Intended for healthcare professionals

Analysis Health, Wealth, and Profits

Economic consequences of better health: insights from clinical data

BMJ 2020; 370 doi: (Published 20 July 2020) Cite this as: BMJ 2020;370:m2186

Health, wealth, and profits

Click here to read the complete collection

  1. Osondu Ogbuoji, assistant professor of global health1,
  2. Sebastian Vollmer, professor of development economics2,
  3. Dean T Jamison, Edward A Dickson professor emeritus3,
  4. Till Bärnighausen, director4 5 6
  1. 1Center for Policy Impact in Global Health, Duke Global Health Institute, Duke University, Durham, NC, USA
  2. 2University of Göttingen, Department of Economics and Centre for Modern Indian Studies, Göttingen, Germany
  3. 3Institute for Global Health Science, University of California, San Francisco, San Francisco, USA
  4. 4Heidelberg Institute of Global Health (HIGH), Faculty of Medicine and University Hospital, Heidelberg University, Heidelberg, Germany
  5. 5Department of Global Health and Population, Harvard TH Chan School of Public Health, Boston, MA, USA
  6. 6Africa Health Research Institute, Durban, South Africa
  1. Correspondence to: Osondu Ogbuoji osondu.ogbuoji{at}

Osondu Ogbuoji and colleagues discuss the links between economic outcomes and health and how to improve our understanding of the effects of specific diseases

Most people will agree that healthier individuals and populations are likely to generate higher economic output than unhealthy ones. In this article we examine the evidence supporting a causal link between better health and future economic wellbeing. This link is plausible for several reasons. Firstly, healthier individuals are more productive at work and less likely to be absent from work. Secondly, health promotes the accumulation of human capital (that is, knowledge, skills, and experience)—healthier children are more likely to go to school, to learn, and to develop to their full human and economic potential. Thirdly, health motivates the accumulation of physical capital (that is, tools of production such as buildings, machines, and technology) because better health implies greater life expectancy, which increases the incentives to save for retirement. These savings will in turn be available for investment in physical capital. Fourthly, declining child mortality often leads to fertility reductions as parents realise that they need fewer children to ensure comfort in old age. However, this reduction in fertility typically lags behind the child health improvement that triggers it. Consequently, the ratio of working age adults to the young and the old who depend on them increases, thus providing favourable conditions for economic growth: the so called demographic dividend.1

A combination of historical, macroeconomic (using aggregate data at state or country level), and household studies provides ample evidence of the association between health and economic outcomes, as well as insights into potential pathways explaining these associations (box 1).456 Yet causal (as opposed to associational) evidence remains sparse and inconclusive, particularly at the macroeconomic level. The microeconomic research (using data on individuals and households) on the causal link between health and economic outcomes has focused mainly on the economic consequences of two particular aspects of health: quasi-experimental evidence on in utero exposure to mothers’ health; and both experimental and quasi-experimental evidence on reductions in various forms of malnutrition, such as stunting or nutrient deficiencies.78910 Largely lacking, however, are empirical individual level studies establishing the economic consequences of specific diseases such as malaria, hypertension, diabetes, pneumonia, diarrhoea, or depression, and the net economic returns to specific medical interventions. A notable exception is HIV treatment.

Box 1

Economic consequences of better health: pathways

Several authors have proposed theoretical frameworks to categorise the most important causal pathways from health to economic outcomes. Some common frameworks include the health to wealth framework by Bloom and Canning,1 subsequently expanded by Bärnighausen et al,2 and the social drift hypothesis for mental illness.3 We highlight some of the common pathways below.

  • Outcome related productivityHealthier children learn and perform better at school, while healthier adults perform better at work and earn higher salaries. Another example is the social drift hypothesis: a major mental health disorder, such as schizophrenia, prevents a person from earning a stable income, leading to impoverishment

  • Behaviour related productivity—Better health can induce individuals to change long term behaviours that affect economic outcomes, such as fertility, education, and saving for future spending. For instance, increased life expectancy should increase educational attainment in a community. In turn, higher educational attainment will increase wages for individuals and boost economic development in the community

  • Healthcare cost and care related productivity—Disease implies a need for treatment, which in turn implies spending on healthcare and loss of productive time for patients and care givers. Prevention can reduce monetary and time losses from ill health. For instance, good glycaemic control among people with diabetes reduces the costs of care to treat diabetic complications

  • Community economic externalities—Healthier communities tend to attract more investment, such as foreign direct investment, and are more likely to provide the conditions for strong economic growth, such as social and political stability


In utero health and economic wellbeing in adulthood

The in utero period was first linked to adult health outcomes by Barker.1112 Almond subsequently proposed an extension of Barker’s hypothesis to explain the effect of in utero conditions on adult economic outcomes. Using quasi-experimental variation in pre-birth exposure to the 1918-19 Spanish flu in the United States, he estimated that the flu led to reduced educational attainment, and subsequently income and socioeconomic status in adulthood.13 Other studies have found similar effects for Brazil, Sweden, Switzerland, and Taiwan.14151617 However, a meta-regression of census data showed that these findings were not consistent with the experience of most other affected countries.18 Almond and coauthors have investigated other natural experiments, such as exposure to Chernobyl’s radioactive fallout in Sweden and temporal variation in Ramadan observance.1920 The Chernobyl study found that children exposed to radiation fallout performed worse than others in mathematics.19 The Ramadan study showed that children whose mothers observed the Ramadan fast during their pregnancy exhibited higher rates of learning disabilities.20

Effects of childhood diets

On average, healthy children have better educational outcomes than unhealthy children, and this educational advantage translates to higher earnings over their lifetime.78921 Experiments around the world found significantly higher maths and language scores and attendance rates in schools that operated a school feeding programme compared with schools that did not.2223242526 A randomised trial in a region of India with widespread iron deficiency anaemia found that maths and language scores were higher among children who regularly participated in a school feeding programme that included salt fortified with iron.27 Another study in Guatemala found that children who were randomly assigned to receive nutritional supplements in their first two years of life earned 46% more in adulthood than their counterparts.10

Effects of adult diets on productivity at work

Quantity and quality of food intake also affect the productivity levels of workers. In separate studies done at different times and in different countries, better diets had a positive effect on the productivity of farm workers when measured as harvest yield or wages.28293031 In some studies, increased energy intake among energy deficient adults led to increased activity levels, which increased farm productivity,283233 while in other studies it led to decreased farm productivity or no observable change.3435 Workers with micronutrient deficiencies, such as iron deficiency anaemia, show improvements in aerobic capacity, production efficiency, and work output, once these specific deficiencies are dealt with.36 By contrast, while some statistical associations exist,3738 there is no evidence that obesity affects long term economic outcomes.

Intergenerational and multigenerational effects

Intergenerational (such as from parents to children) and multigenerational (such as from grandparents to grandchildren) effects of health on economic outcomes have also been studied. While relatively new, this research is increasing, partly because of the growing availability of long term data and increased sophistication of research methods. However, current evidence supporting any link is mostly associational rather than causal.

As expected, the proposed causal links between health status in one generation and economic outcomes in a future generation are complex. In separate reviews of the literature in economics, demography, and sociology, Currie39 and Palloni40 conclude that early child health has a role in the intergenerational inequalities in economic status. Plausible causal pathways run from parental health status to a child’s health status, which in turn affects the child’s economic outcomes, and from parental health status to parental economic outcomes, which then influences child health. While few causal studies have shown these pathways end to end, many provide causal evidence for steps along the chain.39404142

Specific treatments: example of HIV

Untreated HIV leads to illness and eventually premature death, and so it is plausible that HIV treatment has important economic consequences. Quasi-experimental studies over the past 15 years, mostly in sub-Saharan Africa, have established that HIV treatment increases employment43444546 and worker productivity46 and decreases absenteeism.45 These effects are greater if treatment is started soon after HIV infection occurs rather than after the disease has progressed.4748

Economic effects of treatment, however, may also manifest in unexpected directions. For instance, HIV treatment in rural South Africa decreased food security for about three years after treatment initiation, despite positive effects of HIV treatment on employment in the same population.49 The likely explanation is that the financial burden of treatment is immediate and large, while the economic benefits, such as boosting employment, lag treatment initiation by three to five years.49

Implications for clinical research

Overall, evidence showing a causal effect of good health status on economic outcomes is strong. Studies investigating the fetal origins hypothesis have shown that exposure to harmful conditions in utero can greatly reduce future earnings potential. Also, quality of diet in early childhood can have a noticeable effect on cognitive development, academic achievement, and future earnings potential, while the quality of diet in adults directly affects productivity at work. However, there is little causal evidence of the economic consequences of specific diseases and treatments. These knowledge gaps are important because they impede the optimal allocation of resources for healthcare and scientific discovery, and the design of interventions to support economic growth and human development.

A clinical research agenda on economic consequences of specific treatments could begin with existing knowledge—for instance, through systematic reviews and development of evidence and gap maps (box 2).50 For example, what do we know about the long term effects of headaches or chronic back pain on productivity? What are the effects of dementia on grandchildren’s educational attainment? What do we know about the economic effects of treatment of community acquired pneumonia? These types of evidence bases are different from standard health economic estimates of the direct and indirect costs of care; they focus on the downstream effects of health on economic and social outcomes. A good heuristic for selecting diseases to focus on first when doing this type of evidence synthesis are diseases that affect large numbers of people and persist for long durations. These diseases are likely to have major economic effects both at the individual and the population level—for example, chronic diseases such as hypertension, diabetes, and depression.

Box 2

Evidence and gap maps and their potential contributions to research

What are evidence and gap maps?

An evidence and gap map (EGM) visually summarises all available evidence (meta-analysis, systematic reviews, and primary studies) about a particular subject50 and identifies the important knowledge gaps.

Why are they important?

EGMs promote the use of all available evidence to guide research investment and health policy. EGMs also prevent duplication of studies and so ensure more efficient use of research funds.

How can EGMs contribute to establishing the economic consequences of diseases and treatments?

EGMs integrate evidence on the economic consequences of diseases and treatments from different scientific fields. The process of developing EGMs can bring together experts from clinical medicine, epidemiology, and economics. As a result, EGMs ensure that data scattered around different databases, knowledge banks, and repositories are consolidated in an easily accessible form. The visual summary of evidence gaps identified helps to focus researchers on the areas where their activities can add greatest value. EGMs can also show the quality of the evidence—for example, through colour coding—adding important nuance to the evidence synthesis.


As a second step, existing clinical research infrastructures—in particular, clinical trials and cohorts—could be used to learn about economic consequences of specific diseases and treatments (table 1). Adding economic endpoints to clinical trials of novel treatments generates two powerful analytical opportunities. Firstly, it allows empirical estimation of the causal effects of the treatment on economic outcomes in intent-to-treat analyses. Secondly, it enables estimation of the economic effects of the disease that the treatment affects, using instrumental variable analyses. Adding economic outcome assessments to clinical cohorts will enable the estimation of the economic effects of diseases and routine healthcare using quasi-experimental analytic techniques. One strong quasi-experimental approach to analysing clinical cohort data is regression discontinuity,51 which exploits the fact that clinical medicine assigns many treatments by applying a threshold to continuously measured indicators, such as systolic blood pressure, low density lipoprotein cholesterol, or haemoglobin A1c.5253

Table 1

Studies estimating the economic consequences of diseases and treatments

View this table:

A third step in the clinical research agenda on economic consequences of specific treatments is to exploit previous clinical randomised controlled trials. In this approach, cohorts of individuals that participated in trials concluded years (or decades) ago are traced and their economic outcomes are assessed. This allows sufficient time for any long term economic consequences of the intervention to appear. An example is a study in Guatemala that assessed the economic outcomes of adults (aged 25 to 42 years) who had received a nutritional intervention in a randomised controlled trial when they were children (aged 0 to 7 years).10 The long intervals between the original and follow-on studies in this type of research can make it difficult to trace everyone who participated in the original study—the Guatemalan study reached 1424 (60%) of the 2392 people who participated in the original study. Nevertheless, this approach may generate powerful evidence on long term causal effects of treatments on economic outcomes.


There is convincing evidence on the causal effects of health on economic outcomes for a few medical domains. However, for many important diseases and treatments we lack strong evidence. We see major opportunities for clinical research to contribute towards filling these knowledge gaps. Clinical research could complement the existing economic literature, providing greater precision and detail in the evidence on the effects of healthcare beyond health. The funding needs for these scientific advances would not be large because already funded clinical trials and cohorts could be used for this research.

Key messages

  • We have strong evidence that good health in early life leads to better economic outcomes later on

  • The evidence supporting this causal link is mostly related to the economic effects of better diets in utero and early childhood

  • We lack strong evidence on the economic consequences of most specific diseases and treatments over the life course

  • Clinicians and researchers can generate powerful new insights on the economic consequences of specific diseases and treatments by incorporating economic outcomes in clinical trials and identifying and exploiting natural experiments in routine clinical data


  • Contributors and sources: OO is a physician and public health researcher who has studied and reported on the effect of diseases and health states on household economics. SV is a development economist who has conducted extensive studies on the long term impact of natural experiments on health and future economic outcomes. DJ is an economist who has studied and reported widely on the health and economic benefits of disease control and the economic benefits of health. TB is a physician, health economist, and health systems researcher who has done multiple studies on the impact of HIV/AIDS and its treatments on long term economic outcomes. OO and TB conceptualised the article; OO wrote the initial draft. All authors made substantial further contributions to the conception and revised the draft and approved the final version. OO is the guarantor.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and declare that all the authors have no relevant interest to declare.

  • Provenance and peer review: Commissioned; externally peer reviewed.

  • This article is part of a series commissioned by The BMJ. For other articles in the series see


View Abstract