An objective metric of individual health and aging for population surveys

Abstract

Background: We have previously developed and validated a biomarker-based metric of overall health status using Mahalanobis distance (DM) to measure how far from the norm of a reference population (RP) an individual's biomarker profile is. DM is not particularly sensitive to the choice of biomarkers; however, this makes comparison across studies difficult. Here we aimed to identify and validate a standard, optimized version of DM that would be highly stable across populations, while using fewer and more commonly measured biomarkers.

Methods: Using three datasets (the Baltimore Longitudinal Study of Aging, Invecchiare in Chianti and the National Health and Nutrition Examination Survey), we selected the most stable sets of biomarkers in all three populations, notably when interchanging RPs across populations. We performed regression models, using a fourth dataset (the Women's Health and Aging Study), to compare the new DM sets to other well-known metrics [allostatic load (AL) and self-assessed health (SAH)] in their association with diverse health outcomes: mortality, frailty, cardiovascular disease (CVD), diabetes, and comorbidity number.

Results: A nine- (DM9) and a seventeen-biomarker set (DM17) were identified as highly stable regardless of the chosen RP (e.g.: mean correlation among versions generated by interchanging RPs across dataset of r = 0.94 for both DM9 and DM17). In general, DM17 and DM9 were both competitive compared with AL and SAH in predicting aging correlates, with some exceptions for DM9. For example, DM9, DM17, AL, and SAH all predicted mortality to a similar extent (ranges of hazard ratios of 1.15-1.30, 1.21-1.36, 1.17-1.38, and 1.17-1.49, respectively). On the other hand, DM9 predicted CVD less well than DM17 (ranges of odds ratios of 0.97-1.08, 1.07-1.85, respectively).

Conclusions: The metrics we propose here are easy to measure with data that are already available in a wide array of panel, cohort, and clinical studies. The standardized versions here lose a small amount of predictive power compared to more complete versions, but are nonetheless competitive with existing metrics of overall health. DM17 performs slightly better than DM9 and should be preferred in most cases, but DM9 may still be used when a more limited number of biomarkers is available.

Keywords: Allostatic load; Biomarkers; Mahalanobis distance; Physiological dysregulation; Population composition; Self-assessed health.

Fig. 1

Stability of dysregulation scores across populations. For each dataset or a combined set (All), we performed correlations between dysregulation scores (DM) calculated using the study population (columns) as its own reference population or another dataset as the reference population (lines). Correlations were calculated for the three biomarker sets: 9 biomarker-set (DM9), 17-set (DM17), and the entire set (DM31). Mean Pearson correlation coefficients (r) are indicated for each set and ellipses indicate correlations visually, i.e. darker and narrower when stronger

Fig. 2

Relationships between health metrics and aging correlates in the InCHIANTI dataset. Estimations (points) together with 95% confidence intervals (CIs; segments) are plotted for mortality, the number of frailty criteria, cardiovascular diseases (CVD), diabetes, and the number of comorbidities (see text for details). Results are based on regression models adjusting for: (1) age and sex (solid lines); (2) age, sex, as well as physical and cognitive functions (dashed lines); or (3) age, sex, and socioeconomic status (dotted lines). For ease of comparison, each metric was standardized, i.e. divided by its standard deviation. Different colors refer to different health metrics and estimates are indicated on the right. Significant results are plotted in bold, with asterisks indicating the significance level (***p < 0.001; **p < 0.01; *p < 0.05). Abbreviations: AL, allostatic load; DM9, 9-set dysregulation score (DM); DM17, 17-set DM; DM31, 31-set DM; SAH, self-assessed health

Fig. 3

Relationships between health metrics and aging correlates in the WHAS dataset. Estimations (points) together with 95% confidence intervals (CIs; segments) are plotted for mortality, the number of frailty criteria, cardiovascular diseases (CVD), diabetes, and the number of comorbidities (see text for details). Results are based on regression models adjusting for: (1) age (solid lines); (2) age as well as physical and cognitive functions (dashed lines); or (3) age and socioeconomic status (dotted lines). For ease of comparison, each metric was standardized, i.e. divided by its standard deviation. Different colors refer to different health metrics and estimates are indicated on the right. Significant results are plotted in bold, with asterisks indicating the significance level (***p < 0.001; **p < 0.01; *p < 0.05). Abbreviations: AL, allostatic load; DM9, 9-set dysregulation score (DM); DM17, 17-set DM; DM31, 31-set DM; SAH, self-assessed health

Fig. 4

Comparison of predictive performance across health metrics for various health outcomes. Bars represent the means of estimated regression coefficients for the three different analyses performed (see Figs. 2, 3) in InCHIANTI (blue) and WHAS (red), with the corresponding 95% confidence interval. For ease of comparison across health outcomes, we used the log-hazard and log-odds ratios. Numbers above the bars indicate the number of significant associations out of three analyses. Abbreviations: Comorb., number of comorbidities; DM9, 9-set dysregulation score (DM); DM17, 17-set DM; DM31, 31-set DM