December 10, 2019

Education and Health: What Does the Literature Tell Us?

By Kate Raphael

Across many disciplines, greater educational attainment is closely associated with health. People who have obtained more schooling are significantly likelier to live longer, healthier lives. The mediating pathways that facilitate this connection are myriad and complex. A number of pathways have been proposed, including ones involving health literacy and behaviors, employment opportunities, and social and psychological factors. The degree and direction of causality and the details of these proposed mechanisms remains unclear.

What is clearer, is that increasing education is usually associated with a range of benefits, many of them health-related, in addition to more indirect health benefits like earning potential. These benefits become even more important when we consider the fact that the gradient in health outcomes by educational attainment has steepened in the U.S. over the last four decades. This trend is associated with a widening health gap between Americans with high and low levels of education and a widening disparity in mortality rates. Strikingly, people with at least some college education have a mortality rate less than half that of those without any college education.

We need more research to understand if these relationships are causal. Part of why we lack this understanding is because the randomization of educational interventions presents significant logistical and ethical challenges. Even when a degree of randomness is achieved in the experimental design, small sample sizes and limited follow-up place limitations on the findings. As a result, the most rigorous studies of educational interventions are “natural experiments” that exploit changes in policy across place and time. Commonly studied policies of this type are “compulsory schooling laws,” which mandate a minimum level of educational attainment for children.

Even with these policies in the arsenal of studiable interventions, the body of literature on educational interventions is relatively weak and narrow. We took a deep dive into the most recent reviews on educational interventions and their effects on health outcomes. Here’s what we found.

Education is closely associated with improvements in specific health outcomes, but only for certain populations and timespans. Indeed, educational appears to have a significantly larger effect on men. Even when educational attainment is associated with improvements in health, they only appear in a narrow range of health outcomes. Compulsory schooling laws have robust effects on educational attainment and earning potential but minimal effect on outcomes other than mortality.

Most of the literature focuses on improvements in quantity rather than quality of schooling (i.e., more years of education rather than better education). However, limited evidence suggests that interventions targeting school quality have the greatest potential to improve health and behavioral outcomes relating to physical activity.

The authors of the literature we reviewed called for deeper investigation in a number of areas. First, we need to examine a wider range of health outcomes and give particular attention to the heterogeneous effects of education. Since education appears to have varying effects for different populations and in different contexts, it is critical we understand how and why (you can watch a video of an expert discussing these nuances at our meeting in Detroit). Second, we need a more nuanced consideration of the interactions between determinants of education and health, in addition to the interactions with other drivers of health. For example, we understand that socioeconomic status is related to educational attainment, and that both are associated with health, but we don’t have a clear understanding of how.

But, it’s the interventions themselves that also warrant closer scrutiny. Authors of the studies we reviewed called for evaluation of intervention components separately and in combination, as well as evaluation of intervention effects at a more granular level. It’s rare for interventions to be modified and re-implemented, and the opportunity for iteration does not always present itself, but when it does, this strategy could allow for finer tuning and more precise understanding of mechanisms. Even when this is not possible, the authors called for greater reliance on and investigation of theory in the design and implementation of interventions.

Finally, we need more research in a wider range of settings. It’s already been demonstrated that the effects of education are population- and context-dependent, so the conditions of study should also be varied. Authors called for an exploration of effects of educational interventions at higher and lower levels of education, and a greater emphasis on the quality of schooling rather than sole focus on years of schooling completed.

We should not expect the results of educational interventions to be similar (or even positive) across different communities and settings, and this is precisely why we need more and better means of investigation. All these potential research refinements are challenged by the fact that many health outcomes of education are long-term, showing up decades after intervention. This requires consistent investment over an extended period to refine our understanding of education’s health effects.

Want more? You can read the entirety of our literature review here.