Economies of scale
Too Much of a Good Thing? Exploring the Impact of Wealth on Weight
Nicole Au & David Johnston
Health Economics, forthcoming
Abstract:
Obesity, like many health conditions, is more prevalent among the socioeconomically disadvantaged. In our data, very poor women are three times more likely to be obese and five times more likely to be severely obese than rich women. Despite this strong correlation, it remains unclear whether higher wealth causes lower obesity. In this paper, we use nationally representative panel data and exogenous wealth shocks (primarily inheritances and lottery wins) to shed light on this issue. Our estimates show that wealth improvements increase weight for women, but not men. This effect differs by initial wealth and weight — an average-sized wealth shock received by initially poor and obese women is estimated to increase weight by almost 10 lb. Importantly, for some females, the effects appear permanent. We also find that a change in diet is the most likely explanation for the weight gain. Overall, the results suggest that additional wealth may exacerbate rather than alleviate weight problems.
----------------------
Caloric Beverage Intake among Adult Supplemental Nutrition Assistance Program Participants
Jessica Todd & Michele Ver Ploeg
American Journal of Public Health, September 2014, Pages e80-e85
Objectives: We compared sugar-sweetened beverage (SSB), alcohol, and other caloric beverage (juice and milk) consumption of Supplemental Nutrition Assistance Program (SNAP) participants with that of low-income nonparticipants.
Methods: We used 1 day of dietary intake data from the 2005–2008 National Health and Nutrition Examination Survey for 4594 adults aged 20 years and older with household income at or below 250% of the federal poverty line. We used bivariate and multivariate methods to compare the probability of consuming and the amount of calories consumed for each beverage type across 3 groups: current SNAP participants, former participants, and nonparticipants. We used instrumental variable methods to control for unobservable differences in participant groups.
Results: After controlling for observable characteristics, SNAP participants were no more likely to consume SSBs than were nonparticipants. Instrumental variable estimates showed that current participants consumed fewer calories from SSBs than did similar nonparticipants. We found no differences in alcoholic beverage consumption, which cannot be purchased with SNAP benefits.
Conclusions: SNAP participants are not unique in their consumption of SSBs or alcoholic beverages. Purchase restrictions may have little effect on SSB consumption.
----------------------
Heather Scherschel Wagner et al.
Health Psychology, forthcoming
Objective: People seek out their own idiosyncratic comfort foods when in negative moods, and they believe that these foods rapidly improve their mood. The purpose of these studies is to investigate whether comfort foods actually provide psychological benefits, and if so, whether they improve mood better than comparison foods or no food.
Methods: Participants first completed an online questionnaire to indicate their comfort foods and a variety of comparison foods. During two lab sessions a week apart from each other (and at least a week after the online questionnaire, counterbalanced in order), participants watched films that induced negative affect. In one session, participants were then served their comfort food. In the other, participants were served an equally liked noncomfort food (Study 1), a neutral food (Study 2), or no food (Studies 3 and 4). Short-term mood changes were measured so that we could seek out psychological effects of these foods, rather than biochemical effects on mood from particular food components (e.g., sugars or vitamins).
Results: Comfort foods led to significant improvements in mood, but no more than other foods or no food.
Conclusions: Although people believe that comfort foods provide them with mood benefits, comfort foods do not provide comfort beyond that of other foods (or no food). These results are likely not due to a floor effect because participants’ moods did not return to baseline levels. Individuals may be giving comfort food “credit” for mood effects that would have occurred even in the absence of the comfort food.
----------------------
Reducing Childhood Obesity through U.S. Federal Policy: A Microsimulation Analysis
Alyson Kristensen et al.
American Journal of Preventive Medicine, forthcoming
Background: Childhood obesity prevalence remains high in the U.S., especially among racial/ethnic minorities and low-income populations. Federal policy is important in improving public health given its broad reach. Information is needed about federal policies that could reduce childhood obesity rates and by how much.
Purpose: To estimate the impact of three federal policies on childhood obesity prevalence in 2032, after 20 years of implementation.
Methods: Criteria were used to select the three following policies to reduce childhood obesity from 26 recommended policies: afterschool physical activity programs, a $0.01/ounce sugar-sweetened beverage (SSB) excise tax, and a ban on child-directed fast food TV advertising. For each policy, the literature was reviewed from January 2000 through July 2012 to find evidence of effectiveness and create average effect sizes. In 2012, a Markov microsimulation model estimated each policy’s impact on diet or physical activity, and then BMI, in a simulated school-aged population in 2032.
Results: The microsimulation predicted that afterschool physical activity programs would reduce obesity the most among children aged 6–12 years (1.8 percentage points) and the advertising ban would reduce obesity the least (0.9 percentage points). The SSB excise tax would reduce obesity the most among adolescents aged 13–18 years (2.4 percentage points). All three policies would reduce obesity more among blacks and Hispanics than whites, with the SSB excise tax reducing obesity disparities the most.
Conclusions: All three policies would reduce childhood obesity prevalence by 2032. However, a national $0.01/ounce SSB excise tax is the best option.
----------------------
From Coke to Coors: A Field Study of a Fat Tax and its Unintended Consequences
Brian Wansink et al.
Cornell University Working Paper, July 2014
Abstract:
Could taxation of calorie-dense foods such as soft drinks be used to reduce obesity? To address this question, a six-month field experiment was conducted in an American city of 62,000 where half of the 113 households recruited into the study faced a 10% tax on calorie-dense foods and beverages and half did not. The tax resulted in a short-term (1-month) decrease in soft drink purchases, but no decrease over a 3-month or 6-month period. Moreover, in beer-purchasing households, this tax led to increased purchases of beer. To behavior scholars, this underscores the importance of investigating unexpected substitutions. To public health officials and policy makers, this presents an important empirical result and more generally points toward wide ranging contributions that marketing scholarship can make in their decisions.
----------------------
Daniel Taber et al.
PLoS ONE, August 2014
Background: Sodas are widely sold in vending machines and other school venues in the United States, particularly in high school. Research suggests that policy changes have reduced soda access, but the impact of reduced access on consumption is unclear. This study was designed to identify student, environmental, or policy characteristics that modify the associations between school vending machines and student dietary behaviors.
Methods: Data on school vending machine access and student diet were obtained as part of the National Youth Physical Activity and Nutrition Study (NYPANS) and linked to state-level data on soda taxes, restaurant taxes, and state laws governing the sale of soda in schools. Regression models were used to: 1) estimate associations between vending machine access and soda consumption, fast food consumption, and lunch source, and 2) determine if associations were modified by state soda taxes, restaurant taxes, laws banning in-school soda sales, or student characteristics (race/ethnicity, sex, home food access, weight loss behaviors.)
Results: Contrary to the hypothesis, students tended to consume 0.53 fewer servings of soda/week (95% CI: -1.17, 0.11) and consume fast food on 0.24 fewer days/week (95% CI: -0.44, -0.05) if they had in-school access to vending machines. They were also less likely to consume soda daily (23.9% vs. 27.9%, average difference = -4.02, 95% CI: -7.28, -0.76). However, these inverse associations were observed primarily among states with lower soda and restaurant tax rates (relative to general food tax rates) and states that did not ban in-school soda sales. Associations did not vary by any student characteristics except for weight loss behaviors.
Conclusion: Isolated changes to the school food environment may have unintended consequences unless policymakers incorporate other initiatives designed to discourage overall soda consumption.
----------------------
Peer-Effects in Obesity among Public Elementary School Children: A Grade-Level Analysis
Jebaraj Asirvatham, Rodolfo Nayga & Michael Thomsen
Applied Economic Perspectives and Policy, September 2014, Pages 438-459
Abstract:
Using a panel dataset at the grade level from Arkansas public schools, this study finds that changes in the obesity prevalence at the oldest grade are associated with changes in obesity prevalence at younger grades. Furthermore, analysis across different school types shows that the peer effects are statistically significant but the magnitude of the effect is greater in kindergarten to fourth-grade schools than in kindergarten to sixth-grade schools. We also use tests on spatial and temporal dimensions, as well as by weight status grouping, to provide evidence that these peer effects are more than just a statistical correlation.
----------------------
Generational Shift in Parental Perceptions of Overweight Among School-Aged Children
Andrew Hansen et al.
Pediatrics, September 2014, Pages 481 -488
Background: Parental perceptions of child’s weight status may influence family readiness to foster healthy behaviors. This study investigated the generational shifting of parental perceptions about children’s weight.
Methods: Data were collected in the NHANES 1988–1994 (n = 2871) and 2005–2010 (n = 3202). Parents, mainly mothers, were asked whether they considered their child, ages 6 to 11 years, to be overweight, underweight, or just about the right weight. The Centers for Disease Control and Prevention 2000 growth chart was used for reference. We ran Poisson regression to estimate the probability ratio between the 2 surveys for parents perceiving their child as overweight after controlling for actual weight.
Results: The 10th percentile of BMI z scores for children who were parentally perceived as overweight shifted with statistical significance from 84th percentile of reference population in the early survey to 91st percentile of reference population in the recent survey (P < .05). The mean z score of children parentally perceived as overweight also increased between surveys with the largest increase among children from poor families (from 1.60 [SE: 0.20] to 1.98 [0.08], P < .05), followed by African Americans (from 1.65 [0.09] to 2.02 [0.05], P < .05). The probability of overweight/obese children being correctly perceived as overweight by the parents declined by 24% between surveys (probability ratio = 0.76 [95% confidence interval: 0.67–0.87]).
Conclusions: Overweight/obese children were less likely to be perceived as overweight in the recent survey compared with peers of similar weight but surveyed 10+ years earlier. The declining tendency among parents to perceive overweight children appropriately may indicate a generational shift in social norms related to body weight.
----------------------
Comparison of Weight Loss Among Named Diet Programs in Overweight and Obese Adults: A Meta-analysis
Bradley Johnston et al.
Journal of the American Medical Association, September 2014, Pages 923-933
Objective: To determine weight loss outcomes for popular diets based on diet class (macronutrient composition) and named diet.
Data Extraction and Synthesis: Two reviewers independently extracted data on populations, interventions, outcomes, risk of bias, and quality of evidence. A Bayesian framework was used to perform a series of random-effects network meta-analyses with meta-regression to estimate the relative effectiveness of diet classes and programs for change in weight and body mass index from baseline. Our analyses adjusted for behavioral support and exercise.
Results: Among 59 eligible articles reporting 48 unique randomized trials (including 7286 individuals) and compared with no diet, the largest weight loss was associated with low-carbohydrate diets (8.73 kg [95% credible interval {CI}, 7.27 to 10.20 kg] at 6-month follow-up and 7.25 kg [95% CI, 5.33 to 9.25 kg] at 12-month follow-up) and low-fat diets (7.99 kg [95% CI, 6.01 to 9.92 kg] at 6-month follow-up and 7.27 kg [95% CI, 5.26 to 9.34 kg] at 12-month follow-up). Weight loss differences between individual diets were minimal. For example, the Atkins diet resulted in a 1.71 kg greater weight loss than the Zone diet at 6-month follow-up. Between 6- and 12-month follow-up, the influence of behavioral support (3.23 kg [95% CI, 2.23 to 4.23 kg] at 6-month follow-up vs 1.08 kg [95% CI, −1.82 to 3.96 kg] at 12-month follow-up) and exercise (0.64 kg [95% CI, −0.35 to 1.66 kg] vs 2.13 kg [95% CI, 0.43 to 3.85 kg], respectively) on weight loss differed.
Conclusions and Relevance: Significant weight loss was observed with any low-carbohydrate or low-fat diet. Weight loss differences between individual named diets were small. This supports the practice of recommending any diet that a patient will adhere to in order to lose weight.
----------------------
Gaurav Suri et al.
Journal of Experimental Psychology: Applied, forthcoming
Abstract:
To encourage an increase in daily activity, researchers have tried a variety of health-related communications, but with mixed results. In the present research — using the stair escalator choice context — we examined predictions derived from the Heuristic Systematic Model (HSM), Self Determination Theory (SDT), and related theories. Specifically, we tested whether (as predicted by HSM) signs that encourage heuristic processing (“Take the Stairs”) would have greatest impact when placed at the stair/escalator point of choice (when processing time is limited), whereas signs that encourage systematic processing (“Will You Take the Stairs?”) would have greatest impact when placed at some distance from the point of choice (when processing time is less limited). We also tested whether (as predicted by SDT) messages promoting autonomy would be more likely to result in sustained motivated behavior (i.e., stair taking at subsequent uncued choice points) than messages that use commands. A series of studies involving more than 9,000 pedestrians provided support for these predictions.
----------------------
Prompting healthier eating: Testing the use of health and social norm based messages
Eric Robinson, Alexander Fleming & Suzanne Higgs
Health Psychology, September 2014, Pages 1057-1064
Objective: Health based messages are commonly used to promote fruit and vegetable intake, but are limited in their effectiveness. Social norm messages, which suggest other people are eating healthily, may be more effective. Our aim was to compare the effect on food selection of a message containing health related information about fruit and vegetable consumption with a message containing social normative information about consumption of fruit and vegetables.
Method: In two laboratory studies, predominantly young female adult students were exposed to a health or social norm message about fruit and vegetables. In Study 1, lunch meal food selections and intake were assessed and in Study 2, snack food selections and intake were assessed. Study 1 examined the effect of a descriptive social norm (information about what others are eating) versus a health message and Study 2 examined the effect of both a descriptive norm and an injunctive norm message (information about what others approve of) versus a health message.
Results: In Study 1, exposure to a descriptive social norm message resulted in significantly more vegetables being selected and eaten than exposure to a health message. In Study 2, exposure to a descriptive social norm message resulted in significantly more fruit and vegetables and less high energy dense snack food being selected and eaten than exposure to a health message. There was no effect of exposure to the injunctive norm message. In both studies, significant differences between the social norm and health message conditions were observed in low but not high usual consumers of fruit and vegetables.
Conclusions: For the promotion of healthy eating, social norm messages may be more effective than health messages for consumers failing to adhere to dietary guidelines.
----------------------
Impact of non-physician health professionals' BMI on obesity care and beliefs
Sara Bleich et al.
Obesity, forthcoming
Objective: Examine the impact of non-physician health professional body mass index (BMI) on obesity care, self-efficacy, and perceptions of patient trust in weight loss advice.
Methods: A national cross-sectional internet-based survey of 500 US non-physician health professionals specializing in nutrition, nursing, behavioral/mental health, exercise, and pharmacy collected between January 20 and February 5, 2014 was analyzed.
Results: Normal BMI professionals were more likely than overweight/obese professionals to report success in helping patients achieve clinically significant weight loss (52% vs. 29%, P = 0.01). No differences by health professional BMI about the appropriate patient body weight for weight-related care (initiate weight loss discussions and success in helping patients lose weight), confidence in ability to help patients lose weight, or in perceived patient trust in their advice were observed. Most health professionals (71%) do not feel successful in helping patients lose weight until they are morbidly obese, regardless of BMI.
Conclusions: Normal BMI non-physician health professionals report being more successful than overweight and obese health professionals at helping obese patients lose weight. More research is needed to understand how to improve self-efficiency for delivering obesity care, particularly among overweight and class I obese patients.
----------------------
The effect of non-caloric sweeteners on cognition, choice, and post-consumption satisfaction
Sarah Hill et al.
Appetite, December 2014, Pages 82–88
Abstract:
Consumers often turn to non-caloric sweeteners (NCS) as a means of promoting a healthy body weight. However, several studies have now linked their long-term use to increased weight gain, raising the question of whether these products produce unintended psychological, physiological, or behavioral changes that have implications for weight management goals. In the following, we present the results of three experiments bearing on this issue, testing whether NCS-consumption influences how individuals think about and respond to food. Participants in each of our three experiments were randomly assigned to consume a sugar-sweetened beverage, an unsweetened beverage, or a beverage sweetened with NCS. We then measured their cognition (Experiment 1), product choice (Experiment 2), and subjective responses to a sugar-sweetened food (Experiment 3). Results revealed that consuming NCS-sweetened beverages influence psychological processes in ways that – over time – may increase calorie intake.
----------------------
Public and health professionals’ misconceptions about the dynamics of body weight gain/loss
Tarek Abdel-Hamid et al.
System Dynamics Review, January-June 2014, Pages 58–74
Abstract:
Human body energy storage operates as a stock-and-flow system with inflow (food intake) and outflow (energy expenditure). In spite of the ubiquity of stock-and-flow structures, evidence suggests that human beings fail to understand stock accumulation and rates of change, a difficulty called the stock–flow failure. This study examines the influence of health care training and cultural background in overcoming stock–flow failure. A standardized protocol assessed lay people's and health care professionals’ ability to apply stock-and-flow reasoning to infer the dynamics of weight gain/loss during the holiday season (621 subjects from seven countries). Our results indicate that both types of subjects exhibited systematic errors indicative of use of erroneous heuristics. Indeed 76% of lay subjects and 71% of health care professionals failed to understand the simple dynamic impact of energy intake and energy expenditure on body weight. Stock–flow failure was found across cultures and was not improved by professional health training. The problem of stock–flow failure as a transcultural global issue with education and policy implications is discussed.
----------------------
Measures of adiposity predict interleukin-6 responses to repeated psychosocial stress
Christine McInnis et al.
Brain, Behavior, and Immunity, forthcoming
Objective: Overweight and obese individuals, who comprise approximately two-thirds of the U.S. population, are at increased risk for developing a range of diseases. This increased risk may be due in part to maladaptive stress responses within this group, including heightened low-grade inflammation and HPA axis non-habituation. In this study we tested the relationship between adiposity, plasma interleukin-6 (IL-6) and HPA axis responses to repeated stress.
Methods: Sixty-seven healthy participants were exposed to the Trier Social Stress Test (TSST) on two consecutive days. We collected saliva for cortisol measurements at baseline and at 1, 10, 30, 60 and 120 min post-TSST, and blood for plasma IL-6 measurements at baseline and 30 and 120 min post-TSST.
Results: Stress exposure induced significant increases of cortisol and IL-6 on both days (cortisol: F = 38, p < 0.001; IL-6: F = 90.8; p < 0.001), and repeated exposure was related with cortisol habituation (F = 8.2; p < 0.001) and IL-6 sensitization (F = 5.2; p = 0.022). BMI and body fat were related with higher cortisol responses to repeated stress (BMI: beta = 0.34; p = 0.014; body fat: beta = 0.29; p = 0.045), and with higher IL-6 responses to repeated stress (BMI: beta = 0.27, p = 0.044; body fat: beta = 0.37; p = 0.006).
Conclusions: Taken together, individuals with higher measures of adiposity showed less efficient HPA axis habituation as well as sensitization of IL-6 responses to repeated acute stress. These findings point to maladaptive stress response patterns in overweight humans, which, through exposure to higher levels of inflammatory mediators, might partially explain diseases related with overweight and/or obesity.
----------------------
Individual differences in striatum activity to food commercials predict weight gain in adolescents
Sonja Yokum et al.
Obesity, forthcoming
Objective: Adolescents view thousands of food commercials annually, but little is known about how individual differences in neural response to food commercials relate to weight gain. To add to our understanding of individual risk factors for unhealthy weight gain and environmental contributions to the obesity epidemic, we tested the associations between reward region (striatum and orbitofrontal cortex [OFC]) responsivity to food commercials and future change in body mass index (BMI).
Methods: Adolescents (N = 30) underwent a scan session at baseline while watching a television show edited to include 20 food commercials and 20 nonfood commercials. BMI was measured at baseline and 1-year follow-up.
Results: Activation in the striatum, but not OFC, in response to food commercials relative to nonfood commercials and in response to food commercials relative to the television show was positively associated with change in BMI over 1-year follow-up. Baseline BMI did not moderate these effects.
Conclusions: The results suggest that there are individual differences in neural susceptibility to food advertising. These findings highlight a potential mechanism for the impact of food marketing on adolescent obesity.
----------------------
John Milton Adams et al.
Appetite, December 2014, Pages 10–18
Abstract:
In the present research, we offer a novel method for informing consumers about the sugar content in sugar-sweetened beverages (SSBs). With a series of experiments, we present evidence that this method curbs preference for SSBs and leads to more negative attitudes toward SSBs. We propose that people view SSBs more negatively and show less preference for SSBs when they are able to concretely visualize the quantity of sugar in SSBs. For example, we suggest that people might have more negative views toward the idea of consuming 28 sugar cubes (concrete information), compared to consuming “70g” of sugar (abstract information). Indeed, we found that, without any intervention, people struggle to convert sugar grams into a concrete, physical sugar representation (Experiment 1). But, when people are provided ways to convert abstract sugar-nutrition information into a concrete representation, they find SSBs less attractive (Experiment 2) and are less likely to select SSBs in favor of sugar-free beverage options (Experiments 3 and 4). These findings offer direct applications to the design of public-health messages and nutrition-education interventions. Such applications might benefit society in its battle with the obesity epidemic.
----------------------
Tetyana Pudrovska et al.
Journal of Health and Social Behavior, September 2014, Pages 283-301
Abstract:
Using the 1957–1993 data from the Wisconsin Longitudinal Study, we explore reciprocal associations between socioeconomic status (SES) and body mass in the 1939 birth cohort of non-Hispanic white men and women. We integrate the fundamental cause theory, the gender relations theory, and the life course perspective to analyze gender differences in (a) the ways that early socioeconomic disadvantage launches bidirectional associations of body mass and SES and (b) the extent to which these mutually reinforcing effects generate socioeconomic disparities in midlife body mass. Using structural equation modeling, we find that socioeconomic disadvantage at age 18 is related to higher body mass index and a greater risk of obesity at age 54, and that this relationship is significantly stronger for women than men. Moreover, women are more adversely affected by two mechanisms underlying the focal association: the obesogenic effect of socioeconomic disadvantage and the SES-impeding effect of obesity. These patterns were also replicated in propensity score–matching models. We conclude that gender and SES act synergistically over the life course to shape reciprocal chains of two disadvantaged statuses: heavier body mass and lower SES.
----------------------
Occupational Conditions, Self-Care, and Obesity among Clergy in the United States
Todd Ferguson et al.
Social Science Research, forthcoming
Abstract:
Prior research has shown that a variety of occupational conditions such as long work hours are associated with increased likelihood of obesity. In this study, we use the specific case of the clergy to explore how occupational conditions are linked to increased or decreased odds of being obese. We hypothesize that stressful conditions are associated with increased odds of obesity and that self-care practices are associated with decreased odds. Using the 2008/9 U.S. Congregational Life Survey’s national sample of clergy from multiple religious traditions, we find support for our hypotheses. Clergy who experience more stress, work more hours, or are bi-vocational have higher odds of obesity. Those who take a day off each week, have taken a sabbatical, or are involved in a support group experience lower odds. For Protestant clergy, being involved in a support group or taking a day off moderates the association between certain stressful occupational conditions and obesity.
----------------------
Peggy Liu et al.
Duke University Working Paper, May 2014
Abstract:
We introduce a simple solution to help consumers manage choices between healthy and unhealthy food options: vice-virtue bundles. Vice-virtue bundles are item aggregates with varying proportions of both vice and virtue, holding overall quantity constant. Four studies compare choice and perceptions of differently composed vice-virtue bundles relative to one another and to pure vice and pure virtue options. Results suggest that people tend to prefer vice-virtue bundles with small (¼) to medium (½) proportions of vice rather than large (¾) proportions of vice. Moreover, people rate vice-virtue bundles with small vice proportions as healthier but equally tasty as bundles with larger vice proportions. For most individuals, choice patterns are different from those predicted by variety-seeking accounts alone. Instead, these findings provide evidence that bundle choice can be predicted by the identification of a taste-health balance point, determined based on consumers’ perceptions of tastiness and healthiness as a function of the relative proportion of vice in an option.
----------------------
Obesity history and male employment
Francesco Renna
Applied Economics Letters, forthcoming
Abstract:
Using the National Longitudinal Survey of Youth 1979, this article computes the stock of obesity as the number of obese years in the adult life of an individual. Then it estimates the effect of the stock of obesity on the probability of being employed. It is found that the accumulated years of morbid obesity (i.e. obesity associated with a body mass index above 40) has a large negative impact on employment status. This effect remains significant even after conditioning on time-invariant unobserved heterogeneity. The results of the IV probit analysis indicate that the stock of morbid obese years can be regarded as exogenous. Less severe levels of obesity do not seem to have an impact on employment.
----------------------
Explaining educational disparities in adiposity: The role of neighborhood environments
Gavin Abbott et al.
Obesity, forthcoming
Objective: To examine the extent to which characteristics of the neighborhood built environment explain the association between adiposity and educational qualifications in Australian women.
Methods: A community sample of 1,819 women (aged 18-66) from the Melbourne SESAW study provided information regarding their body mass index (BMI) and level of education. Objective measures of participants' residential neighborhood built environments were obtained using a Geographic Information System.
Results: Compared with women with a high school degree or above, women who did not complete high school had higher average BMI, which was partially explained by lower density of sports facilities and living less proximally to the coastline and to supermarkets. In a multiple mediator model, which explained 16.6% of the educational disparity in BMI, the number of sports facilities and presence of the coastline within 2 km of participants' homes were significant mediators of the observed socioeconomic disparity in BMI.
Conclusions: The residential neighborhood environment may help to explain socioeconomic patterning of overweight and obesity in Australian women. These results provide further support for considering the built environment in obesity prevention initiatives, suggesting a potential role in decreasing social inequalities in obesity.