Time-varying consideration of health behaviours explains over 90% of the inequality in mortality associated with socioeconomic status

Rebecca Harounoff, Danyal Sarwar, María Pérez‐Ortiz, Eric J. Brunner, John Shawe‐TaylormedRxiv 2024 · DOI

Abstract

Socioeconomic status (SES) is a well-established predictor of mortality, but the extent to which health behaviors mediate this link is often underestimated. This study utilizes 32 years of longitudinal data from the Whitehall II cohort to demonstrate that accounting for time-varying changes in smoking, diet, alcohol, and physical activity explains up to 91% of the socioeconomic gradient in mortality. These results suggest that lifestyle factors almost entirely account for the mortality gap between social classes when cumulative exposure is properly modeled.

Introduction

Epidemiological research has long identified a "social gradient" in health, where individuals in lower socioeconomic positions face higher mortality risks. While health behaviors like smoking and physical inactivity are known contributors, previous studies often found they explained only 40–70% of this inequality. The authors hypothesized that these estimates were limited by the use of "snapshot" baseline data, which fails to capture how behaviors evolve over decades. This study aims to determine if a longitudinal, time-varying approach provides a more complete explanation of the SES-mortality link.

Methods

The study analyzed data from the Whitehall II cohort, a longitudinal study of 10,308 British civil servants (initially aged 35–55) followed for 32 years. Socioeconomic position was defined by employment grade. The researchers used Cox proportional hazards regression to calculate hazard ratios for mortality, adjusting for health behaviors in two ways:

  1. Baseline Analysis: Using only habits recorded at the start of the study.
  2. Longitudinal Analysis: Incorporating data from 11 waves of collection, treating behaviors as time-varying covariates.

Additionally, the study explored nonlinear interactions between behaviors (e.g., smoking and exercise) and expanded dietary metrics to include factors like meat, cheese, and fat consumption, aligning with modern nutritional guidelines.

Results

The transition from baseline to longitudinal modeling significantly increased the explained portion of the mortality gap.

Comparison of percentage attenuation in mortality risk explained by health behaviors at baseline versus longitudinal assessment.

As illustrated in Figure 1, the "percentage attenuation"—the degree to which health behaviors statistically explain the SES-mortality association—nearly doubled for most factors when moving from a baseline to a longitudinal model. While baseline habits explained 51% of the gap in the 2021 follow-up, the longitudinal model accounted for 87%.

The explanatory power was further enhanced by refining the dietary assessment. By moving from a basic diet score to a more comprehensive one, the model explained 91% of the association between employment grade and mortality.

Hazard ratios and percentage attenuation including longitudinal phases and updated diet scores.

Furthermore, nonlinear modeling of smoking and exercise proved to be a superior predictor of mortality, providing 13% more explanatory power than treating these behaviors as independent, linear factors.

Discussion

The finding that over 90% of the socioeconomic mortality gap can be explained by health behaviors is a significant shift from previous estimates. It suggests that the "unexplained" portion of health inequality is largely a product of data limitations rather than unobserved biological or environmental factors.

However, the authors caution that "reverse causation" may be at play: poor health in lower SES groups may lead to a decline in healthy behaviors, which then accelerates mortality. This highlights a complex feedback loop where social position influences health, which in turn influences behavior. The results underscore the importance of understanding why health behaviors are socially patterned, suggesting that interventions targeting these behaviors in lower SES groups could be highly effective in reducing mortality inequalities.

Conclusion

By accounting for 32 years of behavioral changes, this study shows that lifestyle factors almost entirely mediate the relationship between socioeconomic status and mortality. These findings emphasize the necessity of longitudinal data in public health research and suggest that addressing the social determinants of health behaviors is key to closing the mortality gap.

Summary updated

Abstract

Background: Applying survival analysis techniques to epidemiological inference within research into ageing offers opportunities to estimate the association between exposure and outcome in longitudinal data. This study used Cox regression to investigate how socioeconomic inequality in mortality can be explained by exposure to various factors including smoking, diet, alcohol and physical activity. This study seeks to complement and extend previous work which found that the contribution of the socioeconomic gradient to inequalities in health was underestimated by baseline analysis. Methods: Data was obtained from Whitehall II, a British longitudinal cohort study, which investigated social determinants of health. Analysis is based on 11 waves of data collected over 32 years on 10,308 civil servants aged between 35 and 90. Socioeconomic position was defined by baseline employment grade (1-3). During the follow-up 2,427 participants died. Extensive experimental analysis was conducted using a vast number of health behaviours. Cox regression produced an age-and-sex-adjusted hazard ratio for the socioeconomic inequality in mortality. Health behaviours (smoking, physical activity, alcohol consumption, and diet) were then added as covariates to determine the extent to which they statistically explain this inequality, and how this differed from the last similar analysis from 2009. This was done at baseline and longitudinally. The health behaviours were then combined linearly, nonlinearly and new health behaviours were added. Results: Adding the above health behaviours as covariates statistically explained the socioeconomic gradient in mortality at baseline from 42% to 2009, to 51% to 2021. Longitudinal consideration increased the explanatory power, when all health behaviours were added as time-varying covariates, from 51% to 87%. Adding more variables in the form of a more comprehensive diet score statistically explained the gradient further, to 91%. The nonlinear model of smoking and exercise most accurately predicted mortality and had a 13% higher explanatory power when explaining the gradient compared to the linear model in longitudinal data. Conclusion: In the Whitehall II study, socioeconomic position and mortality showed an association. There is a gain in explanatory power of the set of health behaviours at baseline when follow-up is extended by 12 years, from 42% to 51%. When changes in behaviour over the 32 years of follow-up were also accounted for, this association was now significantly explained by over 90%, compared with 51% when considered at baseline. We suggest that reverse causation is partly responsible for the almost complete explanation of the social gradient in mortality by health behaviours. These results would therefore lead us to question why health behaviours are socially patterned in the way that has been observed, which would be significant for targeting health behaviours in lower socioeconomic statuses.
BibTeX
@misc{harounoff2024time,
  author = {Rebecca Harounoff and Danyal Sarwar and María Pérez‐Ortiz and Eric J. Brunner and John Shawe‐Taylor},
  title = {{Time-varying consideration of health behaviours explains over 90\% of the inequality in mortality associated with socioeconomic status}},
  year = {2024},
  doi = {10.1101/2024.10.01.24314725},
  howpublished = {medRxiv},
}
Figure