scans of two age-adjusted biological age measures (PhenoAgeAcceleration and BioAgeAcceleration), estimated from clinical biochemistry markers (Levine et al., 2018; Levine, 2013) in European-descent participants from UK Biobank. The strongest signals were found in the APOE gene, tagged by the two major protein-coding SNPs, PhenoAgeAccel-rs429358 (APOE e4 determinant) (p = 1.50 × 10-72 ); BioAgeAccel-rs7412 (APOE e2 determinant) (p = 3.16 × 10-60 ). Interestingly, we observed inverse APOE e2 and e4 associations and unique pathway enrichments when comparing the two biological age measures. Genes associated with BioAgeAccel were enriched in lipid related pathways, while genes associated with PhenoAgeAccel showed enrichment for immune system, cell function, and carbohydrate homeostasis pathways, suggesting the two measures capture different aging domains. Our study reaffirms that aging patterns are heterogeneous across individuals, and the manner in which a person ages may be partly attributed to genetic predisposition.
↑ ^40.0 ^40.1 Gialluisi A et al.: Exploring domains, clinical implications and environmental associations of a deep learning marker of biological ageing. Eur J Epidemiol 2022. (PMID 34453631) [PubMed] [DOI] Deep Neural Networks (DNN) have been recently developed for the estimation of Biological Age (BA), the hypothetical underlying age of an organism, which can differ from its chronological age (CA). Although promising, these population-specific algorithms warrant further characterization and validation, since their biological, clinical and environmental correlates remain largely unexplored. Here, an accurate DNN was trained to compute BA based on 36 circulating biomarkers in an Italian population (N = 23,858; age ≥ 35 years; 51.7% women). This estimate was heavily influenced by markers of metabolic, heart, kidney and liver function. The resulting Δage (BA-CA) significantly predicted mortality and hospitalization risk for all and specific causes. Slowed biological aging (Δage < 0) was associated with higher physical and mental wellbeing, healthy lifestyles (e.g. adherence to Mediterranean diet) and higher socioeconomic status (educational attainment, household income and occupational status), while accelerated aging (Δage > 0) was associated with smoking and obesity. Together, lifestyles and socioeconomic variables explained ~48% of the total variance in Δage, potentially suggesting the existence of a genetic basis. These findings validate blood-based biological aging as a marker of public health in adult Italians and provide a robust body of knowledge on its biological architecture, clinical implications and potential environmental influences.
↑ Pyrkov TV et al.: Extracting biological age from biomedical data via deep learning: too much of a good thing?. Sci Rep 2018. (PMID 29581467) [PubMed] [DOI] [Full text] Age-related physiological changes in humans are linearly associated with age. Naturally, linear combinations of physiological measures trained to estimate chronological age have recently emerged as a practical way to quantify aging in the form of biological age. In this work, we used one-week long physical activity records from a 2003-2006 National Health and Nutrition Examination Survey (NHANES) to compare three increasingly accurate biological age models: the unsupervised Principal Components Analysis (PCA) score, a multivariate linear regression, and a state-of-the-art deep convolutional neural network (CNN). We found that the supervised approaches produce better chronological age estimations at the expense of a loss of the association between the aging acceleration and all-cause mortality. Consequently, we turned to the NHANES death register directly and introduced a novel way to train parametric proportional hazards models suitable for out-of-the-box implementation with any modern machine learning software. As a demonstration, we produced a separate deep CNN for mortality risks prediction that outperformed any of the biological age or a simple linear proportional hazards model. Altogether, our findings demonstrate the emerging potential of combined wearable sensors and deep learning technologies for applications involving continuous health risk monitoring and real-time feedback to patients and care providers.
↑ Rahman SA & Adjeroh DA: Deep Learning using Convolutional LSTM estimates Biological Age from Physical Activity. Sci Rep 2019. (PMID 31388024) [PubMed] [DOI] [Full text] Human age estimation is an important and difficult challenge. Different biomarkers and numerous approaches have been studied for biological age estimation, each with its advantages and limitations. In this work, we investigate whether physical activity can be exploited for biological age estimation for adult humans. We introduce an approach based on deep convolutional long short term memory (ConvLSTM) to predict biological age, using human physical activity as recorded by a wearable device. We also demonstrate five deep biological age estimation models including the proposed approach and compare their performance on the NHANES physical activity dataset. Results on mortality hazard analysis using both the Cox proportional hazard model and Kaplan-Meier curves each show that the proposed method for estimating biological age outperforms other state-of-the-art approaches. This work has significant implications in combining wearable sensors and deep learning techniques for improved health monitoring, for instance, in a mobile health environment. Mobile health (mHealth) applications provide patients, caregivers, and administrators continuous information about a patient, even outside the hospital.
↑ Raghu VK et al.: Deep Learning to Estimate Biological Age From Chest Radiographs. JACC Cardiovasc Imaging 2021. (PMID 33744131) [PubMed] [DOI] OBJECTIVES: The goal of this study was to assess whether a deep learning estimate of age from a chest radiograph image (CXR-Age) can predict longevity beyond chronological age. BACKGROUND: Chronological age is an imperfect measure of longevity. Biological age, a measure of overall health, may improve personalized care. This paper proposes a new way to estimate biological age using a convolutional neural network that takes as input a CXR image and outputs a chest x-ray age (in years) as a measure of long-term mortality risk. METHODS: CXR-Age was developed using CXR from 116,035 individuals and validated in 2 held-out testing sets: 1) 75% of the CXR arm of PLCO (Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial) (N = 40,967); and 2) the CXR arm of NLST (National Lung Screening Trial) (N = 5,414). CXR-Age was compared to chronological age and a multivariable regression model of chronological age, risk factors, and radiograph findings to predict all-cause and cardiovascular mortality with a maximum 23 years and 13 years of follow-up, respectively. The primary outcome was observed mortality; results are provided for the testing datasets only. RESULTS: In the PLCO testing dataset, a 5-year increase in CXR-Age carried a higher risk of all-cause mortality than a 5-year increase in chronological age (CXR-Age hazard ratio [HR]: 2.26 [95% confidence interval (CI): 2.24 to 2.29] vs. chronological age HR: 1.77 [95% CI: 1.75 to 1.78]; p < 0.001). A similar pattern was found for cardiovascular mortality (CXR-Age cause-specific HR: 2.45 per 5 years [95% CI: 2.34 to 2.56] vs. chronological age HR: 1.82 per 5 years [95% CI: 1.74 to 1.90]). Similar results were seen for both outcomes in the NLST external testing dataset. Adding CXR-Age to the multivariable model resulted in significant improvements for predicting both outcomes in both testing datasets (p < 0.001 for all comparisons). CONCLUSIONS: Based on a CXR image, CXR-Age predicted long-term all-cause and cardiovascular mortality.

[pmid19940465-1] 1.0 ^1.1 ^1.2 Bai X et al.: Evaluation of biological aging process - a population-based study of healthy people in China. Gerontology 2010. (PMID 19940465) [PubMed] [DOI] BACKGROUND: Although there have been cross-sectional and longitudinal studies examining biological age (BA) with chronological age (CA)-related changes in physical, physiological, biochemical, and hormonal variables, few studies have performed echocardiographic evaluation of the cardiovascular system and inflammatory biomarkers. Furthermore, little is known about biomarkers of aging and BA score (BAS) for healthy people in China. OBJECTIVES: The purpose of this study was to identify the biomarkers of healthy aging and to establish BAS for healthy people in China. METHODS: We examined 2,876 men and women aged 30-98 years old in three Chinese cities, and 852 healthy subjects were assessed with 108 physical, morphological, physiological and biochemical variables. After excluding binary variables, variables that had a correlation coefficient with CA of < or =0.25 and redundant variables, eight variables including CA, arterial pulse pressure (PP), intima-media thickness (IMT), end diastolic velocity (EDV), ratio of peak velocity of early filling to atrial filling (E/A), mitral valve annulus lateral wall of peak velocity of early filling (MVEL), cystatin C (CYSC), and fibrinogen (FIB) were selected as candidate biomarkers of aging based on a factor-weighted BAS composite for predicting BA. RESULTS: The BAS equation was 0.248 (CA) + 0.195 (IMT) - 0.196 (EDV) - 0.167 (E/A) - 0.166 (MVEL) + 0.188 (PP) + 0.182 (FIB) + 0.193 (CYSC). Individual BAS were significantly correlated with CA (r = 0.893, p < 0.001). Biological aging rate predicted by BAS was accelerated with increases in CA, and peaked when healthy men and women reached > or =75 years of age. CONCLUSIONS: Our data suggest that BAS is superior to CA in assessing the rate of aging in healthy Chinese people. The cardiovascular variables play a crucial role in the evaluation of biological aging. Biological aging rate appears to be age specific.

[pmid5841151-2] 2.0 ^2.1 Hollingsworth JW et al.: Correlations between tests of aging in Hiroshima subjects--an attempt to define "physiologic age". Yale J Biol Med 1965. (PMID 5841151) [PubMed] [Full text]

[pmid950448-3] 3.0 ^3.1 Webster IW & Logie AR: A relationship between functional age and health status in female subjects. J Gerontol 1976. (PMID 950448) [PubMed] [DOI] A multiple regression equation was used to predict age from seven clinical variables in 1080 apparently well female subjects. A multiple correlation coefficient of R = 0.77 was achieved by five of the variables: timed forced expiratory volume, systolic blood pressure, plasma urea nitrogen, cholesterol, and alkaline phosphatase. On the basis of selection by medical questionnaire responses and other objective criteria, 9% of the subjects were nonsmokers and healthier than the rest. These selected subjects showed a significant reduction in preducted age. Within this group, subjective perception of health was associated with differences in predicted age: poor health with an increase and good health with a decrease in functional age. This study of functional age was based on the healthiest segment of the population in order to minimize the effect of overt pathological processes on the aging rate. An association has been demonstrated between health impairment and predicted age as a measure of the aging rate.

[pmid7162237-4] 4.0 ^4.1 Takeda H et al.: Evaluation of biological age and physical age by multiple regression analysis. Med Inform (Lond) 1982. (PMID 7162237) [PubMed] [DOI]

[pmid6667707-5] 5.0 ^5.1 Voitenko VP & Tokar AV: The assessment of biological age and sex differences of human aging. Exp Aging Res 1983. (PMID 6667707) [PubMed] [DOI] Biological age can be assessed by means of clinical parameters. Some parameters are more suitable than others; criteria for a rational selection of such parameters are discussed in detail. The main tool used in this essay is multiple linear regression. The data reveal a characteristic sex difference in human and, presumably, mammalian aging.

[pmid6610563-6] 6.0 ^6.1 Dubina TL et al.: Biological age and its estimation. III. Introduction of a correction to the multiple regression model of biological age in cross-sectional and longitudinal studies. Exp Gerontol 1984. (PMID 6610563) [PubMed] [DOI] In order to eliminate systematic error inherent in assessment of an individual biological age (BA), a correction is suggested for the multiple regression model of BA. The larger the difference between an individual's chronological age and middle age of the sample, the greater the correction value. BA estimates of apparently healthy people of 60 to 100 years were assessed using measurements of three physiological indices: hand grip strength, short-time memory, and vibrotactile sensitivity. Introduction of the correction allowed comparison of BA estimates of the same individual obtained from repeated observations. Preliminary results show there are at least four types of the BA changes with advancing age.

[pmid3226152-7] 7.0 ^7.1 ^7.2 Nakamura E et al.: Assessment of biological age by principal component analysis. Mech Ageing Dev 1988. (PMID 3226152) [PubMed] [DOI] A method of assessing biological age by the application of principal component analysis is reported. Healthy individuals (462) randomly selected from about 6000 men who had taken a 2-day health examination were studied. Out of the 30 physiological variables examined in routine check-ups, 11 variables were selected as suitable for the assessment of biological age based on the results of factor analysis and the physiological meaning of each test. This variable set was then submitted to principal component analysis, and the 1st principal component obtained from this analysis was used as an equation for assessing one's biological age. However, the biological age calculated from this equation is expressed as a score, so the estimated score was transformed to years (biological age) using the T-score idea. The biological age estimated by this method is practically useful and theoretically valid in contrast with the multiple regression model, because this approach eliminates and overcomes the following 2 big problems of the multiple regression model: (1) the distortion of the individual biological age at the regression edges; and (2) a theoretical contradiction in that a perfect model will merely be predicting the subject's chronological age, not his biological age.

[pmid2737197-8] 8.0 ^8.1 Nakamura E et al.: Biological age versus physical fitness age. Eur J Appl Physiol Occup Physiol 1989. (PMID 2737197) [PubMed] [DOI] A population of healthy middle-aged (n = 69) and elderly men (n = 12), who participated in a health promotion program, was studied to determine whether really physically fit individuals are in good biological condition, and also whether improvement of physical fitness in the middle-aged and the elderly reduces their "rate of aging". Biological and physical fitness ages of the individuals studied were estimated from the data for 18 physiological function tests and 5 physical fitness tests, respectively, by a principal component model. The correlation coefficient between the estimated biological and physical fitness ages was 0.72 (p less than 0.01). Detailed analyses of the relationship between the estimated biological and physical fitness ages revealed that those who manifested a higher ("older") physical fitness age did not necessarily have a higher biological age, but those who manifested a lower ("younger") physical fitness age were also found to have a lower biological age. These results suggested that there were considerable individual variations in the relationship between biological condition and physical fitness among individuals with an old physical fitness age, but those who were in a state of high physical fitness maintained a relatively good biological condition. The data regarding the elderly men who had maintained a regular exercise program indicated that their estimated biological ages were considerably younger than the expected values. This might suggest that in older individuals regular physical activity may provide physiological improvements which in turn might reduce "the rate of aging".

[pmid2282902-9] 9.0 ^9.1 Nakamura E et al.: Biological age versus physical fitness age in women. Eur J Appl Physiol Occup Physiol 1990. (PMID 2282902) [PubMed] [DOI] The purpose of this study was to determine whether adult women who are in a state of high physical fitness are in a good state biologically, in terms of biological and physical fitness ages as estimated by statistical means. The subjects were 65 healthy Japanese women (aged 20-64 years). Biological and physical fitness ages were estimated from the data for 18 physiological function tests and 5 physical fitness tests, respectively, by a principal component model. The correlation coefficient between biological and physical fitness ages was 0.70 (P less than 0.01), which was generally regarded as a high correlation. Therefore, those who were in a state of high physical fitness were considered to be in good biological condition. This result is in good agreement with the results (r = 0.72) from adult men, on whom we reported previously. A statistical analysis to ascertain the relative importance of each contributory variable associated with the variance in biological age suggested that routine clinical evaluation of blood pressure and lipid metabolism might play an important role in determining not only the presence and severity of vascular disease but also the rate of biological aging in women.

[pmid8803500-10] 10.0 ^10.1 Nakamura E et al.: Effects of habitual physical exercise on physiological age in men aged 20-85 years as estimated using principal component analysis. Eur J Appl Physiol Occup Physiol 1996. (PMID 8803500) [PubMed] [DOI] A population of 221 healthy adult men (aged 20-85 years) was studied to determine whether those who exercised regularly were in good biological condition, and also whether those who were in a state of high physical fitness were in a good state biologically, in terms of physiological age (PA) and physical fitness age (FA) as estimated by principal component analysis. A group of 17 physiological function tests and 5 physical fitness tests were employed to estimate PA and FA, respectively. The results of this study indicated that those who maintained high physical fitness at all age decade groups from 20 to 79 years had a trend towards maintaining a relatively lower PA (physiologically younger). Mean PA and FA of the trained group were younger by 4.7 and 7.3 years, respectively than those of the untrained group. In addition, the slope of regression line of PA on chronological age was more gentle in the trained group than that in the untrained group. These results would suggest that those who are in a state of high physical fitness maintain a relatively good physiological condition, and that regular physical exercise may delay physiological changes normally seen with aging, and consequently may increase the life span.

[pmid12634284-11] 11.0 ^11.1 Nakamura E & Miyao K: Further evaluation of the basic nature of the human biological aging process based on a factor analysis of age-related physiological variables. J Gerontol A Biol Sci Med Sci 2003. (PMID 12634284) [PubMed] [DOI] This study aimed to reexamine whether there exists a primary aging process that controls the rate of aging in a number of different functions. Eighty-six adult males who successively received a 2-day routine health checkup test for 7 years from 1992 to 1998 at the Kyoto Red Cross Hospital were selected as subjects. Nine candidate biomarkers of aging were selected from the 25 physiological variables based on the investigation of age-related changes. A principal factor analysis was applied to the partial correlation matrix for 9 selected biomarkers calculated by controlling for age. Furthermore, a confirmatory factor analysis in testing first- and second-order factor models was applied to the covariance matrix for 9 biomarkers. The results of these factor analyses revealed that there existed one general factor and three system-specific factors. Therefore, biological age changes can be viewed as a time-dependent complex integration of the primary and secondary aging processes.

[pmid12672981-12] 12.0 ^12.1 Ueno LM et al.: Biomarkers of aging in women and the rate of longitudinal changes. J Physiol Anthropol Appl Human Sci 2003. (PMID 12672981) [PubMed] [DOI] The purposes of this study were (1) to estimate biological age score (BAS) in Japanese healthy women based on the 4-7 years longitudinal data for physiological, hematological and biochemical examinations and (2) to examine the rate of aging changes in adult women based on the estimated BAS. The samples consisted of cross-sectional (n=981) and longitudinal (n=110) groups. Out of 31 variables examined, five variables (forced expiratory volume in 1.0 s, systolic blood pressure, mean corpuscular hemoglobin, glucose, albumin/globulin ratio) that met the following criteria: 1) significant cross-sectional correlation with age; 2) significant longitudinal change in the same direction as the cross-sectional correlation; and (3) assessment of redundancy, were selected as candidate biomarkers of aging. This variable set was then submitted into a principal component analysis, and the first principal component obtained from this analysis was used as an equation for assessing one's BAS. Individual BAS showed a high longitudinal stability of age-related changes, suggesting high predictive validity of our newly developed aging measurement equation. However, changes in the aging rate based on the estimated BAS were not constant. The mean slopes of the regression lines of BAS for the three age groups (age<45, 45</=age<65 yrs, 65</=age) were 0.095, 0.065, 0.138, respectively. One-way analysis of variance detected a significant difference (F=5.14, p<0.01) among the three age groups. These results suggest that the rate of aging in adult women is relatively slower until 65 years of age, but after 65, the rate of aging shows a rapid increase. We concluded that the longitudinal method used for selection of variables to compute the BAS was useful and theoretically valid compared to those obtained from cross-sectional data analysis.

[pmid17921421-13] 13.0 ^13.1 Nakamura E & Miyao K: A method for identifying biomarkers of aging and constructing an index of biological age in humans. J Gerontol A Biol Sci Med Sci 2007. (PMID 17921421) [PubMed] [DOI] This study was conducted to identify biomarkers of aging and to construct an index of biological age in humans. Healthy adult men (n = 86) who had received an annual health examination from 1992 through 1998 were studied. From 29 physiological variables, five variables (forced expiratory volume in 1 second, systolic blood pressure, hematocrit, albumin, blood urea nitrogen) were selected as candidate biomarkers of aging. Five candidate biomarkers expressed substantial covariance along one principal component. The first principal component obtained from a principal component analysis was used to calculate biological age scores (BAS). Individual BAS showed high longitudinal stability of age-related changes. Age-related changes of BAS are characterized by three components: age, peak functional capacity, and aging rate. A logistic regression analysis suggested that aging rate was influenced by environmental factors, but peak functional capacity was almost independent of environmental factors.

[pmid17889950-14] 14.0 ^14.1 Bae CY et al.: Development of models for predicting biological age (BA) with physical, biochemical, and hormonal parameters. Arch Gerontol Geriatr 2008. (PMID 17889950) [PubMed] [DOI] Individual differences are the hallmark of aging. Chronological age (CHA) is known that fails to provide an accurate indicator of the aging but biological age (BA) estimates the functional status of an individual in reference to his or her chronological peers on the basis of how well he or she functions in comparison with others of the same CHA. Therefore, we developed models for predicting BA that can be applicable in clinical practice settings. This was a community-based cross-sectional study. Subjects were recruited from the health promotion center in Korea from 2001 to 2005. Among these, data obtained from the 3575 participants (1302 men and 2273 women) was used for clinical evaluation and statistical analysis. For our test battery we selected 25 parameters among the routine tests. For males, the best models were developed using 15, 7, 5, and 4 of the 25 chosen parameters for total, physical, biochemical and hormonal characteristics, respectively (R(2)=0.62, 0.38, 0.33, and 0.36, respectively). Similar to males, for the females, 14, 6, 8, and 3 parameters were developed as the models (R(2)=0.66, 0.40, 0.42, and 0.37, respectively). Our BA prediction models may be used as supplementary tools adding knowledge in the evaluation of aging status.

[pmid18840798-15] 15.0 ^15.1 Nakamura E & Miyao K: Sex differences in human biological aging. J Gerontol A Biol Sci Med Sci 2008. (PMID 18840798) [PubMed] [DOI] This study aims to clarify sex differences in human biological aging and to explore the gender gaps in health and longevity. Eighty-six men and 93 women who received a 2-day routine health checkup for 6-7 years beginning in 1992 at the Kyoto Red Cross Hospital were selected. Five candidate biomarkers of aging (forced expiratory volume in 1.0 second per square of height [FEV(1)/Ht(2)], systolic blood pressure [SBP], red blood cells [RBC], albumin [ALBU], and blood urea nitrogen [BUN]) were selected from 29 physiological variables. Individual biological ages (BAS) were estimated from these five biomarkers by a principal component model. From the investigation of the longitudinal changes of individual BAS, it was suggested that (i) beyond 65 years, the rate of aging showed a rapid increase, and (ii) women had relatively lower functional capabilities compared with men, but the rate of aging was slower than that of men, suggesting that these differences might present both disadvantages and advantages for women with regard to health and longevity.

[pmid18597867-16] 16.0 ^16.1 Park J et al.: Developing a biological age assessment equation using principal component analysis and clinical biomarkers of aging in Korean men. Arch Gerontol Geriatr 2009. (PMID 18597867) [PubMed] [DOI] The purpose of the present study is to find clinically useful candidate biomarkers of aging, and using these to develop an equation measuring biological age (BA) in Korean men, then to validate the clinical usefulness of it. Among 4288 men who received medical health examinations, we selected 1588 men who met the normality criteria of each variable. We assumed that chronological ages (CA) of healthy persons represent the BA of them. Variables showing significant correlations with CA were selected. Redundant variables were excluded. We selected 11 variables: VO(2)max, percent body fat (%BF), waist circumference (WC), forced expiratory volume in 1 s (FEV1), systolic blood pressure (SBP), low density cholesterol (LDLCH), blood urea nitrogen (BUN), serum albumin (SA), erythrocyte sedimentation rate(ESR) hearing threshold (HT), and glycosylated hemoglobin (HBA1C). These 11 variables were then submitted into principal component analysis (PCA) and standardized BA scores were obtained. Using them and T-scale idea, the following equation to assess BA was developed: BA=-28.7+0.83(%BF)+0.48(WC)+0.13(SBP)-0.27(VO(2)max)+0.19(HT)-3.1(FEV1)+0.32(BUN)+0.06(LDLCH)-3.0(SA)+0.34(ESR)+4.6(HBA1C). We compared the BA of 3122 men by their fasting glucose and age level. The BA of the higher glucose level group was significantly higher than that of others at all CA levels. The selected 11 biomarkers encompassed known clinically important factors of adult diseases and functional disabilities. This BA assessment equation can be used in the general Korean male population and it proved to be clinically useful.

[pmid20005245-17] 17.0 ^17.1 ^17.2 Cho IH et al.: An empirical comparative study on biological age estimation algorithms with an application of Work Ability Index (WAI). Mech Ageing Dev 2010. (PMID 20005245) [PubMed] [DOI] In this study, we described the characteristics of five different biological age (BA) estimation algorithms, including (i) multiple linear regression, (ii) principal component analysis, and somewhat unique methods developed by (iii) Hochschild, (iv) Klemera and Doubal, and (v) a variant of Klemera and Doubal's method. The objective of this study is to find the most appropriate method of BA estimation by examining the association between Work Ability Index (WAI) and the differences of each algorithm's estimates from chronological age (CA). The WAI was found to be a measure that reflects an individual's current health status rather than the deterioration caused by a serious dependency with the age. Experiments were conducted on 200 Korean male participants using a BA estimation system developed principally under the concept of non-invasive, simple to operate and human function-based. Using the empirical data, BA estimation as well as various analyses including correlation analysis and discriminant function analysis was performed. As a result, it had been confirmed by the empirical data that Klemera and Doubal's method with uncorrelated variables from principal component analysis produces relatively reliable and acceptable BA estimates.

[pmid22433233-18] 18.0 ^18.1 Jee H et al.: Development and application of biological age prediction models with physical fitness and physiological components in Korean adults. Gerontology 2012. (PMID 22433233) [PubMed] [DOI] BACKGROUND: Several biological age (BA) prediction models have been suggested with a variety of biomarkers. Valid models should be able to measure BA in a relatively short time period and predict subsequent physiological capability. Physiological and physical fitness variables have been shown to be distinctive markers for predicting BA and morbidity. The practical and noninvasive nature of such variables makes them useful as clinical assessment tools in estimating BA for in-depth diagnosis and corresponding intervention. OBJECTIVE: To identify, develop and evaluate biomarkers and BA prediction models and validate their clinical usefulness for the practical diagnosis of functional aging. METHODS: Fourteen variables were measured in 3,112 male and 1,233 female participants aged 30 and older between the years 2004 and 2007. Through a series of parsimonious stepwise elimination processes, two sets of 8 gender-specific variables were selected as candidate biomarkers for 1,604 men and 760 women. Principal component analysis, linear regression analysis and adjustment methods were further applied to obtain two sets of true BA (TBA) prediction models. The TBA models were examined for validity by comparing TBA to the corresponding chronological age (CA) with clinical risk factors. RESULTS: TBA prediction models with r(2) values of 0.638 and 0.672 were developed, each unique to men and women, respectively. The overall mean TBA and CA of the participants were 53.9 and 51.8 years, respectively, with a marginal difference of -2.1 and -1.3 years. The regression slopes or rates of TBA as a function of CA were 1.00 and 1.28 for men and women with r values of 0.799 and 0.820 (p < 0.001), respectively. In comparing TBA to CA rates between healthy and clinical risk groups, both sarcopenic and obese groups showed significant increases in TBA. CONCLUSIONS: The selected biomarkers encompass various complex physiopathological factors related to intrinsic and extrinsic physiological and functional aging. The BA prediction models based on the selected biomarkers could be practical in assessing BA for Korean adults.

[pmid23213031-19] 19.0 ^19.1 ^19.2 ^19.3 Levine ME: Modeling the rate of senescence: can estimated biological age predict mortality more accurately than chronological age?. J Gerontol A Biol Sci Med Sci 2013. (PMID 23213031) [PubMed] [DOI] [Full text] Biological age (BA) is useful for examining differences in aging rates. Nevertheless, little consensus exists regarding optimal methods for calculating BA. The aim of this study is to compare the predictive ability of five BA algorithms. The sample included 9,389 persons, aged 30-75 years, from National Health and Nutrition Examination Survey III. During the 18-year follow-up, 1,843 deaths were counted. Each BA algorithm was compared with chronological age on the basis of predictive sensitivity and strength of association with mortality. Results found that the Klemera and Doubal method was the most reliable predictor of mortality and performed significantly better than chronological age. Furthermore, when included with chronological age in a model, Klemera and Doubal method had more robust predictive ability and caused chronological age to no longer be significantly associated with mortality. Given the potential of BA to highlight heterogeneity, the Klemera and Doubal method algorithm may be useful for studying a number of questions regarding the biology of aging.

[pmid23642770-20] 20.0 ^20.1 Bae CY et al.: Models for estimating the biological age of five organs using clinical biomarkers that are commonly measured in clinical practice settings. Maturitas 2013. (PMID 23642770) [PubMed] [DOI] OBJECTIVES: To date, no worldwide studies have been conducted to estimate the biological age of five organs using clinical biomarkers that are associated with the aging status. Therefore, we conducted this study to develop the models for estimating the biological age of five organs (heart, lung, liver, pancreas, and kidney) using clinical biomarkers which are commonly measured in clinical practice. DESIGN: A cross sectional study. METHODS: Subjects were recruited from the routine health check-up centers in Korea from 2004 through 2010. Data obtained from 121,189 subjects (66,168 men and 55,021 women) were used for clinical evaluation and statistical analysis. We examined the relations between clinical biomarkers associated with five organs and the chronological age and proposed a model for estimating the biological age of five organs. RESULTS: In the models for predicting the biological ages of the heart, lung, liver, pancreas and kidney in men, 12, 2, 8, 3, and 5 parameters were respectively included (R(2)=0.652, 0.427, 0.107, 0.245, and 0.651). In contrast to men, 10, 2, 8, 3, and 5 parameters in women were respectively included (R(2)=0.780, 0.435, 0.140, 0.384, and 0.501). CONCLUSION: We first proposed the models for predicting the biological age of five organs in the current study. We developed those using clinical parameters that can be easily obtained in clinical practice settings. Our biological age prediction models may be used as supplementary tools to assess the aging status of five organs in clinical practice settings.

[pmid24522464-21] 21.0 ^21.1 Zhang WG et al.: Construction of an integral formula of biological age for a healthy Chinese population using principle component analysis. J Nutr Health Aging 2014. (PMID 24522464) [PubMed] [DOI] BACKGROUND: Whereas chronological age (CA) cannot distinguish functional differences among individuals of the same age, the biological age (BA) may be used to reflect the functional state of the body. The purpose of this study was to construct an integral formula of the BA, by using principle component analysis (PCA). METHODS: The vital organ function of 505 healthy individuals of Han origin (age 35-91 years) was examined. A total of 114 indicators of cardiovascular, pulmonary, and brain functions, and clinical, inflammatory, genetic, psychological, and life habit factors were assessed as candidate indicators of aging. Candidate indicators were submitted with CA to correlation and redundancy analyses. The PCA method was used to build an integral formula of the BA for the population. RESULTS: Seven biomarkers were selected in accordance with a certain load standard. These biomarkers included the trail making test (TMT), pulse pressure (PP), mitral valve annulus ventricular septum of the peak velocity of early filling (MVES), minimum carotid artery intimal-medial thickness (IMTmin), maximum internal diameter of the carotid artery (Dmax), maximal midexpiratory flow rate 75/25 (MMEF75/25), and Cystatin C (CysC). The formula for the BA was: BA = 0.0685 (TMT) + 0.267 (PP) - 1.375 (MVES) + 22.443 (IMTmin) + 2.962 (Dmax) - 2.332 (MMEF75/25) + 16.104 (CysC) + 0.137 (CA) + 0.492. CONCLUSION: Several genetic and lifestyle indicators were considered as candidate markers of aging. However, ultimately, only markers reflecting the function of the vital organs were included in the BA formula. This study represents a useful attempt to employ multiple indicators to build a comprehensive BA evaluation formula of aging populations.

[pmid24659482-22] 22.0 ^22.1 Zhang WG et al.: Select aging biomarkers based on telomere length and chronological age to build a biological age equation. Age (Dordr) 2014. (PMID 24659482) [PubMed] [DOI] [Full text] The purpose of this study is to build a biological age (BA) equation combining telomere length with chronological age (CA) and associated aging biomarkers. In total, 139 healthy volunteers were recruited from a Chinese Han cohort in Beijing. A genetic index, renal function indices, cardiovascular function indices, brain function indices, and oxidative stress and inflammation indices (C-reactive protein [CRP]) were measured and analyzed. A BA equation was proposed based on selected parameters, with terminal telomere restriction fragment (TRF) and CA as the two principal components. The selected aging markers included mitral annulus peak E anterior wall (MVEA), intima-media thickness (IMT), cystatin C (CYSC), D-dimer (DD), and digital symbol test (DST). The BA equation was: BA = −2.281TRF + 26.321CYSC + 0.025DD − 104.419MVEA + 34.863IMT − 0.265DST + 0.305CA + 26.346. To conclude, telomere length and CA as double benchmarks may be a new method to build a BA.

[pmid26150497-23] 23.0 ^23.1 Belsky DW et al.: Quantification of biological aging in young adults. Proc Natl Acad Sci U S A 2015. (PMID 26150497) [PubMed] [DOI] [Full text] Antiaging therapies show promise in model organism research. Translation to humans is needed to address the challenges of an aging global population. Interventions to slow human aging will need to be applied to still-young individuals. However, most human aging research examines older adults, many with chronic disease. As a result, little is known about aging in young humans. We studied aging in 954 young humans, the Dunedin Study birth cohort, tracking multiple biomarkers across three time points spanning their third and fourth decades of life. We developed and validated two methods by which aging can be measured in young adults, one cross-sectional and one longitudinal. Our longitudinal measure allows quantification of the pace of coordinated physiological deterioration across multiple organ systems (e.g., pulmonary, periodontal, cardiovascular, renal, hepatic, and immune function). We applied these methods to assess biological aging in young humans who had not yet developed age-related diseases. Young individuals of the same chronological age varied in their "biological aging" (declining integrity of multiple organ systems). Already, before midlife, individuals who were aging more rapidly were less physically able, showed cognitive decline and brain aging, self-reported worse health, and looked older. Measured biological aging in young adults can be used to identify causes of aging and evaluate rejuvenation therapies.

[pmid27191382-24] 24.0 ^24.1 Putin E et al.: Deep biomarkers of human aging: Application of deep neural networks to biomarker development. Aging (Albany NY) 2016. (PMID 27191382) [PubMed] [DOI] [Full text] One of the major impediments in human aging research is the absence of a comprehensive and actionable set of biomarkers that may be targeted and measured to track the effectiveness of therapeutic interventions. In this study, we designed a modular ensemble of 21 deep neural networks (DNNs) of varying depth, structure and optimization to predict human chronological age using a basic blood test. To train the DNNs, we used over 60,000 samples from common blood biochemistry and cell count tests from routine health exams performed by a single laboratory and linked to chronological age and sex. The best performing DNN in the ensemble demonstrated 81.5 % epsilon-accuracy r = 0.90 with R(2) = 0.80 and MAE = 6.07 years in predicting chronological age within a 10 year frame, while the entire ensemble achieved 83.5% epsilon-accuracy r = 0.91 with R(2) = 0.82 and MAE = 5.55 years. The ensemble also identified the 5 most important markers for predicting human chronological age: albumin, glucose, alkaline phosphatase, urea and erythrocytes. To allow for public testing and evaluate real-life performance of the predictor, we developed an online system available at http://www.aging.ai. The ensemble approach may facilitate integration of multi-modal data linked to chronological age and sex that may lead to simple, minimally invasive, and affordable methods of tracking integrated biomarkers of aging in humans and performing cross-species feature importance analysis.

[pmid27216811-25] 25.0 ^25.1 Mitnitski A et al.: Heterogeneity of Human Aging and Its Assessment. J Gerontol A Biol Sci Med Sci 2017. (PMID 27216811) [PubMed] [DOI] [Full text] Understanding the heterogeneity in health of older adults is a compelling question in the biology of aging. We analyzed the performance of five measures of health heterogeneity, judging them by their ability to predict mortality. Using clinical and biomarker data on 1,013 participants of the Canadian Study of Health and Aging who were followed for up to 6 years, we calculated two indices of biological age using the Klemera and Doubal method, which controversially includes using chronological age as a "biomarker," and three frailty indices (FIs) that do not include chronological age: a standard clinical FI, an FI from standard laboratory blood tests and blood pressure, and their combination (FI-combined). Predictive validity was tested using Cox proportional hazards analysis and discriminative ability by the area under the receiver-operating characteristic curves. All five measures showed moderate performance that was improved by combining measures to evaluate larger numbers of items. The greatest addition in explanatory power came from the FI-combined that showed the best mortality prediction in an age-adjusted model. More extensive comparisons across different databases are required, but these results do not support including chronological age as a biomarker.

[pmid28110151-26] 26.0 ^26.1 ^26.2 ^26.3 Jee H & Park J: Selection of an optimal set of biomarkers and comparative analyses of biological age estimation models in Korean females. Arch Gerontol Geriatr 2017. (PMID 28110151) [PubMed] [DOI] To date, an optimal working model which predicts biological age (BA) with a set of working biomarkers has not been devised to represent the Korean female population. Accuracy of prediction and applicability are required of an optimal set of commonly assessed biomarkers to provide information on the health status. The goal of this study was to identify a set of biomarkers that represent the aging process and to develop and compare different BA prediction models to elucidate the most fitting and applicable model for providing information on health status in the Korean female population. Using a series of selection processes, eight clinically assessable variables were selected by analyzing relations between 31 clinical variables and chronologic age in 912 normal, healthy individuals among 3642 female participants with ages ranging from 30 to 80 years. The multiple linear regression (MLR), principal component analysis (PCA), and the Klemera-Doubal (KDM) statistical methods were applied to obtain three different sets of BA prediction models. These three models were assessed by calculating and performing the coefficient determinations (r2), regression slopes, effect sizes, pairwise t-tests, and Bland-Altman plots. The BA models were further compared for the applicability by calculating the BAs of clinical risk groups. MLR showed the narrowing effects at the either ends of the age spectrum with greatest effect sizes. PCA showed the greatest degree of dispersion and deviation from the regression center. These MLR and PCA trends were also exhibited by clinically risk groups. In conclusion, the KDM BA prediction model based on the selected biomarkers was found to provide the most reliable and stable results for the practical assessment of BA.

[pmid28203066-27] 27.0 ^27.1 Kang YG et al.: Models for estimating the metabolic syndrome biological age as the new index for evaluation and management of metabolic syndrome. Clin Interv Aging 2017. (PMID 28203066) [PubMed] [DOI] [Full text] PURPOSE: This study aims to propose a metabolic syndrome (MS) biological age model, through which overall evaluation and management of the health status and aging state in MS can be done easily. Through this model, we hope to provide a novel evaluation and management health index that can be utilized in various health care fields. PATIENT AND METHODS: MS parameters from American Heart Association/National Heart, Lung, and Blood Institute guidelines in 2005 were used as biomarkers for the estimation of MS biological age. MS biological age model development was done by analyzing data of 263,828 participants and clinical application of the developed MS biological age was assessed by analyzing the data of 188,886 subjects. RESULTS: The principal component accounted for 36.1% in male and 38.9% in female of the total variance in the battery of five variables. The correlation coefficient between corrected biological age and chronological age in males and females were 0.711 and 0.737, respectively. Significant difference for mean MS biological age and chronological age between the three groups, normal, at risk and MS, was seen (P<0.001). CONCLUSION: For the comprehensive approach in MS management, MS biological age is expected to be additionally utilized as a novel evaluation and management index along with the traditional MS diagnosis.

[pmid29188884-28] 28.0 ^28.1 Zhang W et al.: Model Construction for Biological Age Based on a Cross-Sectional Study of a Healthy Chinese Han population. J Nutr Health Aging 2017. (PMID 29188884) [PubMed] [DOI] OBJECTIVES: Biological age (BA) has been proposed to evaluate the aging status in an objective way instead of chronological age (CA). The purpose of our study is to construct a more precise formula of BA in the cross-sectional study based on a largest-ever sample of our studies. This formula aims at better evaluation of body function and exploring the disciplines of aging in different genders and age stages. METHODS: A total of 1,373 healthy Chinese Han (age range, 19-93 years) were recruited from five cities in China, including 581 males and 792 females. Physical examination, blood routine, blood chemistry, and other lab tests were performed to obtain a total of 74 clinical variables. Then, the principal component analysis (PCA) was used to select variables and estimate BA. The BA formula was further validated in a population with some diseases (n=266), including cardiovascular diseases, type 2 diabetes, kidney diseases, pulmonary diseases, cancer and disorders in nervous system. RESULTS: The BA formula was constructed as follows: BA = 0.358 (pulse pressure) + 0.258 (trail making test) - 11.552 (mitral valve E/A peak) + 26.383 (minimum intima-media thickness) + 31.965 (Cystatin C) + 0.163 (CA) - 3.902. In validation of the formula, BAs of patients were older than those of healthy persons. The BA accelerates faster in the middle-aged population than in the elderly population (>75 years old). CONCLUSION: This BA formula can reflect health condition changes of aging better than CA in a Chinese Han population.

[pmid28958059-29] 29.0 ^29.1 Brown PJ et al.: Biological Age, Not Chronological Age, Is Associated with Late-Life Depression. J Gerontol A Biol Sci Med Sci 2018. (PMID 28958059) [PubMed] [DOI] [Full text] BACKGROUND: The pathophysiology of late-life depression (LLD) is complex and heterogeneous, with age-related processes implicated in its pathogenesis. This study examined the cross-sectional and longitudinal association between depressive symptoms and a baseline multibiomarker algorithm of biological age (BA) that aggregates indicators of inflammatory, metabolic, cardiovascular, lung, liver, and kidney functioning. METHOD: Data were analyzed from 2,776 men and women from the prospective observational Health Aging and Body Composition Study, who had both evaluable chronological age (CA) and BA. Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression (CES-D) scale. RESULTS: A covariate-adjusted regression model showed that BA (B = 0.03, p = .0471) but not CA (B = -0.01, p = .7185) is associated with baseline CES-D scores. The mean baseline BA for individuals with a CES-D ≥ 10 was 1.28 years greater than in those with a CES-D < 10. Comparatively, there is only a 0.05-year difference in mean CA between the two depression groups. A covariate-adjusted longitudinal model found that baseline BA predicts CES-D score at follow-up (B = 0.04, p = .0058), whereas CA does not (B = 0.03, p = .4125). Additionally, an older BA significantly predicted a CES-D ≥ 10 (B = 0.02, p = .032) over a 10-year period. CONCLUSIONS: A multibiomarker index of an older adult's BA outperformed their CA in predicting subsequent increased and clinically significant depressive symptoms. This result supports the evolving view of LLD as a brain disorder resulting from deleterious age-associated changes across numerous physiological systems.

[pmid28977464-30] 30.0 ^30.1 Murabito JM et al.: Measures of Biologic Age in a Community Sample Predict Mortality and Age-Related Disease: The Framingham Offspring Study. J Gerontol A Biol Sci Med Sci 2018. (PMID 28977464) [PubMed] [DOI] [Full text] BACKGROUND: We tested the association of biologic age (BA) measures constructed from different types of biomarkers with mortality and disease in a community-based sample. METHODS: In Framingham Offspring participants at Exams 7 (1998-2001, mean age 62 ± 10) and 8 (2005-2008, mean age 67 ± 9), we used the Klemera-Doubal method to estimate clinical BA and inflammatory BA and computed the difference (∆age) between BA and CA. Clinical ∆age was computed at Exam 2 (1979-1983, mean age 45 ± 10). At Exam 8, we computed measures of intrinsic and extrinsic epigenetic age. Participants were followed through 2014 for outcomes. Cox proportional hazards models tested the association of each BA estimate with each outcome adjusting for covariates. RESULTS: Sample sizes ranged from 2532 to 3417 participants. In multivariable models, each 1-year increase in clinical ∆age at Exam 2 (hazard ratio [HR] = 1.04-1.06, p < 2 × 10-16) and clinical ∆age and inflammatory ∆age at Exam 7 significantly increased the hazards of mortality and incident cardiovascular disease (HR = 1.01-1.05, p < 2 × 10-7), whereas inflammatory ∆age increased the hazards of cancer (HR = 1.01, p < .05). At Exam 8, increased clinical ∆age, inflammatory ∆age, and extrinsic epigenetic age all significantly increased the hazard of mortality (HR = 1.03-1.05, all p < .05); clinical ∆age and inflammatory ∆age increased cardiovascular disease risk (HR = 1.04-1.05, all p < .01); and clinical ∆age increased cancer risk (HR = 1.03, p < .01) when all three BA measures were included in the model. Intrinsic epigenetic age was not significantly associated with any outcome. CONCLUSIONS: Our findings suggest BA measures may be complementary in predicting risk for mortality and age-related disease.

[pmid29340580-31] 31.0 ^31.1 Mamoshina P et al.: Population Specific Biomarkers of Human Aging: A Big Data Study Using South Korean, Canadian, and Eastern European Patient Populations. J Gerontol A Biol Sci Med Sci 2018. (PMID 29340580) [PubMed] [DOI] [Full text] Accurate and physiologically meaningful biomarkers for human aging are key to assessing antiaging therapies. Given ethnic differences in health, diet, lifestyle, behavior, environmental exposures, and even average rate of biological aging, it stands to reason that aging clocks trained on datasets obtained from specific ethnic populations are more likely to account for these potential confounding factors, resulting in an enhanced capacity to predict chronological age and quantify biological age. Here, we present a deep learning-based hematological aging clock modeled using the large combined dataset of Canadian, South Korean, and Eastern European population blood samples that show increased predictive accuracy in individual populations compared to population specific hematologic aging clocks. The performance of models was also evaluated on publicly available samples of the American population from the National Health and Nutrition Examination Survey (NHANES). In addition, we explored the association between age predicted by both population specific and combined hematological clocks and all-cause mortality. Overall, this study suggests (a) the population specificity of aging patterns and (b) hematologic clocks predicts all-cause mortality. The proposed models were added to the freely-available Aging.AI system expanding the range of tools for analysis of human aging.

[pmid30644411-32] 32.0 ^32.1 Mamoshina P et al.: Blood Biochemistry Analysis to Detect Smoking Status and Quantify Accelerated Aging in Smokers. Sci Rep 2019. (PMID 30644411) [PubMed] [DOI] [Full text] There is an association between smoking and cancer, cardiovascular disease and all-cause mortality. However, currently, there are no affordable and informative tests for assessing the effects of smoking on the rate of biological aging. In this study we demonstrate for the first time that smoking status can be predicted using blood biochemistry and cell count results andthe recent advances in artificial intelligence (AI). By employing age-prediction models developed using supervised deep learning techniques, we found that smokers exhibited higher aging rates than nonsmokers, regardless of their cholesterol ratios and fasting glucose levels. We further used those models to quantify the acceleration of biological aging due to tobacco use. Female smokers were predicted to be twice as old as their chronological age compared to nonsmokers, whereas male smokers were predicted to be one and a half times as old as their chronological age compared to nonsmokers. Our findings suggest that deep learning analysis of routine blood tests could complement or even replace the current error-prone method of self-reporting of smoking status and could be expanded to assess the effect of other lifestyle and environmental factors on aging.

[pmid30899733-33] 33.0 ^33.1 ^33.2 ^33.3 Jee H: Selection of a set of biomarkers and comparisons of biological age estimation models for Korean men. J Exerc Rehabil 2019. (PMID 30899733) [PubMed] [DOI] [Full text] Biological age (BA) represents the rate of the senescence with a set of biomarkers. The BA prediction models have not been compared to obtain an optimal BA prediction model with BA biomarkers for Korean men. The study aims to obtain a set of BA biomarkers and compare three of the reported statistical approaches for an optimal BA prediction model. The Korea National Health and Nutrition Examination Surveys data of 2009 to 2011 were used to select six BA biomarkers from 940 healthy subjects aged between 30 to 80 years. The multiple linear regression (MLR), principal component analysis (PCA), and Klemera and Doubal methods (KDM) were used to obtain three BA prediction models. Correlation coefficients (r) with 95% confidence intervals (CI) and regression slopes were assessed. One of the Euro Quality of Life-5 Dimensions, mobility, was compared for feasibility test of each BA models. KDM showed greatest correlation (r=0.88 [P<0.05]) with smallest 95% CI and regression slope (1.00). PCA also showed strong correlation (r=0.79 [P<0.05]) with small 95% CI and regression slope (0.94). MLR (r=0.68 [P<0.05]) showed over and underestimated BA results at the end of the age spectrum. Estimations of BA were most reliable with KDM. The PCA and MLR approaches were comparatively simple to devise for Korean men.

[pmid30999227-34] 34.0 ^34.1 Hastings WJ et al.: Comparability of biological aging measures in the National Health and Nutrition Examination Study, 1999-2002. Psychoneuroendocrinology 2019. (PMID 30999227) [PubMed] [DOI] [Full text] BACKGROUND: Biological processes of aging are thought to be modifiable causes of many different chronic diseases. Measures of biological aging could provide sensitive endpoints for studies of risk factors hypothesized to shorten healthy lifespan and/or interventions that extend it. But uncertainty remains about how to measure biological aging and if proposed measures assess the same thing. METHOD: We tested four proposed measures of biological aging that could be quantified with available data from the National Health and Nutrition Examination Survey (NHANES), Klemera-Doubal method (KDM) Biological Age, homeostatic dysregulation, Levine Method (LM) Biological Age, and leukocyte telomere length. RESULTS: We analyzed data collected during 1999-2002, when all four biological aging meausres could be taken. Participants' KDM biological ages, homeostatic dysregulation levels, LM biological ages, and telomere length were all correlated with their chronological ages. KDM Biological Age, homeostatic dysregulation, and LM Biological Age were all correlated with one another, but these measures were uncorrelated with telomere length. Participants' with more advanced biological aging performed worse on tests of physical, cognitive, and perceptual functioning and reported more limitations to their daily activities and more pain, and rated themselves as being in worse health. In parallel, participants with risk factors for shorter healthy lifespan exhibited more advanced biological aging. In both sets of analyses, effect-sizes tended to be larger for KDM Biological Age, homeostatic dysregulation, and LM Biological Age as compared to telomere length. DISCUSSION: The cellular-level aging biomarker telomere length may measure different aspects of the aging process as compared to the patient-level physiological composite measures KDM Biological Age, homeostatic dysregulation, and LM Biological Age. Studies aiming to test if risk factors accelerate aging or if interventions may slow aging should not treat proposed measures of aging as interchangeable.

[pmid31179487-35] 35.0 ^35.1 ^35.2 ^35.3 Zhong X et al.: Estimating Biological Age in the Singapore Longitudinal Aging Study. J Gerontol A Biol Sci Med Sci 2020. (PMID 31179487) [PubMed] [DOI] BACKGROUND: Biological age (BA) is a more accurate measure of the rate of human aging than chronological age (CA). However, there is limited consensus regarding measures of BA in life span and healthspan. METHODS: This study investigated measurement sets of 68 physiological biomarkers using data from 2,844 Chinese Singaporeans in two age subgroups (55-70 and 71-94 years) in the Singapore Longitudinal Aging Study (SLAS-2) with 8-year follow-up frailty and mortality data. We computed BA estimate using three commonly used algorithms: Principal Component Analysis (PCA), Multiple Linear Regression (MLR), and Klemera and Doubal (KD) method, and additionally, explored the use of machine learning methods for prediction of mortality and frailty. The most optimal algorithmic estimate of BA compared to CA was evaluated for their associations with risk factors and health outcome. RESULTS: Stepwise selection procedures resulted in the final selection of 8 biomarkers in males and 10 biomarkers in females. The highest-ranking biomarkers were estimated glomerular filtration rate for both genders, and the forced expiratory volume in 1 second in males and females. The BA estimates robustly predicted frailty and mortality and outperformed CA. The best performing KD measure of BA was notably predictive in the younger group (aged 55-70 years). BA estimates obtained using a machine learning train-test method were not more accurate than conventional BA estimates in predicting mortality and frailty in most situations. Biologically older people with the same CA as biologically younger individuals had higher prevalence of frailty and 8-year mortality, and worse health, behavioral, and functional characteristics. CONCLUSIONS: BA is better than CA for measuring life span (mortality) and healthspan (frailty). This measurement set of physiological markers of biological aging among Chinese robustly differentiate biologically old from younger individuals with the same CA.

[pmid31566204-36] 36.0 ^36.1 Gaydosh L et al.: Testing Proposed Quantifications of Biological Aging in Taiwanese Older Adults. J Gerontol A Biol Sci Med Sci 2020. (PMID 31566204) [PubMed] [DOI] [Full text] Quantification of biological aging is of interest in gerontology as a means to surveil aging rates in the population and to evaluate the effects of interventions to increase healthy life span. Analysis of proposed methods to quantify biological aging has focused on samples of midlife or mixed-age adults in the West. Research is needed to test whether quantifications of biological aging can differentiate aging rates among older adults and if quantifications of biological aging developed in Western samples can differentiate aging rates in non-Western populations. We conducted analysis of Klemera-Doubal method (KDM) Biological Age and homeostatic dysregulation measures of biological aging developed in the U.S. NHANES and tested in a sample of older Taiwanese adults in the Social Environment and Biomarkers of Aging Study. We conducted analysis of physical and cognitive function and mortality, comparing quantifications of biological aging to a biomarker index based on norms within our analysis sample and to participants' ratings of their own health. Results showed that quantifications of biological aging (a) predicted differences in physical and cognitive function and in mortality risk among Taiwanese older adults and (b) performed as well as a traditional biomarker index and participant self-rated health for prediction of these outcomes.

[pmid31693736-37] 37.0 ^37.1 Parker DC et al.: Association of Blood Chemistry Quantifications of Biological Aging With Disability and Mortality in Older Adults. J Gerontol A Biol Sci Med Sci 2020. (PMID 31693736) [PubMed] [DOI] [Full text] Quantification of biological aging has been proposed for population surveillance of age-related decline in system integrity and evaluation of geroprotective therapies. However, methods of quantifying biological aging have been little studied in geriatric populations. We analyzed three clinical-biomarker-algorithm methods to quantify biological aging. Klemera-Doubal method Biological Age and homeostatic dysregulation algorithms were parameterized from analysis of U.S. National Health and Nutrition Examination Surveys (NHANES) data (N = 36,207) based on published methods. Levine method Biological Age was adapted from published analysis of NHANES data. Algorithms were applied to biomarker data from the Duke Established Populations for Epidemiologic Studies of the Elderly (Duke-EPESE) cohort of older adults (N = 1,374, aged 71-102 years, 35% male, 52% African American). We tested associations of biological aging measures with participant reported Activities of daily living (ADL), instrumental activities of daily living (IADL) dependencies, and mortality. We evaluated the sensitivity of results to the demographic composition of reference samples and biomarker sets used to develop biological aging algorithms. African American and white Duke-EPESE participants with more advanced biological aging reported dependence in more ADLs and IADLs and were at increased risk of death over follow-up through 2017. Effect sizes were similar across algorithms, but were strongest for Levine method Biological Age (per-quintile increase in ADL incidence rate ratio = 1.25, 95% confidence interval [1.17-1.37], IADL incidence rate ratio = 1.23 [1.15-1.32], mortality hazard ratio = 1.12 [1.08-1.16]). Results were insensitive to demographic composition of reference samples, but modestly sensitive to the biomarker sets used to develop biological aging algorithms. Blood-chemistry-based quantifications of biological aging show promise for evaluating the effectiveness of interventions to extend healthy life span in older adults.

[pmid32946548-38] 38.0 ^38.1 Liu Z: Development and Validation of 2 Composite Aging Measures Using Routine Clinical Biomarkers in the Chinese Population: Analyses From 2 Prospective Cohort Studies. J Gerontol A Biol Sci Med Sci 2021. (PMID 32946548) [PubMed] [DOI] [Full text] BACKGROUND: This study aimed to: (i) develop 2 composite aging measures in the Chinese population using 2 recent advanced algorithms (the Klemera and Doubal method and Mahalanobis distance); and (ii) validate the 2 measures by examining their associations with mortality and disease counts. METHODS: Based on data from the China Nutrition and Health Survey (CHNS) 2009 wave (N = 8119, aged 20-79 years, 53.5% women), a nationwide prospective cohort study of the Chinese population, we developed Klemera and Doubal method-biological age (KDM-BA) and physiological dysregulation (PD, derived from Mahalanobis distance) using 12 biomarkers. For the validation analysis, we used Cox proportional hazard regression models (for mortality) and linear, Poisson, and logistic regression models (for disease counts) to examine the associations. We replicated the validation analysis in the China Health and Retirement Longitudinal Study (CHARLS, N = 9304, aged 45-99 years, 53.4% women). RESULTS: Both aging measures were predictive of mortality after accounting for age and gender (KDM-BA, per 1-year, hazard ratio [HR] = 1.14, 95% confidence interval [CI] = 1.08, 1.19; PD, per 1-SD, HR = 1.50, 95% CI = 1.33, 1.69). With few exceptions, these mortality predictions were robust across stratifications by age, gender, education, and health behaviors. The 2 aging measures were associated with disease counts both cross-sectionally and longitudinally. These results were generally replicable in CHARLS although 4 biomarkers were not available. CONCLUSIONS: We successfully developed and validated 2 composite aging measures-KDM-BA and PD, which have great potentials for applications in early identifications and preventions of aging and aging-related diseases in China.

[pmid34038024-39] 39.0 ^39.1 Kuo CL et al.: Genetic associations for two biological age measures point to distinct aging phenotypes. Aging Cell 2021. (PMID 34038024) [PubMed] [DOI] [Full text] Biological age measures outperform chronological age in predicting various aging outcomes, yet little is known regarding genetic predisposition. We performed genome-wide association scans of two age-adjusted biological age measures (PhenoAgeAcceleration and BioAgeAcceleration), estimated from clinical biochemistry markers (Levine et al., 2018; Levine, 2013) in European-descent participants from UK Biobank. The strongest signals were found in the APOE gene, tagged by the two major protein-coding SNPs, PhenoAgeAccel-rs429358 (APOE e4 determinant) (p = 1.50 × 10-72 ); BioAgeAccel-rs7412 (APOE e2 determinant) (p = 3.16 × 10-60 ). Interestingly, we observed inverse APOE e2 and e4 associations and unique pathway enrichments when comparing the two biological age measures. Genes associated with BioAgeAccel were enriched in lipid related pathways, while genes associated with PhenoAgeAccel showed enrichment for immune system, cell function, and carbohydrate homeostasis pathways, suggesting the two measures capture different aging domains. Our study reaffirms that aging patterns are heterogeneous across individuals, and the manner in which a person ages may be partly attributed to genetic predisposition.

[pmid34453631-40] 40.0 ^40.1 Gialluisi A et al.: Exploring domains, clinical implications and environmental associations of a deep learning marker of biological ageing. Eur J Epidemiol 2022. (PMID 34453631) [PubMed] [DOI] Deep Neural Networks (DNN) have been recently developed for the estimation of Biological Age (BA), the hypothetical underlying age of an organism, which can differ from its chronological age (CA). Although promising, these population-specific algorithms warrant further characterization and validation, since their biological, clinical and environmental correlates remain largely unexplored. Here, an accurate DNN was trained to compute BA based on 36 circulating biomarkers in an Italian population (N = 23,858; age ≥ 35 years; 51.7% women). This estimate was heavily influenced by markers of metabolic, heart, kidney and liver function. The resulting Δage (BA-CA) significantly predicted mortality and hospitalization risk for all and specific causes. Slowed biological aging (Δage < 0) was associated with higher physical and mental wellbeing, healthy lifestyles (e.g. adherence to Mediterranean diet) and higher socioeconomic status (educational attainment, household income and occupational status), while accelerated aging (Δage > 0) was associated with smoking and obesity. Together, lifestyles and socioeconomic variables explained ~48% of the total variance in Δage, potentially suggesting the existence of a genetic basis. These findings validate blood-based biological aging as a marker of public health in adult Italians and provide a robust body of knowledge on its biological architecture, clinical implications and potential environmental influences.

[pmid29581467-41] Pyrkov TV et al.: Extracting biological age from biomedical data via deep learning: too much of a good thing?. Sci Rep 2018. (PMID 29581467) [PubMed] [DOI] [Full text] Age-related physiological changes in humans are linearly associated with age. Naturally, linear combinations of physiological measures trained to estimate chronological age have recently emerged as a practical way to quantify aging in the form of biological age. In this work, we used one-week long physical activity records from a 2003-2006 National Health and Nutrition Examination Survey (NHANES) to compare three increasingly accurate biological age models: the unsupervised Principal Components Analysis (PCA) score, a multivariate linear regression, and a state-of-the-art deep convolutional neural network (CNN). We found that the supervised approaches produce better chronological age estimations at the expense of a loss of the association between the aging acceleration and all-cause mortality. Consequently, we turned to the NHANES death register directly and introduced a novel way to train parametric proportional hazards models suitable for out-of-the-box implementation with any modern machine learning software. As a demonstration, we produced a separate deep CNN for mortality risks prediction that outperformed any of the biological age or a simple linear proportional hazards model. Altogether, our findings demonstrate the emerging potential of combined wearable sensors and deep learning technologies for applications involving continuous health risk monitoring and real-time feedback to patients and care providers.

[pmid31388024-42] Rahman SA & Adjeroh DA: Deep Learning using Convolutional LSTM estimates Biological Age from Physical Activity. Sci Rep 2019. (PMID 31388024) [PubMed] [DOI] [Full text] Human age estimation is an important and difficult challenge. Different biomarkers and numerous approaches have been studied for biological age estimation, each with its advantages and limitations. In this work, we investigate whether physical activity can be exploited for biological age estimation for adult humans. We introduce an approach based on deep convolutional long short term memory (ConvLSTM) to predict biological age, using human physical activity as recorded by a wearable device. We also demonstrate five deep biological age estimation models including the proposed approach and compare their performance on the NHANES physical activity dataset. Results on mortality hazard analysis using both the Cox proportional hazard model and Kaplan-Meier curves each show that the proposed method for estimating biological age outperforms other state-of-the-art approaches. This work has significant implications in combining wearable sensors and deep learning techniques for improved health monitoring, for instance, in a mobile health environment. Mobile health (mHealth) applications provide patients, caregivers, and administrators continuous information about a patient, even outside the hospital.

[pmid33744131-43] Raghu VK et al.: Deep Learning to Estimate Biological Age From Chest Radiographs. JACC Cardiovasc Imaging 2021. (PMID 33744131) [PubMed] [DOI] OBJECTIVES: The goal of this study was to assess whether a deep learning estimate of age from a chest radiograph image (CXR-Age) can predict longevity beyond chronological age. BACKGROUND: Chronological age is an imperfect measure of longevity. Biological age, a measure of overall health, may improve personalized care. This paper proposes a new way to estimate biological age using a convolutional neural network that takes as input a CXR image and outputs a chest x-ray age (in years) as a measure of long-term mortality risk. METHODS: CXR-Age was developed using CXR from 116,035 individuals and validated in 2 held-out testing sets: 1) 75% of the CXR arm of PLCO (Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial) (N = 40,967); and 2) the CXR arm of NLST (National Lung Screening Trial) (N = 5,414). CXR-Age was compared to chronological age and a multivariable regression model of chronological age, risk factors, and radiograph findings to predict all-cause and cardiovascular mortality with a maximum 23 years and 13 years of follow-up, respectively. The primary outcome was observed mortality; results are provided for the testing datasets only. RESULTS: In the PLCO testing dataset, a 5-year increase in CXR-Age carried a higher risk of all-cause mortality than a 5-year increase in chronological age (CXR-Age hazard ratio [HR]: 2.26 [95% confidence interval (CI): 2.24 to 2.29] vs. chronological age HR: 1.77 [95% CI: 1.75 to 1.78]; p < 0.001). A similar pattern was found for cardiovascular mortality (CXR-Age cause-specific HR: 2.45 per 5 years [95% CI: 2.34 to 2.56] vs. chronological age HR: 1.82 per 5 years [95% CI: 1.74 to 1.90]). Similar results were seen for both outcomes in the NLST external testing dataset. Adding CXR-Age to the multivariable model resulted in significant improvements for predicting both outcomes in both testing datasets (p < 0.001 for all comparisons). CONCLUSIONS: Based on a CXR image, CXR-Age predicted long-term all-cause and cardiovascular mortality.

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]