In Western cultures it is well established that individuals who struggle to access or afford nutritious food (food insecurity), paradoxically tend to have higher body mass index (BMI) (Franklin et al., 2012; Nettle et al., 2017). The most common explanations are that when individuals are resource poor, they tend to purchase low costs foods that are higher in calorie content, and these foods also tend to be more readily available in deprived areas where food insecurity is highest. While these explanations are undoubtedly important, they fail to explain two phenomena. Firstly, evidence reliably shows that females are at greater risk of weight gain when they encounter food insecurity (Nettle, Andrews & Bateson, 2017; Townsend, Peerson, Love, Achterberg & Murphy, 2001). Secondly, individuals who have experienced food insecurity in childhood, tend to consume more when they encounter food insecurity in later life (Sim et al., 2018; Nettle et al., 2019). If the link between food insecurity and weight gain is solely due to availability and access to low-cost high-energy foods, then females and those with experience of childhood food poverty should not be at greater risk. In this research we seek to further test a model that could offer better understanding of the association between food insecurity, eating behaviour, BMI and why different groups of individuals appear to be affected differently
In a preliminary study run in food banks and via social media in the Northwest of England (Keenan, Christiansen & Hardman, 2020), we found evidence that the association between food insecurity and BMI might be mediated by distress and eating to cope. Specifically, the more instances of food insecurity an individual had experienced over the last year (e.g. having to skip meals, eat smaller portions than desired) the greater their levels of reported distress. Where individuals responded to this distress by consuming high energy foods as a means of coping, this was associated with a higher BMI. An additional observation was that females were significantly more likely to use food as a coping mechanism, with post-hoc analysis also revealing a trend towards males who had experienced distress more likely to turn to alcohol, although this was not statistically significant. These results raise the possibility that females and males might use different coping strategies when they encounter food insecurity related distress. Unfortunately, because our sample was 90% female, we were unable to fully test the impact of gender on the model. Using an online platform like Prolific, we hope a representative and more gender balanced sample can be recruited, which will allow us to fully test the hypothesis that gender differences in coping style might explain why females are more likely to gain weight in response to food insecurity.
The current study will also seek to measure inequalities in childhood to ascertain whether this might moderate the relationships between food insecurity and BMI in the model. It is predicted that those who have encountered food poverty in childhood will report greater distress in response to food insecurity in adulthood and be more likely to consume food or alcohol as a way of coping.
To test these predictions, we aim to run two studies. The original pilot study was run in a small region in the North West of England, so for study 1 we are seeking to run across the UK with a representative sample that is as gender balanced. In study 2, we would seek to run the model again in the USA, to establish if the pathways within the model apply to an international population.
In both studies, participants will provide informed consent and complete questionnaires relating to the experience of: food insecurity, emotional and physical distress, eating to cope, drinking to cope, perceptions of childhood inequalities, questions about eating behaviours and height and weight (to calculate BMI). They will also provide demographic information including gender, age, household income and education. However, we will not collect any personally identifiable information.
The following hypotheses will be tested:
H1: Food insecurity will be indirectly associated with BMI via distress and using food as a coping mechanism
H2: The pathway in H1 will be stronger for individuals who have experienced childhood food insecurity
H3: There will be a gender difference in the coping mechanisms used, with females more likely to eat to cope and males drinking to cope.
Analysis will use structural equation modelling to test the hypothesised pathways between food insecurity and BMI in H1, H2, and H3.
Attention measures will be included to identify distracted participants.
Impact of research
If gender differences and early childhood experiences are observed to be important, this would be the first time that distress and coping mechanisms will have been identified as possible explanatory variables. Findings will be of use for the development of interventions or safeguards for those most at risk of food insecurity. The knowledge generated can also be used to influence policy makers and governments, both within the UK and internationally on the importance of lifting individuals out of food poverty and preventing food insecurity.
Based on the estimates by Kim (2005), we calculate that a minimum of 467 will be needed to observe a close fitting root-mean-square error of approximation (RMSEA) (df = 24, α < 0.05, 90% power) of model fit for both studies. To account for potential incomplete responses and those not passing the attention checks, we aim to recruit 550 participants for each study. To ensure a representative sample in the UK for study 1 and the USA for study 2, and assuming the survey takes 25 minutes to complete, Prolific provides a total estimate of £5,719.90 (£2,89.95 per study).
Pre-registration and publication
Details of both studies will be pre-registered on AsPredicted.com under the title ‘food insecurity and BMI – May 2021’. Findings will be published in an open access journal, with all stimuli and data made available via a web appendix. This will include details and reasons for any data exclusions. The exact stimuli used on JISC online surveys, raw data and code to reproduce all analyses will be made available on an Open Science Framework repository.
Franklin, B., Jones, A., Love, D., Puckett, S., & Macklin, J. (2012). White-Means S. Exploring mediators of food insecurity and obesity: a review of recent literature. Journal of Community Health, 37(1) , 253-64.
Keenan, G. S., Christiansen, P., & Hardman, C. A. (2020). Household food insecurity, diet quality, and obesity: An explanatory model. Obesity, 29 (1 ), 143-150
Kim, K. H. (2005). The relation among fit indexes, power, and sample size in structural equation modelling. Structural Equation Modeling , 12, 368- 390.
Nettle, D., Andrews, C., Bateson, M. (2017). Food insecurity as a driver of obesity in humans: The insurance hypothesis . Behaviour and Brain Science, 40(2) , 1-53.
Nettle, D., Joly, M., Broadbent, E., Smith, C., Tittle, E., Bateson, M. (2019). Opportunistic food consumption in relation to childhood and adult food insecurity: An exploratory correlational study. Appetite, 132 (1) , 222-9.
Sim, A. Y., Lim, E. X., Forde, C. G., Cheon, B. K., (2018). Personal relative deprivation increases self-selected portion sizes and food intake. Appetite, 121(1) , 268-274
Townsend, M.S., Peerson, J., Love, B., Achterberg, C., Murphy, S.P. (2001). Food insecurity is positively related to overweight in women. Journal of Nutrition. 131(6), 1738-45.