Tackling Obesity and Reducing Health Inequalities: A Failure of Evidence-Based Policy-Making

Published On: October 2016Categories: 2016 Featured Opinions: EBDM, Editorials



University of Edinburgh

Chair, Public Health Research and Policy

Dalla Lana School of Public Health, University of Toronto

Professor Emeritus


Public health research boasts few success stories of its policy influence in recent years. Consider the examples of the obesity pandemic, and health inequalities by social class. The thousands of scientific studies about these problems provide little conclusive evidence that any specific policy is guaranteed to deal with the problem effectively. Even when we know a lot about the causation of the problem – as in the case of the global obesity pandemic — we are still not able to provide policy makers with explicit, well-evidenced guidance on how to tackle it. As I recently wrote in a commentary in Nature (Frank 2016), part of the challenge with obesity is that science is not certain which specific policy levers, aimed at changing the way people eat and drink, as well as how they engage in physical activity, are the most important ones to pull. The obesity challenge is compounded by the fact that vested interests in the food and beverage sector are fundamentally opposed to strong state action that might impact on their profits – such as subsidizing the prices of healthy foods and drinks, and taxing unhealthy ones. It is thus not surprising to hear in the news this week that the UK Government, under Prime Minister Theresa May, has decided that “the time is not right,” in the face of the UK’s current economic fragility, for such firm action on obesity, even in children. Given the many prior announcements of the opposite intention by May’s predecessor, David Cameron, this is a major climb-down – and a clear defeat for evidence-based policy-making.

In the case of health inequalities by social class, there are equally great challenges to providing policy-makers with scientifically sound evidence on what to do about them. A recent, widely-cited review of reviews, of all public health interventions which have been evaluated for their impact on health inequalities (Lorenc et al. 2013), was able to find clear evidence of inequality reduction for only three specific policies and programs:

  • the provision of free folic acid supplements to women of reproductive age (to prevent neural tube defects in their offspring 1);

  • increasing tobacco prices (largely by taxation);

  • workplace interventions to “increase employee control or participation, or change working hours.”

For fourteen other specific public health interventions that have been the subject of high-quality structured reviews in recent years, Lorenc et al. found either no data on whether those policies and programs changed health inequalities by social class (e.g. the studies did not measure inequalities), or clear evidence that such inequalities were increased. Yet many of the latter group of inequality-generating interventions are among the most widely used public health actions in many Western democracies: the mass-provision of education, communication and information (printed materials, media campaigns) about health risks and health-improving practices; school-based programmes to increase physical activity and/or improve eating behaviours; and workplace smoking bans!

In short, the science behind reducing health inequalities by social class is not yet sufficiently developed for fully informing policy, although it is able to warn us when an intervention might make inequalities worse. The most cogent studies on this topic suggest that the most promising interventions for tackling health inequalities come not from health-sector actions at all – but rather from fiscal (tax and transfer) policies, as well as social welfare programmes at the nation-state level. For example, Lynn McIntyre and colleagues in Alberta have just published (McIntyre et al. 2016) a compelling analysis of 2009-10 Canadian Community Health Survey data. These analyses show that governmental provision of universal pension payments (OAS/GAS) to those Canadians over 65 who make less than $20,000 annually, from all sources, has clearly improved their self-reported health and mental health, as well as the functional health, compared to the pension-ineligible age-group who are just slightly younger (age 55-64). The authors point out that these pension benefits operate like a Guaranteed Annual Income (GAI) program for all Canadians over 65, so that the same sort of health improvements could be expected from a comparably generous GAI program for younger Canadians at low income levels.

Whether this kind of observational evidence will convince federal and provincial policy-makers to implement GAI for other age-groups in Canada is hard to say. Budgetary pressures will certainly be cited as a reason for doing nothing of the sort. But there can be no doubt that reducing health inequalities in the long run, for a country such as Canada, will first require the reduction of socio-economic inequalities. That broader policy goal, thoroughly backed now by authoritative research (Pickett and Wilkinson 2015) as a sound strategy to improve the performance of any society, across all sectors, seems as elusive as ever.

Public health researchers committed to influencing policy therefore find themselves in a tough situation. When the science can as yet only point in a general direction of travel (as in the case of obesity control) the evidence seems powerless to outweigh vested interests poised to stop such policies. Yet when the science clearly indicates that a controversial policy (such as GAI) could make a huge impact on health inequalities – or, that many widely implemented policies and programs, such as media campaigns about staying healthy, are likely to only worsen the problem — few governments seem to have the political courage to take researchers’ advice. One can only hope that the next generation of public health investigators and policy-makers will be better able to overcome these impasses, by joint working that increases the two communities understanding of each other’s perspectives.


1  Unfortunately, this is an outdated intervention, since it is now known that only by fortifying basic foods with folic acid (as is done in Canada and the USA with flour) can one be sure to reach those women who do not yet know they are pregnant, which is when the folic acid must already be in their systems in order to prevent these serious birth anomalies (anencephaly and spina bifida).


Frank J. Origins of the obesity pandemic can be analysed. (World View commentary) Nature 2016 (April 14); 532:149.

Lorenc T, Petticrew M, Welch V, Tugwell P. What types of interventions generate inequalities? Evidence from systematic reviews. J Epidemiol Community Health 2013; 67:190-3 doi: 10.1136/jech-2012-201257.

McIntyre L, Kwok C, Emery JCH, Dutton DJ. Impact of a guaranteed annual income program on Canadian seniors’ physical, mental and functional health. Can J Public Health 2016; 107(2):e176-e182 doi: 10.17269/CJPH.107.5372.

Pickett KE, Wilkinson RG. Income inequality and health: a causal review. Social Science & Medicine 2015; 28: 316-326.