Nettle et al. note that “high levels of stored reserves ought to be found not among those whose access to food is assured, but exactly among those whose access to food is insecure” (sect. 3.3, para. 4). We agree with this and commend the thorough analysis and discussion presented in the target article that builds on similar concepts presented previously (e.g., the economic insecurity hypothesis in Smith et al. Reference Smith, Stoddard and Barnes2007), the resource scarcity hypothesis (Dhurandhar Reference Dhurandhar2016), and others (Kaiser et al. Reference Kaiser, Smith and Allison2012). Herein, we extend the discussion of the insurance hypothesis in two areas given little or no attention by Nettle et al.: (1) the relationship between food insecurity and obesity among racial and ethnic minorities in the United States and abroad and (2) the role of randomized studies to test the hypothesis in humans.
First, patterns of poverty and food insecurity, which disproportionately affect racial and ethnic minorities in the United States and abroad, yield results consistent with the insurance hypothesis. In 2015, 24.1% of Blacks and 21.4% of Hispanics were in poverty, compared to 10.1% of non-Hispanic Whites (U.S. Census Bureau 2016). Correspondingly, food insecurity is higher in racial and ethnic minority populations (Coleman-Jensen et al. Reference Coleman-Jensen, Gregory and Singh2014; Nord et al. Reference Nord, Andrews and Carlson2005). In 2013, an estimated 10.6% of non-Hispanic White households were food insecure compared with 26.1% of Black households and 23.7% of Hispanic households respectively. In line with the insurance hypothesis, rates of obesity are significantly higher in most racial and ethnic minority groups, though exceptions exist: The estimated prevalence of obesity among Asians (10.8%) is one-third that of non-Hispanic Whites (32.6%) (Ogden et al. Reference Ogden, Carroll, Kit and Flegal2014). This exception does not contradict the insurance hypothesis: 2015 median income among U.S. Asian families was 18% higher than in non-Hispanic White households (U.S. Census Bureau 2015). The authors also cite the well-documented existence of heavier body ideals in subsistence populations (Anderson et al. Reference Anderson, Crawford, Nadeau and Lindberg1992; Tovée et al. Reference Tovée, Swami, Furnham and Mangalparsad2006; Wetsman & Marlowe Reference Wetsman and Marlowe1999), suggesting cultural adaptations that complement individual behavioral responses to energetic uncertainty. Similarly, body weight ideals are higher in Blacks and Hispanics relative to the ideals of thinness reported in non-Hispanic White counterparts (Bennett & Wolin Reference Bennett and Wolin2006; Chang & Christakis Reference Chang and Christakis2003). Although a multitude of factors contribute to racial and ethnic disparities observed in the United States and abroad, the insurance hypothesis suggests an additional contributing factor – energetic uncertainty and adaptations made in response to perceived food insecurity.
Given the complex etiology of human obesity, we wish to emphasize the value of randomized controlled studies in testing the insurance hypothesis and isolating perceived energetic uncertainty as a contributing factor. Although Nettle et al. argue that designs such as the “the stronger ‘doubly longitudinal’ approach … come as close to the experimental approaches used in birds as is possible with human participants” (sect. 5.3, para. 2), we are aware of at least two published studies that have experimentally manipulated some aspect of social status, a potential contributor to energetic uncertainty, in human populations and tested subsequent obesity-related eating behavior. Both found that experimentally induced lower social status in humans caused consumption of a significantly greater percent of daily calorie needs (Cardel et al. 2016) or total calories consumed (Bratanova et al. Reference Bratanova, Loughnan, Klein, Claassen and Wood2016). The Moving to Opportunity Study was a randomized social experiment with a voucher-based intervention that allowed the intervention group to move to a nicer neighborhood, and this intervention was sufficient to reduce mean body mass index (BMI) compared to the control group that received a traditional voucher (Ludwig et al. Reference Ludwig, Sanbonmatsu, Gennetian, Adam, Duncan, Katz, Ludwig, Sanbonmatsu, Gennetian, Adam, Duncan, Katz, Kessler, Kling, Lindau, Whitaker and McDade2011). Further, the consumption of foods believed to be high in calories was higher in participants primed to the concept of environmental harshness (Laran & Salerno Reference Laran and Salerno2013). Although the aforementioned studies are clearly related to the insurance hypothesis, neither was a pure operationalization of energetic uncertainty. We believe that other cues of food insecurity, or food insecurity itself, could be manipulated and tested for their effect on food selection and eating behavior, and future research should strive to operationalize perceived energetic uncertainty such that it offers a more direct test of the insurance hypothesis.
Based on recent findings in human nutrition and obesity research, it's clear that simply providing access to more resources and/or education alone is not enough to attenuate the effect of food insecurity on the development of obesity. In some cases, the provision of financial resources, which presumably reduce food insecurity, have led to weight gain (Leroy Reference Leroy, Gadsden, González de Cosío and Gertler2013), particularly in individuals who were already overweight or obese. Additional research should identify the predictors of perceived food security, whether the experimental manipulation of these and other factors improves one's sense of control over perceived access to their food supply (e.g., gardening, financial planning, and budget-focused dietary interventions; Dhurandhar Reference Dhurandhar2016) and their potential value to a population-level intervention strategy for obesity related to food insecurity in the “obesogenic” land of plenty in high-income countries.
Nettle et al. note that “high levels of stored reserves ought to be found not among those whose access to food is assured, but exactly among those whose access to food is insecure” (sect. 3.3, para. 4). We agree with this and commend the thorough analysis and discussion presented in the target article that builds on similar concepts presented previously (e.g., the economic insecurity hypothesis in Smith et al. Reference Smith, Stoddard and Barnes2007), the resource scarcity hypothesis (Dhurandhar Reference Dhurandhar2016), and others (Kaiser et al. Reference Kaiser, Smith and Allison2012). Herein, we extend the discussion of the insurance hypothesis in two areas given little or no attention by Nettle et al.: (1) the relationship between food insecurity and obesity among racial and ethnic minorities in the United States and abroad and (2) the role of randomized studies to test the hypothesis in humans.
First, patterns of poverty and food insecurity, which disproportionately affect racial and ethnic minorities in the United States and abroad, yield results consistent with the insurance hypothesis. In 2015, 24.1% of Blacks and 21.4% of Hispanics were in poverty, compared to 10.1% of non-Hispanic Whites (U.S. Census Bureau 2016). Correspondingly, food insecurity is higher in racial and ethnic minority populations (Coleman-Jensen et al. Reference Coleman-Jensen, Gregory and Singh2014; Nord et al. Reference Nord, Andrews and Carlson2005). In 2013, an estimated 10.6% of non-Hispanic White households were food insecure compared with 26.1% of Black households and 23.7% of Hispanic households respectively. In line with the insurance hypothesis, rates of obesity are significantly higher in most racial and ethnic minority groups, though exceptions exist: The estimated prevalence of obesity among Asians (10.8%) is one-third that of non-Hispanic Whites (32.6%) (Ogden et al. Reference Ogden, Carroll, Kit and Flegal2014). This exception does not contradict the insurance hypothesis: 2015 median income among U.S. Asian families was 18% higher than in non-Hispanic White households (U.S. Census Bureau 2015). The authors also cite the well-documented existence of heavier body ideals in subsistence populations (Anderson et al. Reference Anderson, Crawford, Nadeau and Lindberg1992; Tovée et al. Reference Tovée, Swami, Furnham and Mangalparsad2006; Wetsman & Marlowe Reference Wetsman and Marlowe1999), suggesting cultural adaptations that complement individual behavioral responses to energetic uncertainty. Similarly, body weight ideals are higher in Blacks and Hispanics relative to the ideals of thinness reported in non-Hispanic White counterparts (Bennett & Wolin Reference Bennett and Wolin2006; Chang & Christakis Reference Chang and Christakis2003). Although a multitude of factors contribute to racial and ethnic disparities observed in the United States and abroad, the insurance hypothesis suggests an additional contributing factor – energetic uncertainty and adaptations made in response to perceived food insecurity.
Given the complex etiology of human obesity, we wish to emphasize the value of randomized controlled studies in testing the insurance hypothesis and isolating perceived energetic uncertainty as a contributing factor. Although Nettle et al. argue that designs such as the “the stronger ‘doubly longitudinal’ approach … come as close to the experimental approaches used in birds as is possible with human participants” (sect. 5.3, para. 2), we are aware of at least two published studies that have experimentally manipulated some aspect of social status, a potential contributor to energetic uncertainty, in human populations and tested subsequent obesity-related eating behavior. Both found that experimentally induced lower social status in humans caused consumption of a significantly greater percent of daily calorie needs (Cardel et al. 2016) or total calories consumed (Bratanova et al. Reference Bratanova, Loughnan, Klein, Claassen and Wood2016). The Moving to Opportunity Study was a randomized social experiment with a voucher-based intervention that allowed the intervention group to move to a nicer neighborhood, and this intervention was sufficient to reduce mean body mass index (BMI) compared to the control group that received a traditional voucher (Ludwig et al. Reference Ludwig, Sanbonmatsu, Gennetian, Adam, Duncan, Katz, Ludwig, Sanbonmatsu, Gennetian, Adam, Duncan, Katz, Kessler, Kling, Lindau, Whitaker and McDade2011). Further, the consumption of foods believed to be high in calories was higher in participants primed to the concept of environmental harshness (Laran & Salerno Reference Laran and Salerno2013). Although the aforementioned studies are clearly related to the insurance hypothesis, neither was a pure operationalization of energetic uncertainty. We believe that other cues of food insecurity, or food insecurity itself, could be manipulated and tested for their effect on food selection and eating behavior, and future research should strive to operationalize perceived energetic uncertainty such that it offers a more direct test of the insurance hypothesis.
Based on recent findings in human nutrition and obesity research, it's clear that simply providing access to more resources and/or education alone is not enough to attenuate the effect of food insecurity on the development of obesity. In some cases, the provision of financial resources, which presumably reduce food insecurity, have led to weight gain (Leroy Reference Leroy, Gadsden, González de Cosío and Gertler2013), particularly in individuals who were already overweight or obese. Additional research should identify the predictors of perceived food security, whether the experimental manipulation of these and other factors improves one's sense of control over perceived access to their food supply (e.g., gardening, financial planning, and budget-focused dietary interventions; Dhurandhar Reference Dhurandhar2016) and their potential value to a population-level intervention strategy for obesity related to food insecurity in the “obesogenic” land of plenty in high-income countries.