Measuring food insecurity: An introduction to tools for human biologists and ecologists.

OBJECTIVE: Food insecurity is a significant and growing concern undermining the wellbeing of 30% of the global population. Food in/security is a complex construct consisting of four dimensions: availability, access, utilization, and stability, making it challenging to measure. We provide a toolkit human biologists/ecologists can use to advance research on this topic. METHODS: We review the strengths and limitations of common tools used to measure food access and utilization, the two dimensions most proximate to people's lived experience, and emphasize tools that provide data needed to best link food security with human biological outcomes. We also discuss methods that provide contextual data human biologists/ecologists will find useful for study design, ensuring instrument validity, and improving data quality. RESULTS: Food access is principally measured using experience-based instruments that emphasize economic access. Social access, such as food sharing, is under-studied and we recommend using social network analysis to explore this dimension. In terms of utilization, emphasis has been on food choice measured as dietary diversity. Food preparation and intrahousehold distribution, also part of the utilization dimension, are less studied and standardized instruments for measuring both are lacking. The embodiment of food insecurity has focused on child growth, although a growing literature addresses adult mental wellbeing and chronic and infectious disease risk. CONCLUSIONS: We see the potential to expand outcomes to include reproductive and immune function, physical activity, and the gut microbiome. Human biologists/ecologists are well-positioned to advance understanding of the human health impacts of food insecurity and provide data to support intervention efforts.

Shift in body fat distribution from lower body to upper body among urban Colombian women, 1988-1989 to 2007-2008.

OBJECTIVE: Body fat distribution may be a stronger predictor of metabolic risk than BMI. Yet, few studies have investigated secular changes in body fat distribution in middle-income countries or how those changes vary by socioeconomic status (SES). This study evaluated changes in body fat distribution by SES in Colombia, a middle-income country where BMI is increasing rapidly. DESIGN: We applied factor analysis to previously published data to assess secular changes in adiposity and body fat distribution in cross-sectional samples of urban Colombian women. Anthropometry was used to assess weight, height and skinfolds (biceps, triceps, subscapular, suprailiac, thigh, calf). SETTING: Cali, Colombia. PARTICIPANTS: Women (18-44 years) in 1988-1989 (n 1533) and 2007-2009 (n 577) from three SES groups. RESULTS: We identified an overall adiposity factor, which increased between 1988-1989 and 2007-2008 in all SES groups, particularly in the middle SES group. We also identified arm, leg and trunk adiposity factors. In all SES groups, leg adiposity decreased, while trunk and arm adiposity increased. CONCLUSIONS: Factor analysis highlighted three trends that were not readily visible in BMI data and variable-by-variable analysis of skinfolds: (i) overall adiposity increased between time periods in all SES groups; (ii) the adiposity increase was driven by a shift from lower body to upper body; (iii) the adiposity increase was greatest in the middle SES group. Factor analysis provided novel insights into secular changes and socioeconomic variation in body fat distribution during a period of rapid economic development in a middle-income country.

Variation in dietary intake and body fatness by socioeconomic status among women in the context of Costa Rican nutrition transitions.

The Nutrition Transition model posits that vegetable oils, animal source foods (ASFs) and caloric sweeteners contribute to increases in adiposity and hence body mass index. Body mass index (BMI) is increasing more rapidly among Latin American populations of low versus high socioeconomic status (SES). The objectives of this study among Costa Rican women were to: (1) compare indicators of adiposity and dietary intake by SES and (2) evaluate the relationship between intake of foods high in vegetable oils, ASFs or caloric sweeteners and body fatness. This cross-sectional study, conducted in 2014-2015, included 128 low-, middle- and high-SES non-pregnant, non-lactating women aged between 25 and 45 years with 1-4 live births. Anthropometry was used to assess BMI, body composition and body fat distribution. Dietary recalls (n = 379) were used to assess dietary intake. Percentage body fat was greater in low- versus high-SES women (31.5 ± 3.9 vs 28.2 ± 4.7%). Skinfold measurements at four sites on the upper and lower body were greater in low- versus high-SES women. Body mass index did not vary in low- versus high-SES women. Intake frequency of foods high in vegetable oils was greater in low- and middle- (1.8 and 1.8 times/day, respectively) versus high- (1.1 times/day) SES women. For individual foods, intake frequency varied significantly by SES for high-fat condiments, fried vegetables, dairy, sweetened coffee/tea and pastries and desserts. Intake frequency of Nutrition Transition food categories was not associated with percentage body fat after adjustment for energy intake. Indicators of body composition provide additional information beyond BMI that are useful in understanding SES-adiposity associations in Latin America. Approaches to understanding diet and adiposity in Latin America that focus on vegetable oils, ASFs and caloric sweeteners should consider within-country variation in the pace of the Nutrition Transition, especially when explaining variation in adiposity by SES.

No change in energy efficiency in lactation: Insights from a longitudinal study.

OBJECTIVES: Lactation is the most energy-demanding phase of reproduction for human females, but it is still unclear how women in different environments are able to meet this additional energy demand. Previous studies have investigated whether changes in metabolism could have an energy-sparing effect in lactation, with conflicting results. Here, we asked whether increased energy efficiency in physical activity serves as an energy-sparing mechanism in lactation. METHODS: We used a longitudinal design with a control group. Participants were 33 well-nourished, exclusively breastfeeding women and 29 non-pregnant, non-lactating (NPNL) controls aged 32 ± 4 years. Lactating women were measured at peak- and post-lactation. NPNL controls completed a baseline measurement and a follow-up visit. Energy efficiency in physical activity was assessed using a graded submaximal exercise test and calculated as delta efficiency (change in work accomplished over change in energy expended) and gross efficiency (work accomplished over energy expended). RESULTS: There was no significant change in either delta efficiency or gross efficiency from peak to post lactation in lactating women, and no significant difference in delta efficiency between lactating women and NPNL controls at any time period. However, lactating women showed greater between-visit variation in delta efficiency than the NPNL controls. Additionally, 79% of lactating participants lost weight between visits (mean weight loss -3.6 ± 2.3kg), consistent with a mobilization of body tissues to support lactation. CONCLUSION: We found no support for the idea that lactating women undergo an increase in energy efficiency to support the energy costs of lactation.

Using the protein leverage hypothesis to understand socioeconomic variation in obesity.

OBJECTIVES: The protein leverage hypothesis (PLH) predicts that protein appetite will stimulate excess energy intake, and consequently obesity, when the proportion of protein in the diet is low. Experimental studies support the PLH, but whether protein leverage can be used to understand socioeconomic (SES) variation in obesity is unknown. The objective of this study was to test two hypotheses from the PLH under non-experimental conditions. Consistent with the PLH, we expect that (1) absolute protein intake will be similar across populations, here defined as SES groups and, (2) the proportion of protein in the diet will be inversely associated with energy intake. METHODS: This was a cross-sectional study conducted in a random sample of 135 low-, middle-, and high-SES women in Costa Rica. Anthropometry was used to calculate body mass index (BMI). Twenty-four-hour dietary recalls were used to measure dietary intake. RESULTS: The prevalence of obesity varied between low- (38.8%), middle- (43.9%), and high- (17.8%) SES women. Absolute protein intake was similar across low- (58.5 g), middle- (59.4 g), and high- (65.6 g) SES women (p = 0.12). Protein intake as a proportion of total energy intake was inversely associated with total energy intake only among middle- (r = -0.37, p = 0.02) and high- (r = -0.36, p = 0.01) SES women. CONCLUSIONS: Consistent with the PLH, absolute protein intake was similar across SES groups. The relationship between the proportion of protein in the diet and total energy intake should be studied further in the context of real world conditions that may influence protein leverage.

Amazonian foods and implications for human biology.

CONTEXT: Diets of subsistence-based Amazonian populations have been linked to local resources, but are changing with market penetration. OBJECTIVE: To review the available data on traditional Amazonian foods and diets and evaluate their implications for human biology as a step toward understanding nutrition transitions in the region. METHODS: This study used the Human Relations Area Files for information on the diets of Amerindian groups in the Amazon Basin from 1950 to the present, and used other published sources and the authors' own data. RESULTS: Data on food use was identified for only nine groups and dietary intake data for individuals in only three of the groups. A diet based on starchy staples (manioc and plantains) and fish, supplemented with a limited variety of other plant and animal foods, was found. Bitter manioc-based foods were associated with the consumption of cyanogens and fish with the consumption of mercury. Diets of adults appear to be adequate in energy and protein and low in fats. Children's diets were not well documented. CONCLUSION: Based on the limited available data, Amazonian diets are restricted in variety, but appear to be adequate in energy and protein for adults, but likely insufficiently nutrient-dense for children.

Local trends in diet in urban Colombia, 1990-1995 to 2008: little evidence of a nutrition transition among low-income women.

OBJECTIVES: To quantify changes in the diets of low-income women in Cali, Colombia between 1990-1995 and 2008, a period of increases in body size, and to situate these changes within national-level trends in food availability, as well as to compare these changes with those expected in countries undergoing a nutrition transition. METHODS: Individual dietary intake was assessed via 24-hour recalls in both 1990-1995 (n = 85) and 2008 (n = 88). Dietary data were analyzed for intake of energy, macronutrients, and specific food items. National-level trends in food availability were evaluated using data from the Food and Agriculture Organization. RESULTS: Total energy and protein intake did not change over time, but in 2008 women consumed proportionally more fat (23.0 vs. 19.1% of calories; P = 0.002) and less carbohydrate (66.5 vs. 71.0% of calories; P < 0.001) than in 1990-1995. The increased fat consumption is attributable to vegetable oils. This increase in vegetable oil consumption, and a decrease in starchy vegetable consumption, fit with both national-level trends in food supply, and the expectations of a nutrition transition. On the other hand, the increased consumption of non-starchy vegetables, and the stability in consumption of added sugars and animal-source proteins was contrary to the expectations of a nutrition transition. CONCLUSIONS: The changes in diet among low-income women in Cali, Colombia between 1990-1995 and 2008 partially match national-level trends in food supply and the theoretical expectations of a nutrition transition, but are nonetheless a localized phenomenon. They do not help explain concurrent changes in body size.

Stable isotopes (carbon, nitrogen, sulfur), diet, and anthropometry in urban Colombian women: investigating socioeconomic differences.

OBJECTIVES: We conducted stable isotope and dietary analyses of women from higher and lower socioeconomic status (SES) groups in Cali, Colombia. The objectives were to test between-group differences in stable isotope, dietary, and anthropometric characteristics, and to evaluate relationships between diet and stable isotope values. METHODS: Hair samples from 38 women (mean age 33.4) from higher and lower SES groups were analyzed for δ(13) C, δ(15) N, and δ(34) S values. Dietary intake was assessed via 24-h recalls. Anthropometric variables measured were body mass index, five body circumferences, and six skinfold thicknesses. RESULTS: Mean δ(13) C and δ(15) N values of the higher SES group (-16.4 and 10.3‰) were significantly greater than those of the lower SES group (-17.2 and 9.6‰; P < 0.01), but mean δ(34) S values did not differ significantly between groups (higher SES: 4.6‰; lower SES: 5.1‰). The higher SES group consumed a greater percentage of protein than the lower SES group (14% vs. 12% of energy; P = 0.03), but the groups did not differ in other dietary characteristics or in anthropometric characteristics. δ(13) C, δ(15) N, and δ(34) S values were not correlated with intake of the dietary items predicted (sugars, animal-source protein, and marine foods, respectively). The lower SES group was more variable in all three stable isotope values (P < 0.05), mirroring a trend toward greater dietary variability in this group. CONCLUSIONS: Stable isotope values revealed a difference between SES groups that was not explained by the dietary data. The relationship between diet and stable isotope composition is complex.

Exploring the multidimensionality of stature variation in the past through comparisons of archaeological and living populations.

Adult stature variation is commonly attributed to differential stress-levels during development. However, due to selective mortality and heterogeneous frailty, a population's tall stature may be more indicative of high selective pressures than of positive life conditions. This article examines stature in a biocultural context and draws parallels between bioarchaeological and living populations to explore the multidimensionality of stature variation in the past. This study investigates: 1) stature differences between archaeological populations exposed to low or high stress (inferred from skeletal indicators); 2) similarities in growth retardation patterns between archaeological and living groups; and 3) the apportionment of variance in growth outcomes at the regional level in archaeological and living populations. Anatomical stature estimates were examined in relation to skeletal stress indicators (cribra orbitalia, porotic hyperostosis, linear enamel hypoplasia) in two medieval bioarchaeological populations. Stature and biocultural information were gathered for comparative living samples from South America. Results indicate 1) significant (P < 0.01) differences in stature between groups exposed to different levels of skeletal stress; 2) greater prevalence of stunting among living groups, with similar patterns in socially stratified archaeological and modern groups; and 3) a degree of regional variance in growth outcomes consistent with that observed for highly selected traits. The relationship between early stress and growth is confounded by several factors-including catch-up growth, cultural buffering, and social inequality. The interpretations of early life conditions based on the relationship between stress and stature should be advanced with caution.

Socioeconomic status, stature, and obesity in women: 20-year trends in urban Colombia.

OBJECTIVES: Ongoing social and economic changes in developing countries are associated with increases in body size, and most notably increases in the prevalence of obesity. The social patterning of these changes in terms of socioeconomic status (SES) is not well documented. The aim of this study was to assess the changes in stature, body mass index (BMI) and fatness in adult women in urban Cali, Colombia between 1988-1989 and 2007-2008. METHODS: We compared the results of anthropometric surveys completed in 1988-1989 and 2007-2008 of nonpregnant, nonlactating women, 18-44 years of age. Samples in both studies were stratified by SES. We calculated age-standardized prevalence rates to assess time trends in obesity. Body fatness was assessed by skinfold thicknesses. RESULTS: Stature increased in all SES groups and remained positively associated with SES. BMI increased only in the lower SES group, from 24.4 to 25.9 kg/m(2) and remained negatively associated with SES. The age-standardized prevalence of obesity increased from 7.9 to 17.0% in the lower SES group, but only from 4.5 to 8.2% in the middle SES group, and was unchanged in the upper SES group. Body fatness increased in all SES groups, but only in the upper body. CONCLUSION: The increased stature in all SES groups is indicative of general improvements in socioeconomic conditions. The increased prevalence of obesity in the lower SES groups is in keeping with the findings in other middle-income developing countries.

Assessing diet in populations at risk for konzo and neurolathyrism.

Although both konzo and neurolathyrism are diseases associated with diet, we know surprising little about the diets of the groups at risk. The objective of this paper is to discuss methods for assessing dietary intake in populations at risk for konzo and lathyrism. These methods include weighed food records and interview based techniques like 24-h recalls and food frequency questionnaires (FFQs). Food records have the potential to provide accurate information on food quantities, and are generally the method of choice. Interview based methods provide less precise information on the quantities of foods ingested, and are subject to recall bias, but may be useful in some studies or for surveillance. Sample size needs to be adequate to account for day-to-day and seasonal variability in food intake, and differences between age and sex groups. Adequate data on the composition of foods, as actually consumed, are needed to evaluate the food intake information. This is especially important in the case of cassava and grass pea where the toxins in the diet is a function of processing. Biomarkers for assessing the cyanogen exposure from cassava-based diets are available; biomarkers for the ß-ODAP exposure from grass pea diets need development.

Energy expenditure among farmers in developing countries: what do we know?

The trend toward overweight and obesity in developing countries is often assumed to be due, in part, to reductions in energy expenditure associated with the transition from agrarian to urban lifestyles. In this article we first review the published studies on energy expenditure in farming populations living in developing countries, populations generally assumed to have high levels of energy expenditure. To facilitate comparison we express energy expenditure as physical activity level (PAL), i.e. the ratio of total daily energy expenditure to basal metabolic rate. Then, with the goal of better understanding variability in energy expenditure between different human groups, we focus on case studies of women farmers in Colombia and Brazil for whom we have good ethnographic data. The published studies reviewed indicate that most farmers have PAL values in the moderate physical activity range, but toward the high end of that range. PAL values of male farmers tend to be higher than female, and show greater seasonal variation. The case studies illustrate that women farmers, living in broadly similar environments, and dependent on the cultivation of the same crop can have quite different patterns of physical activity and hence PAL values. These differences are a function of differences in behavior related to social and cultural variables like the organization of work at the household level and perceptions of how food crops should be processed, as well as micro-level ecological factors.

Diet, energy expenditure, and body composition of lactating Ribeirinha women in the Brazilian Amazon.

Lactation is the most energetically demanding part of human reproduction; yet, compared with pregnancy, we know little about the strategies women in different settings employ to cope with these increased energy demands. This paper takes a biocultural approach and reports longitudinal data on the anthropometry, dietary intakes and energy expenditure of a sample of 23 rural, lactating Ribeirinha women living in subsistence-based communities in the eastern Amazon. The dietary intakes of these lactating women were insufficient to meet their lactating energy needs and were least sufficient during resguardo, a 40-day period in the immediate postpartum when the women observed a series of food taboos and work restrictions. Instead, the women in this study met the increased energy demands of lactation by drawing on their energy reserves and reducing their energy expenditure in physical activity. The women showed a significant reduction in weight (P<0.001), BMI (P<0.001) and in circumferences (hip, P=0.01; waist, P=0.03) and skinfolds (thigh, P=0.03) in the gluteal femoral region. Total daily energy expenditure (TDEE) was lowest during resguardo and increased as lactation progressed (P=0.01). While the practice of resguardo reduced maternal energy expenditure and allowed women more time to spend with their newborn infants, it came at a cost (low dietary intake), which appears to be related to the loss of the adult woman from subsistence activities. By taking a biocultural approach this study illustrates the role the social environment plays in shaping the experience of lactating women.

Biocultural approaches in human biology.

Biocultural approaches recognize the pervasiveness and dynamism of interactions between biological and cultural phenomena, and they explicitly strive to integrate biological, sociocultural, environmental, and other kinds of data. They have been part of human biology at least since 1958, when Frank Livingstone so elegantly explained the linkages among population growth, subsistence strategy, and the distribution of the sickle cell gene in West Africa. These approaches developed further with the advent of human adaptability studies in the 1960s as part of the Human Biological Program and have become increasingly focused on understanding the impacts of everyday life on human biological variation. Biocultural approaches generate explanations that are intuitively appealing to many because they offer a kind of holistic view. They can, however, be very challenging approaches to implement, perhaps in part because we are more experienced in measuring the biological than the cultural. Some of the challenges include (1) defining precisely what we mean by constructs like socioeconomic status, poverty, rural, and urban; (2) operationalizing key variables so that they can be measured in ways that are ethnographically valid as well as replicable; (3) defining and measuring multiple causal pathways. In this paper, I briefly review the history of biocultural approaches and then illustrate some of the challenges that these approaches present with examples from my own research on nutrition and energetics as well as that of other practitioners.

Anthropometric characteristics of pregnant women in Cali, Colombia and relationship to birth weight.

Anthropometric dimensions were collected from 46 pregnant women living in Cali, Colombia to gain a better understanding of how poor, urban women deal with the demands of pregnancy and to identify relationships between maternal characteristics and infant birth weight. Height, weight, skinfold thicknesses (subscapular, suprailiac, thigh, calf, and triceps), and circumferences (hip, thigh, calf, and mid-upper arm) were measured on all women. Infant measurements were weight and length. The women were measured in the second and third trimesters, and a subsample (n = 16) was measured twice in the third trimester. Mean birth weight was 3,137.6 +/- 488.5 g (n = 44), and mean length was 49.8 +/- 3.0 cm. All but three of the infants were full-term, and the incidence of low birth weight (LBW) was 9%. The 46 women showed a significant increase in weight (P < 0.001); subscapular, suprailiac, and mid-thigh skinfold thicknesses (P < or = 0.01) and in hip, thigh, and calf circumferences (P < or = 0.01) between trimesters 2 and 3. Women who gave birth to both normal birth weight (NBW) and LBW infants showed significant increases in weight (P < 0.001 and P = 0.02, respectively), but only women who had NBW infants showed significant increases in the suprailiac skinfold and hip circumference (P < 0.001). In the third trimester, attained weight, skinfold thicknesses, and circumferences tended to be greater in women who had NBW infants. In general, this group of women gained less weight and had a greater incidence of LBW infants compared with women in developed countries, but changes in skinfold thicknesses over the course of pregnancy were similar compared with other studies.

Food and macronutrient intake of economically disadvantaged pregnant women in Colombia.

The objective of this article is to assess changes in diet composition, defined in terms of macronutrient intake and types of foods consumed, in pregnancy in poor urban women in Colombia. The subjects were 20 pregnant and 20 matched nonpregnant, nonlactating (NPNL) women 19 to 35 years of age. The pregnant women were studied in three measurement rounds at 14.0 +/- 3.6, 27 +/- 2.2, and 35 +/- 1.7 weeks gestation, and the NPNL women in three measurement rounds approximately 3 months apart. Dietary intake was obtained from estimated food records and macronutrient composition from published sources. Types of foods consumed were aggregated into 16 groups: alcohol; breads; candy; coffee, chocolate; juices; fruit; legumes; meat, fish, offal; dairy; vegetable dishes; other; rice, pasta; tubers, plantains; salads; soft drinks; and soups. Macronutrient intakes showed nonsignificant increases in pregnancy. There were no significant differences between pregnant and NPNL women, except for carbohydrate intake in late pregnancy (P = 0.03). Carbohydrate, fat, and protein provided 74%, 17%, and 12% of dietary energy, respectively, in pregnant women at baseline, and did not change significantly. Except for a decrease in fruits, the types of foods consumed did not change significantly in pregnancy. There were no between-group differences in types of foods consumed except for the greater number of fruits consumed by pregnant women at baseline (P = 0.004). We conclude that in this population there were no changes in diet composition in pregnancy, except for an increase in fruit consumption in Round 1. Am. J. Hum. Biol. 11:753-762, 1999. Copyright 1999 Wiley-Liss, Inc.

Household headship and nutritional status: Female-headed versus male/dual-headed households.

Approximately one-third of all poor urban households in Latin America are headed by females. This study compares the nutritional status of women and children from female-headed (FHH) and male/dual headed (MHH/DHH) households in economically disadvantaged neighborhoods in Cali, Colombia. Data collected included socioeconomic indicators, anthropometric dimensions, dietary intake, and interviews documenting household composition. Classification of household headship was based on the acknowledged head and the primary economic provider of the household. Of the households in this study, 81 (80%) were from MHH/DHHs and 20 (20%) from FHHs. FHHs had fewer material possessions and lived in houses made of less expensive materials compared to MHH/DHHs. However, there were no significant differences in the nutritional status of women or children in FHHs vs MHH/DHHs. Most women had BMIs within the normal range (19-29 kg/m2 ). The majority of children (95%) had weight-for-height z-scores within ±2 of the NCHS references. These results suggest that while FHHs in Cali were economically poorer, the women studied appeared to have had access to sufficient food sources for themselves and their children. Am. J. Hum. Biol. 10:699-709, 1998. © 1998 Wiley-Liss, Inc.