Home > Blog posts > Methods and modelling

The prevalence of painfully thin in the unemployed

Underweight may be an overlooked but crucial factor linking unemployment and poor health in the UK

Both outside and inside academia, it is often assumed that unemployed people are heavier than other people. In society at large, I expect this is tied up with media narratives about benefit claimants being lazy or lacking self-control, while academics researching the links between unemployment and health have their own reasons. Changes in dietary quality following job loss might well cause weight gain, since people often choose energy-dense but nutrient-poor foods to stretch a restricted food budget(1). At the same time, employers are known to be biased against larger candidates(2) when hiring staff; such discriminatory ‘selection’ processes should also increase the weight of the average jobseeker, as the slimmer people tend to be re-employed first, leaving the heavier ones in the pool for longer. There is also the elevated mortality rate(3) among jobseekers, which excess overweight and obesity would help explain. Nevertheless, the evidence does not support the assumption especially well. Some studies find evidence for weight gain with unemployment, but only for one gender(4, 5) or only for people whose pre-job loss BMI was already higher(5, 6); other studies suggest a fall in BMI ranking during unemployment(7, 8). Clearly, the picture is complex.

While I was doing my PhD I volunteered at a foodbank, and noticed that there were more people coming in who were painfully thin than too heavy. There were people who had not eaten that day or the day before, or who had walked for two hours to get there, because paying for a return bus journey was out of the question. Granted, by no means all people who experience unemployment turn to food banks, and not all people who turn to foodbanks are unemployed; these things are of course modified by personal circumstances and support networks. But it got me thinking: have researchers been so concerned with obesity that they have missed a crucial part of the story? If weight loss or weight gain can occur during unemployment depending on personal circumstances, might there be an overlooked ‘U-shaped’ association of unemployment and body weight, with excess obesity and excess underweight among jobseekers? Since previous studies have compared mean differences in body weight between unemployed people and controls, a U-shaped association may have been missed, since the excess obesity and excess underweight would cancel each other out in the overall average.

I work for Understanding Society, a nationally-representative survey of over 40,000 UK households, whose work and health circumstances we have followed since 2009. We used this data to look at the BMI (body mass index) of 10,737 working-age adults in 2010-2012, a period when effects of the recession were being keenly felt and substantial changes being made to the benefits system. Crucially, we used a kind of modelling that did not assume unemployment would impact BMI in the same direction for everyone. Rather, we allowed for a simultaneous elevation among jobseekers in both underweight and obesity, by comparing the probabilities having any BMI status other than the recommended one of 18.5-24.9 between people who were currently unemployed, people who had recently been unemployed, and people who had not recently been unemployed (the control group). To isolate the impact of unemployment itself, we took into account other factors such as demographics, chronic health conditions and mental health, smoking and physical activity.

For currently unemployed people, we found a hugely elevated likelihood (Odds Ratio of >4) of being medically underweight, compared to controls. This was accompanied by a reduced likelihood of being overweight (Odds Ratio: 0.71).

Certain groups were especially at risk: there were more extreme effects for longer-term unemployed people, for men, and people from lower-income households, suggesting household reserves or the support of family members may buffer against weight-loss effects. Meanwhile, there was an elevated likelihood of obesity for currently unemployed people (Odds Ratio:1.52), but only among non-smokers, possibly reflecting competing priorities between tobacco, food, and other essentials in the context of a severely restricted budget. Together, these results point to a complex picture in which jobseekers, depending on the complexities of individual lives, are at increased risk of both underweight and obesity, each with their own associated health risks.

The elevated underweight and reduced overweight among current jobseekers are quantitative evidence that many unemployed people are not eating enough in simple caloric terms. Despite the political importance of this question, evidence of this effect has so far been fairly anecdotal. Our results make an important contribution to research trying to explain the increased risk of chronic illness and mortality for unemployed people – suggesting that, at least in contemporary Britain, being underweight may contribute to that much more than previously realised.

The research, by Amanda Hughes and Meena Kumari at the Institute for Social and Economic Research at the University of Essex, published by the journal Preventive Health is available here.

Read an article about this research in The Guardian

More information on the data from Understanding Society is available here

  1. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. American Journal of Clinical Nutrition. 2004;79(1):6-16. PubMed PMID: WOS:000187569500003.

  2. Puhl RM, Heuer CA. The Stigma of Obesity: A Review and Update. Obesity. 2009;17(5):941-64. doi: 10.1038/oby.2008.636. PubMed PMID: WOS:000265709800023.

  3. Roelfs DJ, Shor E, Davidson KW, Schwartz JE. Losing Life and Livelihood: A Systematic Review and Meta-Analysis of Unemployment and All-Cause Mortality. Social science & medicine (1982). 2011;72(6):840-54. doi: 10.1016/j.socscimed.2011.01.005. PubMed PMID: PMC3070776.

  4. Monsivais P, Martin A, Suhrcke M, Forouhi NG, Wareham NJ. Job-loss and weight gain in British adults: Evidence from two longitudinal studies. Social Science & Medicine. 2015;143:223-31. doi: 10.1016/j.socscimed.2015.08.052. PubMed PMID: WOS:000364245600026.

  5. Marcus J. Does Job Loss Make You Smoke and Gain Weight? Economica. 2014;81(324):626-48. doi: 10.1111/ecca.12095. PubMed PMID: WOS:000341588800002.

  6. Deb P, Gallo WT, Ayyagari P, Fletcher JM, Sindelar JL. The effect of job loss on overweight and drinking. Journal of Health Economics. 2011;30(2):317-27. doi: 10.1016/j.jhealeco.2010.12.009. PubMed PMID: WOS:000290885400008.

  7. Montgomery S, Cook D, Bartley M, Wadsworth M. Unemployment, cigarette smoking, alcohol consumption and body weight in young British men. European Journal of Public Health. 1998;8(1):21-7. doi: 10.1093/eurpub/8.1.21. PubMed PMID: WOS:000073132000006.

  8. Jonsdottir S, Asgeirsdottir TL. The effect of job loss on body weight during an economic collapse. European Journal of Health Economics. 2014;15(6):567-76. doi: 10.1007/s10198-013-0494-z. PubMed PMID: WOS:000338212100002.

https://www.theguardian.com/society/2017/jan/29/unemployed-more-likely-to-be-underweight-than-obese