Impact of Early Personal History Characteristics on the Pace of Aging

Implications for Clinical Trials of Therapies to Slow Aging and Extend Healthspan

 

Belsky DW*, Caspi A, Kraus W, Cohen HJ, Ramrakha S, Poulton R, Moffitt TE. Impact of early personal-history characteristics on the Pace of Aging: Implications for clinical trials of therapies to slow aging and extend healthspan. Aging Cell, published online April 12, 2017.

 

 

Summary

Therapies to extend healthspan are poised to move from laboratory animal models to human clinical trials. Translation from mouse to human will entail challenges, among them the multifactorial heterogeneity of human aging. To inform clinical trials about this heterogeneity, we report how humans’ pace of biological aging relates to personal-history characteristics. Because geroprotective therapies must be delivered by midlife to prevent age-related disease onset, we studied young-adult members of the Dunedin Study 1972–73 birth cohort (n = 954). Cohort members’ Pace of Aging was measured as coordinated decline in the integrity of multiple organ systems, by quantifying rate of decline across repeated measurements of 18 biomarkers assayed when cohort members were ages 26, 32, and 38 years. The childhood personal-history characteristics studied were known predictors of age-related disease and mortality, and were measured prospectively during childhood. Personal-history char- acteristics of familial longevity, childhood social class, adverse childhood experiences, and childhood health, intelligence, and self-control all predicted differences in cohort members’ adult- hood Pace of Aging. Accumulation of more personal-history risks predicted faster Pace of Aging. Because trials of anti-aging therapies will need to ascertain personal histories retrospectively, we replicated results using cohort members’ retrospective per- sonal-history reports made in adulthood. Because many trials recruit participants from clinical settings, we replicated results in the cohort subset who had recent health system contact accord- ing to electronic medical records. Quick, inexpensive measures of trial participants’ early personal histories can enable clinical trials to study who volunteers for trials, who adheres to treatment, and who responds to anti-aging therapies.

 

 

Subgroups of normally distributed human pace of aging relevant to design of trials for healthspan-extension therapies. Pace of aging is the rate of coordinated decline in the integrity of bodily systems occurring with advancing chronological age. We showed that variation in the pace of aging could be quantified already in individuals still too young to have age-related disease by tracking changes in biomarkers of organ system functioning over time (Belsky et al., 2015a). We earlier reported that individuals whose physiology changed more slowly with the passage of chronological time (light blue segment of distribution) experienced better physical and cognitive functional outcomes in aging, and also showed fewer subjective signs of aging. The opposite was true of those whose physiology changed more rapidly (red segment of distribution). Both slower-aging and faster-aging population segments are needed in research to develop healthspan-extension therapies. Slower-aging populations may provide clues to novel therapeutic targets. Faster-aging populations are those who therapies must benefit.

 

 

 

 

 

Family and childhood characteristics are associated with Study members’ Pace of Aging from age 26 to 38 years. Figure cells graph associations between six family and childhood characteristics (x-axis variables) and Study members’ Pace of Aging measured from changes in 18 biomarkers measured across ages 26, 32, and 38 years (y-axis). Age of longest-lived grandparent was measured from reports by Study members’ parents. Childhood social class, exposure to adverse childhood experiences, childhood health, childhood IQ, and childhood self-control were assessed using previously established methodology applied to archival Dunedin Study records including examinations and testing, reports by parents and teachers, clinician ratings, and direct observations. Figures show ‘binned’ scatterplots in which each plotted point reflects average x- and y-coordinates for “bins” of approximately 20 Study members. Regression lines and effect size estimates were estimated from the original, unbinned data. 

 

Cumulative prospectively assessed personal-history risks in Study members with slow, average, and fast Pace of Aging. Panel A graphs density plots of cumulative risk scores for Study members with slow, average, and fast Pace of Aging. The cumulative risk score reflects total burden of risk across six personal-history characteristics (grandparent longevity, family social class during childhood, adverse childhood experiences, childhood IQ score, childhood self-control, and childhood health). For each characteristic, values were standardized to a T-distribution (M = 50, SD = 10) with high scores reflecting increased risk (e.g., shorter-lived grandparents, lower childhood social class). Standardized values were summed to calculate the cumulative risk score. Thus, the expected cumulative risk level was 300. The graph shows that two-thirds of the slow-aging group had below this expected level of risk. In contrast, less than one-third of the fast-aging group did. Panel B graphs proportions of Study members with slow, average, and fast Pace of Aging (see Fig. 1) who were classified as high risk on 0, 1, or 2 or more of the six characteristics. High-risk classifications were for having short-lived grandparents (no grandparent survived past age 80 years), growing up in a low social class family, exposure to four or more adverse childhood experiences, childhood IQ score ≤1 SD below the population mean (a score of 85 or below), childhood self-control score ≤1 SD below the population mean, and childhood health score ≤1 SD below the population mean. The graph shows that most slow-aging Study members had no high-risk classifications. In contrast, more than 40% of the fast-aging Study members were classified as high risk on multiple family and childhood characteristics.

 

Proportions of slow, average, and fast Pace of Aging Study members classified as high risk on 0, 1, or 2 or more family and childhood characteristics based on contemporaneous assessments conducted in adulthood. Risk factors were having short-lived grandparents (no grandparent survived past age 80 years), retrospective report by the Study member that their parents held low-status occupations during the Study member’s childhood, retrospective report of exposure to four or more adverse childhood experiences, not holding any educational credential, and being rated by an examining nurse as having low levels of the personality trait conscientiousness. Panel A graphs results for the full cohort. The pattern is the same as when risk was classified from assessments during childhood. Most slow-aging Study members were not classified as high risk on any family or childhood characteristic. In contrast, more than 40% of the fast-aging Study members were classified as high risk on multiple family and childhood characteristics. Panels B and C repeat the analysis for subsamples of cohort members with recent contacts with the healthcare system and who may reflect the population most accessible to recruitment into clinical trials. Panel B graphs results for Study members with a recent prescription fill. Panel C graphs results for Study members with a recent hospital admission (excluding for pregnancy-related services).

For the full manuscript, click here

For the supplementary information, click here