The insurance hypothesis offers an intriguing environment-based account of the global obesity crisis. Considering the mismatch between ancestral food scarcity and the unprecedented energy-density of contemporary food environments, it attributes socioeconomic differences in obesity to an evolved adaptive mechanism. Specifically, individuals are hypothesized to store more fat when cues indicate that access to food is uncertain, thus buffering against future shortages. The authors acknowledge the multicausality of obesity, with no single factor explaining all or most of the variance. Against this background, they also consider the role of genes and psychological factors such as impulsivity and inhibition in explaining the complex phenomenon of obesity. However, one important factor is missing from their account: the role of others, namely, nutritional gatekeepers.

Although Nettle et al. emphasize the role of society in explaining obesity, they depict food choice – perhaps the single most important behavioral act associated with obesity or lack thereof – as an individualistic decision. They thus overlook social dimensions that are crucial in explaining some of the perplexing patterns surrounding obesity. Eating is not a solitary intake of energy but often a social activity shaped by others’ dietary behavior and choices (Herman et al. Reference Herman, Polivy and Roth2003). “Company” literally means “with bread” – company is those with whom we break bread. Few, if any, health-related behaviors are as closely embedded in the social context as eating – especially where children are concerned. By the age of 10, a child has eaten about 10,000 meals, most of them in the company of their family. Yet, children rarely enjoy autonomy in their food choices. Nutritional gatekeepers – not only parents and grandparents, but also institutions (e.g., kindergartens, schools, policymakers) determine the food choice architecture. One of the most important food contexts for children is the family, with nutritional gatekeepers determining more than 70% of what the family eats (Wansink Reference Wansink2006) through, for example, their economic resources (Keane et al. Reference Keane, Layte, Harrington, Kearney and Parry2012), family mealtime practices (Dallacker et al., Reference Dallacker, Hertwig and Mataunder review), nutritional knowledge, and numerical abilities, which are associated with comprehension of nutrition labels or portion size estimation skills (Dallacker et al. Reference Dallacker, Hertwig, Peters and Mata2016; Mata et al. Reference Mata, Scheibehenne and Todd2008).

Why do the authors not find a relationship between food insurance and obesity in children? We suggest that recognizing the social nature of eating – and, in particular, of children's food choices – can offer answers for this missing link. Not all participants in the shared activity of eating (e.g., Sobal & Nelson Reference Sobal and Nelson2003) will be equally affected by (perceived) food insecurity. Despite eating at the same table, the last-born child is often less well-nourished than the first-born. For example, in a family of seven, the height for age of the last-born child is up to 2.5 standard deviations less than that of the first-born (Hertwig et al. Reference Hertwig, Davis and Sulloway2002). Yet ethical norms and legal policies aim to protect children from malnutrition or starvation. For example, mothers report abstaining from food to ensure that their children are adequately nourished (McIntyre et al. Reference McIntyre, Glanville, Raine, Dayle, Anderson and Battaglia2003; Piperata et al. Reference Piperata, Schmeer, Hadley and Ritchie-Ewing2013). In wealthy societies, institutional settings such as daycare centers and schools often provide free lunches or free milk to children from impoverished families. In the United States, for example, 16 different food assistance programs were funded in 2002, and one in five Americans participated in them at one point during that year (Fox et al. Reference Fox, Hamilton and Lin2004). Thus, relatively rich societies aim to protect children from the detrimental effects of food insecurity through social norms, welfare assistance, and institutionalized arrangements. Admittedly, despite these efforts, even high-income countries appear to suffer from “hidden hunger” and malnutrition caused by vitamin and mineral deficiencies that threaten to impair children's intellectual and physical development (Biesalski & Black Reference Biesalski and Black2016).

The authors suggest a methodological explanation for the missing link between food insecurity and body mass index: The studies included in the meta-analysis measured a child's food insecurity through parental reports, which are likely to differ from the child's perception (Connell et al. Reference Connell, Lofton, Yadrick and Rehner2005; Fram et al. Reference Fram, Frongillo, Jones, Williams, Burke, DeLoach and Blake2011). Importantly, this is not only a methodological, but also, again, a social explanation. As described above, wealthy societies aim to protect children from hunger and food poverty both within the family and beyond (Fox et al. Reference Fox, Hamilton and Lin2004). As a consequence of the multiple individual and institutional nutritional gatekeepers involved in children's nutrition, parental perceptions – being just one piece of the social puzzle – may not be a veridical and integrative proxy of a child's food (in-)security – even more so when parents equate food security with lack of hunger and thus neglect the risk of malnutrition.

To conclude, eating was, has been, and will likely continue to be a shared activity – not always, but often. Any comprehensive model of obesity therefore needs to account for the social nature of food choice and consumption. This is particularly the case for children, whose food choice autonomy is restricted. The authors did not find a link between food insecurity and children's obesity. The reason could be that this link simply does not exist or is relatively weak because parents, institutions, and policymakers buffer children from food insecurity. Alternatively, a link may exist, but it may be moderated by who is competent to gauge children's experience of food security or lack thereof: the children, their parents, institutional settings, policymakers? A stringent test of the food insecurity hypothesis in children demands that proper attention be paid to the social dynamics of food choice and eating.