Figure. Relative odds of obesity associated with migration in Mexico and the United States. A, Men. B, Women.
Flórez KR, Dubowitz T, Saito N, Borges G, Breslau J. United States–Mexico migration and the prevalence of obesity: a transnational perspective. Arch Intern Med.. Published online November 12, 2012. doi:10.1001/2013.jamainternmed.77.
eTable Odds of obesity by migration category and sex
Flórez KR, Dubowitz T, Saito N, Borges G, Breslau J. Mexico–United States Migration and the Prevalence of Obesity: A Transnational Perspective. Arch Intern Med. 2012;172(22):1760-1762. doi:10.1001/2013.jamainternmed.77
Author Affiliations: RAND Corp, Santa Monica, California, and Pittsburgh, Pennsylvania (Drs Flórez, Dubowitz, and Breslau); Department of Public Health Sciences, University of California Davis, Sacramento (Ms Saito); and Instituto Nacional Psiquiatría Ramón de la Fuente, México City, Mexico (Dr Borges).
Country of birth and length of stay in the United States have proven to be strong predictors of obesity among Mexican Americans,1 suggesting the US environment may be distinctively “obesogenic.”2 For example, a 12-oz bottle of American-made Coca-Cola has 240 calories with 65 g of sugar, whereas Mexican-made Coca-Cola has 150 calories per 12-oz bottle with 39 g of sugar (the former is made from high-fructose corn syrup).3,4 However, there is also evidence that immigrants are resistant to these influences: growth in body mass index (BMI), calculated as weight in kilograms divided by height in meters squared, is slower among immigrants than among US-born Mexican Americans.5 Studies have yet to examine the relationship between migration and obesity in a transnational perspective, including comparisons with the Mexican source population to help identify patterns distinctive to the United States.
Data from epidemiological surveys in Mexico (the Mexican National Comorbidity Survey6 [MNCS]) and the United States (the Collaborative Psychiatric Epidemiology Surveys7 [CPES]) were combined (N = 3244 respondents). Obesity was defined as a BMI greater than 30 using self-reported height and weight. Respondents with missing weight or height (n = 266 respondents), implausibly high BMI (>65) (n = 3 respondents), and current pregnancy (n = 62 respondents) were excluded. Comparison groups were defined using information on respondents' personal and familial connection to Mexico-US migration. The MNCS respondents were divided into 3 groups: (1) Mexicans who have never been to the United States and do not have a migrant in their immediate family (living in Mexico, no migrant in family; n = 1050); (2) Mexicans who have not been to the United States but have a migrant in their immediate family (living in Mexico, migrant in family; n = 955); and (3) Mexicans who have previously been migrants in the United States (living in Mexico, previous migrant; n = 126). Respondents in the United States were also divided into 3 groups: (1) Mexican-born immigrants (first generation in the United States; n = 509); (2) US-born with 1 or more Mexican-born parent (second generation in the United States; n = 285); and (3) US-born with US-born parents who self-identified themselves as Mexican American (third generation in the United States; n = 319). Covariates included age (continuous), marital status (married, divorced, never married), educational attainment (0-5, 6-8, 9-11, or ≥12 years), and current smoking status. Analyzes were conducted using SUDAAN software (SAS Institute Inc) to adjust for the complex survey design.
With statistical adjustment for age, marital status, education, and smoking, the odds of obesity among men were higher among the first generation in the United States (odds ratio [OR], 1.66 [95% CI, 1.10-2.52]), the second generation in the United States (OR, 3.38 [95% CI, 1.84-6.20]), and the third generation in the United States (OR, 2.68 [95% CI, 1.48-4.86]), relative to men living in Mexico, with no migrant in family (Figure). Among women, the adjusted odds of obesity were higher for the first generation in the United States (OR, 2.62 [95% CI, 1.72-4.00]), second generation in the United States (OR, 3.08 [95% CI, 1.81-5.23]), and third generation in the United States (OR, 3.79 [95% CI, 2.19-6.57]) relative to women living in Mexico with no migrant in family. Among women but not among men, respondents living in Mexico with a family member in the United States were more likely to be obese than those with no migrants in their family (OR, 1.73 [95% CI, 1.14-2.62]). Also see the eTable .
Consistent evidence reveals greater odds of obesity among US-born Mexican Americans relative to their first-generation counterparts. This study extended this comparison by including those in Mexico and revealed that the gap between first-generation immigrants and the US-born is one part of a graded increase in obesity associated with migration to the United States. This is important in light of a longitudinal analysis that suggested that first-generation immigrants may be resistant to the obesogenic environment in the United States.5 This cross-sectional comparison suggests otherwise. We found slight differences by sex, but results indicate a roughly 3-fold increase in obesity from one extreme to the other for both sexes.
Second, we found that among Mexicans with no direct migration experience, having a migrant in the immediate family is associated with a higher risk for obesity among women but not for men. This finding may reflect economic influences on diet, such as cash remittances sent by migrants working in the United States.
Findings should be interpreted in light of the use of cross-sectional data and reliance on self-report of height and weight. Self-reports tend to underestimate the prevalence of obesity, but evidence suggests that self-report does not differ between immigrant and US-born Mexican Americans, except for those who are underweight.8
Migration is a transnational process that is likely to have a range of health effects in both sending and receiving countries, including diet. Given that obesity is a risk factor for the major causes of mortality in this country, growing rates among Mexican Americans is of public health and clinical urgency.
Correspondence: Dr Flórez, RAND Corp, 1776 Main St, Santa Monica, CA 90407-2138 (email@example.com).
Published Online: November 12, 2012. doi:10.1001/2013.jamainternmed.77
Author Contributions:Study concept and design: Flórez, Dubowitz, Borges, and Breslau. Analysis and interpretation of data: Flórez, Dubowitz, Saito, and Breslau. Drafting of the manuscript: Flórez, Dubowitz, and Breslau. Critical revision of the manuscript for important intellectual content: Flórez, Dubowitz, Saito, Borges, and Breslau. Statistical analysis: Flórez, Saito, Borges, and Breslau. Obtained funding: Breslau. Administrative, technical, and material support: Saito. Study supervision: Dubowitz and Breslau.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grant 7R01MH082023-04 from the National Institute of Mental Health (NIMH) (Dr Breslau).
Addition Contributions: Kathryn Derose, PhD, MPH, and Kristin Leuschner, PhD (RAND Corp); Sergio Aguilar-Gaxiola, MD PhD (Center for Reducing Health Disparities, University of California, Davis); and Maria Elena Medina-Mora, PhD (Instituto Nacional Psiquiatría Ramón de la Fuente), provided comments on the manuscript.
Disclaimer: The NIMH had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.