- Email: [email protected]
Life course, sociocultural factors and disordered eating in adult Mexican women
Appetite 121 (2018) 207e214
Contents lists available at ScienceDirect
Appetite journal homepage: www.elsevier.com/locate/appet
Life course, sociocultural factors and disordered eating in adult Mexican women nica Valdez c, Claudia Unikel b, * Ietza Bojorquez a, Joyce Bustos b, Vero El Colegio de la Frontera Norte, Mexico, Km. 18.5 Carretera Esc enica Tijuana-Ensenada, San Antonio del Mar, Tijuana CP 22506, BC, Mexico n de la Fuente Mun ~ iz, Mexico, Calz M Instituto Nacional de Psiquiatría Ramo exico-Xochimilco 101, Huipulco, 14370 Ciudad de M exico, Mexico c n, Copilco Universidad, 04510 Ciudad de M noma de M Universidad Nacional Auto exico, Mexico, Avenida Universidad 3004, Coyoaca exico, Mexico a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 May 2017 Received in revised form 17 October 2017 Accepted 7 November 2017 Available online 9 November 2017
Disordered eating (DE) can appear in women of all ages and in diverse sociocultural contexts, however most research focuses on younger women in higher income countries. The purpose of this article was to explore the association of life course markers with DE, considering the effects of sociocultural factors, in a sample of adult women in Tijuana, Mexico. We employed data from a household survey (n ¼ 2322) conducted in 2014, to evaluate the associations of DE with age, occupation, marital status and having children (life course markers), and indicators of social position and exposure to modernization (sociocultural factors). The prevalence of weight preoccupation was 69.2% (CI95% 67.3,71.1), the prevalence of dieting 24.8% (CI95% 22.4,27.3), and 2.0% (CI95% 1.4,3.0) had a probable eating disorder according to the questionnaire cutoff score. In the adjusted model, younger age, being employed, higher social position and indicators of exposure to modernization had positive associations with DE. There were interactions between marital status and body mass index, and between age and region of birth. The interaction terms showed that overweight was positively associated with DE among single and cohabiting participants, but not among the married ones; and that the negative association between DE and age was apparent from younger age groups in women born in less developed regions of Mexico. Our results replicate others in showing DE to be present in women through the life course, and point to at-risk groups in the confluences of life course, social position and modernization. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Disordered eating Adult women Life course Social factors
1. Introduction The term “disordered eating” (DE) refers to attitudes and behaviors that are similar to those that appear in eating disorders, without meeting the clinical criteria for the latter. Examples include weight-related preoccupation, dieting, use of medical and nonmedical substances such as laxatives or diuretics, and selfprovoked vomiting. Whether DE is a subclinical form, a prodrome or a risk factor for eating disorders is still unclear (Stice, Ng, & Shaw, 2010), but it can be harmful to health in itself, and people with DE report similar levels of affective disorders, anxiety and suicidality as those with well-established eating disorder diagnoses (Solmi, Hatch, Hotopf, Treasure, & Micali, 2014). DE is also associated
n de Investigaciones Epidemiolo gicas y Psico* Corresponding author. Direccio n de la Fuente Mun ~ iz, Calzada sociales, Instituto Nacional de Psiquiatría Ramo Mexico-Xochimilco 101, Tlalpan, D.F., CP 14370, Mexico. E-mail addresses: [email protected] (I. Bojorquez), [email protected]. mx (C. Unikel). https://doi.org/10.1016/j.appet.2017.11.085 0195-6663/© 2017 Elsevier Ltd. All rights reserved.
with poorer quality of life (Herpertz-Dahlmann et al., 2008; Wade, Wilksch, & Lee, 2012). Risk factors for DE are manifold, including genetics, personality, cognitive style and adverse life events (Kally & Cumella, 2008; Stice, Marti, & Durant, 2011; Stice et al., 2010). At the same time, the increase in the prevalence of DE in the 20th century evidences the importance of sociocultural aspects, and particularly of the social construction of femininity, as DE appears to have increased in parallel with changes in female gender roles (Bordo, 1993; Nichter, 2001; Paquette & Raine, 2004; Rodin, Silberstein, & Striegel-Moore, 1984). These changes are part of a general process of social and cultural modernization, in which the production of a successful identity that includes a certain type of body becomes more and more a central personal project (Giddens, 1991; Turner, 1984). In combination with the thin ideal (the cultural norm that considers slimness as essential for female beauty), the importance attributed to personal cultivation via practices that modify the body (Crossley, 2006) provides the sociocultural context in which DE arises. In more recent times, the public health discourse against obesity has
208
I. Bojorquez et al. / Appetite 121 (2018) 207e214
added to the discourses that promote the thin ideal (Kwan, 2009), increasing the risk of DE. Given the strong association of DE with gender-related social expectations, and the changes in these expectations as women pass through the life course, the manifestations and factors associated with DE in women of different age groups are topics worth of study. While most research on DE focuses on adolescents or young women, an increasing number of studies show that DE behaviors and attitudes are present well into adulthood (Fiske, Fallon, Blissmer, & Redding, 2014; Kilpela, Becker, Wesley, & Stewart, 2015; Slevec & Tiggemann, 2011). In this regard, most research shows a negative association between age and DE (Gadalla & Piran, 2008; Solmi et al., 2014), or an inverted U-shaped relationship that peaks between the ages of 30 and 40 (Matthiasdottir, Jonsson, & Kristjansson, 2012; Mitchison, Hay, Slewa-Younan, & Mond, 2014; Runfola et al., 2013; Tiggemann & Lynch, 2001). However, body shape concerns can also increase as the aging body deviates from the cultural ideal of slimness and youth (Becker, Diedrichs, Jankowski, & Werchan, 2013; Slevec & Tiggemann, 2011) and preoccupation over the body's functioning might also increase (McLean, Paxton, & Wertheim, 2010). At the same time, some studies have found that age decreases the importance attributed to appearance, which might result in less body dissatisfaction and DE (Kilpela et al., 2015). Most of this research has been conducted in upper-income countries (mainly English-speaking), so there is a lack of information about the association of age and DE in other societies. As for the effect of other life course markers, some studies have reported that marriage and motherhood are associated with decrease in DE (Keel, Baxter, Heatherton, & Joiner, 2007), while others find that pregnancy and the postpartum period increase it (Lai, Tang, & Tse, 2006). Some authors suggest that achievement of life course markers such as establishing a career, having children and having a committed long-term relationship might decrease the perceived importance of body shape (Tiggemann, 2015), and qualitative studies have shown that the significance that women attach to their relational identities as mothers can be a protective factor as it helps them to oppose the discourse of self-enhancement through weight control (Bojorquez-Chapela, Unikel, Mendoza, & de Lachica, 2014; Warin, Turner, Moore, & Davies, 2008). While there can be similarities between the experiences of women in different social groups, the life course also interacts with other aspects of the social position to produce experiences of the body, of the importance of weight control and of DE, that are shared with persons in similar positions. Thus, quantitative studies of the association of social position and DE in adults in different regions of the world show that the upper classes are more likely to engage in weight control practices and to be dissatisfied with their appearance even while at a normal weight (Alwan, Viswanathan, Williams, Paccaud, & Bovet, 2010; Machado, Silveira, & Silveira, 2012; Mangweth-Matzek et al., 2014; von Lengerke, Mielck & KORA Study Group, 2012; Williams, Germov, & Young, 2011). The qualitative studies mentioned above also show that upper-class mothers are more conscious of the social expectation of thinness (Bojorquez-Chapela et al., 2014; Warin et al., 2008). On the other hand, while differences by social class in the prevalence of weight-related preoccupation and weight control practices are clear, comparisons between countries show a more complicated picture. If DE is related to modernization, a gradient from lower to higher income countries should be apparent, inasmuch as economic development is associated with other modernizing social processes. However, the World Health Organization worldwide surveys found no clear gradient in the case of
eating disorders (Kessler et al., 2013), and the use of different instruments and special-population samples limits comparisons between countries of the prevalence of DE. The association of exposure to social modernization with DE is still an interesting subject of study, and countries with ample differences in socioeconomic and human development, and varying local and regional cultures, are sites where these associations can be explored (Lee & Lee, 2000). To summarize, the life course, social position, and exposure to modernization, all play a role in DE, but their independent effects and possible interactions are still unclear. Most research in this area has been conducted in high-income countries of Europe and North America, with adolescents or young women, or with special populations such as students. In this article, we assess the association of life course markers with DE in adult women, at the same time exploring the independent effects of indicators of social position and of exposure to modernization, and their interactions with the life course. Following from the literature referenced above, we hypothesized that higher social position and indicators of exposure to modernization would be positively associated with DE. As most literature reports a negative association between life course markers and DE, our second set of hypotheses was that older age, being employed, being married and having children would be associated negatively with DE. Finally, we expected that social position and indicators of exposure to modernization would modify the association between life course markers and DE. 2. Methods 2.1. Site and sample In FebruaryeApril 2014, we conducted a cross-sectional survey on a representative sample of the population of women ages 18 to 65 living in Tijuana, Mexico. Tijuana is a city with over 1.5 million inhabitants on the Mexico-United States (US) border, it has a high rate of internal migration, and is also a transit city for migrants heading to the US(International Organization for Migration (IOM) (2015)). The city has intense dynamics of cross-border relations, with over 40 million annual border crossings (del Castillo, Peschard-Sverdrup, & Fuentes, 2007), and can in many respects be regarded as an intermediate place between Mexican and US culture (Valenzuela, 2003). The survey had a probabilistic, stratified, multi-stage cluster sample design. In the first stage, Basic Statistical Geographical Areas as defined by Mexico's National Institute of Statistics and Geography (INEGI) were divided into strata according to a Marginalization Index reported by Mexico's National Population Council (CONAPO). The Marginalization Index classifies geographical areas by level of social development, and is computed from indicators such as literacy rate, percentage of dwellings with electricity and percentage of population with low income. In the second stage, blocks were randomly selected from each area. All households on the block were visited, and one woman of the required age was randomly selected from each household and invited to participate in the survey. A required sample size of 2500 respondents was predefined. The eligibility criteria were: 1) being 18e65 years old, and 2) agreeing to take part in the survey after a process of informed consent. Potential participants were told of the purpose of the project and the nature of participation, and asked to sign a letter of consent. All procedures were approved by the Ethics Committee of El Colegio de la Frontera Norte. Questionnaires were administered face to face by previously trained interviewers, who also weighed and measured participants using Tanita BF-683W
I. Bojorquez et al. / Appetite 121 (2018) 207e214
scales, and portable SECA 213 stadiometers.
2.2. Variables We measured DE with the Brief Questionnaire of Risky Eating Behaviors (BQREB) (Unikel-Santoncini, Bojorquez-Chapela, & ~ o-Garcia, 2004). The BQREB includes 10 items in a Likert Carren scale, asking about the frequency of restrictive and purging behaviors, binge eating and weight-related concerns over the past three months (see items in Table 2). Response options range from “never or almost never” to “very often (three or more times per week)” and are rated from 0 to 3. Possible scores on the questionnaire thus range from 0 to 30 where a higher score indicates greater frequency. The BQREB was developed and validated in Mexican population and has been used in the National Health rrez Surveys in Mexico (Barriguete-Melendez et al., 2009; Gutie et al., 2012). In a validation study comparing scores in a sample of students and a clinical sample, internal consistency as measured by Cronbach's a was 0.83, and a cutoff >10 had a sensitivity of 0.81 and specificity of 0.78 for the identification of cases of eating disorders (Unikel-Santoncini et al., 2004). While there is no one single accepted definition of DE in the literature, in this article we employed a score >10 in the BQREB as an indicator of a DE level that could be of clinical significance. As indicators of the life course we employed age and life course markers that are important milestones in the passage to adulthood in Mexico (de Oliveira & Ariza, 2002; de Oliveira & Mora, 2008): having children, marital status, and occupation. In exploratory bivariate analysis we observed an inverted-U association of DE with age (continuous variable), but it disappeared after adjustment by other predictors in the model and subsequently we modeled age as a categorical variable (18e29, 30e40 and 41e65 years) for easiness of presentation. As indicators of social position we employed education (in categories by years of education) and an index of socioeconomic level (SEL) computed through principal components analysis of questions on the presence of goods and services in the household. As indicators of exposure to modernization we included region (within Mexico) or country of origin, Internet access at home, and frequency of crossing to the US. We categorized the region of origin following a classification that separates Mexican states in four groups according to their geographic, economic and social characteristics (Palma-Coca et al., 2011), and after exploratory analysis combined these into two groups: the northern (more developed) region, and the central-southern (less developed) region. Out of 36 participants born in other countries, 34 had been born in the US. Therefore, for analysis, we excluded the two participants born elsewhere and considered three categories: northern Mexico, southern Mexico, and US. Since body mass index (BMI) (weight in kilos/height in meters2) is associated with both age and DE, in the final analysis we controlled for this variable as calculated from anthropometric measurements. Values outside the range 10e58 (five cases) were ndez-Bareliminated as invalid (Barquera, Campos-Nonato, Herna rera, Pedroza, & Rivera-Dommarco, 2013), and BMI was recoded into a categorical variable, according to the cutoff points of the World Health Organization (low <18.5, normal 18.5e24.9, overweight 25e29, obesity 30, frequencies shown in Table 1). For multivariate analysis, because of small counts in some combinations of age and BMI, we merged low and normal into one category. The results of analyses separating the low and normal categories were similar to the ones reported, and there were no cases of DE in the low category. Pregnant women (121 cases) were excluded from analyses that considered BMI.
209
2.3. Analysis First, we explored the distribution of the variables of interest, and of their joint distributions. As the BQREB score is a discrete numerical variable, and the over-dispersion parameter (alpha) was statistically significant, we employed a negative binomial regression model for the association analysis. Following a theoretical criterion, variables that were part of our conceptual framework were included in the model regardless of their p-value. We explored nonlinear associations and interactions (Ganzach, 1997), testing for interactions between each of the life course variables (age group, having children, marital status and occupation), and the covariates. Interactions with p < 0.05 were kept in the model. All analyses took into account the sample design, using Taylor linearized series to adjust the standard errors, and were carried out using the “svy” module in Stata, version 13.1. To assess the model fit, we considered the likelihood ratio test comparing it to a Poisson model, comparison of observed vs. predicted counts, and the Akaike Information Criteria (AIC) in unweighted models. 3. Results A total of 2500 women were selected to participate in the study, of whom 2345 (93.8%) answered the questionnaire. Of these, 2322 (99.0%) provided complete answers in the BQREB. The characteristics of the respondents who answered the BQREB are given in Table 1. Most participants had children, were either married or cohabiting, and half were homemakers or stayed at home for other reasons. Over half of participants had less than 10 years of education, and only 13.0% had some college education or more. The percentage born in central-southern Mexico was 16.6%, and 1.6% had been born in the US. Slightly over half of the sample had Internet access at home. About two thirds were either overweight or obese. The 23 respondents who did not give complete answers to the BQREB were similar in respect to observed characteristics (not shown in Table). In Table 1, we also show the percentage scoring above the BQREB cutoff point (i.e. probable clinically significant cases), by covariate levels. Among all participants with complete responses in the BQREB, the percentage scoring above the cutoff point was 2.0% (CI95% 1.4,3.0), and the mean score in the BQREB was 2.8 (CI95% 2.6,3.0). The most frequently reported item of the BQREB was concern about putting on weight, but dieting, exercising and feeling that one had ate too much were also relatively frequent (Table 2). In bivariate analysis (Table 3, second and third columns), DE was negatively associated with age, being separated or widowed, and being from central-southern Mexico; and positively associated with being employed, years of education, SEL, Internet access, border crossing and BMI. In the main effects multivariate model (Table 3, fourth and fifth columns), the negative associations between DE and age and being from central-southern Mexico were still significant, as were the positive associations between DE and being employed, having more than 12 years of education, being in the upper quintile of SEL, having Internet access at home, and higher BMI. In the model including interactions (Table 3, sixth and seventh columns) the effects of employment, education, SEL and Internet access remained significant. The coefficients for interactions between age group and region of origin showed that among participants born in northern Mexico, only those in the 41e65 age group had a significantly lower score as compared with the younger group, while for participants from central-southern Mexico the score was significantly lower also for the 30e40 age group. As the plot of predicted scores in Fig. 1 shows, this means that the negative slope of the association between age and DE started earlier for
210
I. Bojorquez et al. / Appetite 121 (2018) 207e214
Table 1 Characteristics of participants, and percentage with score over questionnaire cutoff.a Variable
Age group 18e29 years 30e40 years 41e65 years Has no children Has children Marital status Single Married Cohabiting Separated/widowed Occupation Home/not employed Student Employed Years of education 0-9 10-12 12þ Socioeconomic level (quintile) 1 (lower) 2 3 4 5 (upper) Born in Northern Mexico Central-southern Mexico United States No Internet access at home Internet access at home No international border crossing International border crossing Body Mass Index:c Low Normal Overweight Obese a b c
Characteristics of participants
Percentage over cutoff, by characteristic
Numberb
Percentageb
Confidence Interval 95%b
Numberb
Percentageb
Confidence Interval 95%b
903 557 862 475 1847
38.8 24.1 37.1 20.3 79.8
38.0,39.6 22.5,25.8 35.2,39.0 18.0,22.7 77.3,82.0
22 18 8 5 43
2.5 3.0 0.9 1.2 2.3
1.5,4.1 1.8,4.8 0.3,3.0 0.4,3.1 1.6,3.2
522 1125 424 251
22.6 49.1 17.5 10.8
21.1,24.2 46.3,51.9 15.7,19.4 9.6,12.2
9 22 14 3
2.3 1.8 3.0 1.0
1.0,5.1 1.2,2.7 1.4,6.3 0.3,3.1
1156 163 1003
50.0 6.9 43.1
49.2,50.8 5.9, 8.0 41.7,44.6
18 2 28
1.6 2.0 2.6
1.0,2.5 0.4,9.0 1.6,4.0
1300 653 271
57.3 29.7 13.0
52.4,62.0 27.1,32.4 9.9,16.9
22 19 4
1.7 2.7 1.6
0.9,3.3 1.7,4.3 0.5,4.7
362 463 542 302 615
14.8 19.4 24.4 13.6 27.9
12.6,17.3 17.0,22.0 21.3,27.7 12.0,15.4 23.1,33.2
7 10 6 8 16
1.7 2.6 1.0 2.0 2.8
0.8,3.6 1.1,5.9 0.4,2.6 0.8,5.0 1.7,4.5
1861 406 32 1154 1164 1466 810
81.8 16.6 1.6 48.2 51.8 62.2 37.8
79.7,83.7 14.8,18.5 1.1, 2.4 43.9,52.5 47.5,56.1 57.5,66.7 33.3,42.5
38 6 3 19 29 32 16
2.0 1.1 12.5 1.6 2.4 2.2 1.9
1.2,3.3 0.5,2.7 4.4,30.5 0.7,3.4 1.7,3.5 1.2,3.9 1.1,3.2
18 506 700 590
1.1 28.4 38.5 32.0
0.7, 1.9 25.8,31.1 35.8,41.2 29.6,34.6
0 8 16 17
0 2.3 2.0 2.6
1.1,4.6 1.1,3.4 1.3,5.5
Participants with complete answers in the Brief Questionnaire of Risky Eating Behaviors (Unikel-Santoncini et al., 2004). Unweighted n ¼ 2322. Number is unweighted n, percentages and means are weighted, confidence intervals consider sampling design. Only for participants who were not pregnant at the time of interview.
Table 2 Prevalence of disordered eating attitudes and behaviors.a In the past three months, how often have you … ?
1/week or more Number
Worried about gaining weight Eaten too much Lost control over eating Vomited to lose weight Fasted to lose weight Dieted to lose weight Exercised to lose weight Taken pills to lose weight Taken diuretics to lose weight Taken laxatives to lose weight a b
1588 617 349 32 140 556 726 193 71 112
b
Percentage
3/week or more b
69.2 27.4 15.2 1.4 6.5 24.8 32.9 8.5 3.4 5.0
b
Confidence Interval 95%
Numberb
Percentageb
Confidence Interval 95%b
67.3,71.1 25.4,30.0 13.4,17.3 1.0,2.1 5.1,8.1 22.4,27.3 29.9,36.1 7.4,9.9 2.6,4.4 4.0,6.3
375 15 9 3 6 51 110 33 12 11
16.5 0.7 0.4 0.1 0.2 2.0 5.1 1.3 0.5 0.6
14.7,18.5 0.4,1.1 0.2,0.7 0.0,0.2 0.1,0.6 1.5,2.7 4.0,6.6 0.8,1.9 0.2,0.9 0.3,1.2
Unweighted n ¼ 2322. Number is unweighted n, percentages and means are weighted, confidence intervals consider sampling design.
those born in the less developed areas of Mexico. The interaction between marital status and BMI showed that for single and cohabiting participants there was a significant positive effect of being in either the overweight or obese categories, while for married participants the effect was only significant for those in the obese category. Besides, the difference in mean scores between the low/normal category and the obese category was more marked for
cohabiting participants (Fig. 2). 4. Discussion As others before (Gadalla & Piran, 2008; Solmi et al., 2014), in this study we found that scores in the BEQRB were lower in the older groups. We also found the previously reported inverted U-
I. Bojorquez et al. / Appetite 121 (2018) 207e214
211
Table 3 Association between life course and other variables, and disordered eating.a Variable
Main effects modelb
Bivariate models Regression Coefficient
Age group 18-29 Ref. 30-40 0.05 41-65 0.33 Occupation Home/not employed Ref. Student 0.17 Employed 0.19 Marital status Single Ref. Married 0.05 Cohabiting 0.13 Separated/Widowed 0.15 Has children 0.00 Years of education 0-9 Ref. 10-12 0.19 12þ 0.34 Socioeconomic level (quintile) 1 (lower) Ref. 2 0.15 3 0.07 4 0.12 5 (upper) 0.36 Born in Northern Mexico Ref. Central-southern Mexico 0.13 United States 0.37 Internet access at home 0.24 Border crossing 0.14 Body Mass Index: Low/Normal Ref. Overweight 0.12 Obese 0.26 Age group * born in 30e40yrs * CS Mexicoc Marital status * BMI Single*Overweight Cohabiting*Overweight Cohabiting*Obese
Model with interaction termsb
Confidence Interval 95%
Regression Coefficient
Confidence Interval 95%
Regression Coefficient
Confidence Interval 95%
0.06, 0.16 0.43,-0.23
Ref. 0.07 0.46
0.22, 0.08 0.63,-0.30
Ref. 0.01 0.41
0.19, 0.16 0.58,-0.23
0.02, 0.36 0.11, 0.28
Ref. 0.23 0.23
0.11, 0.56 0.11, 0.36
Ref. 0.21 0.21
0.11, 0.53 0.09, 0.32
Ref. 0.07, 0.16 0.02, 0.29 0.30,-0.00 0.08, 0.09
Ref. 0.03 0.18 0.01 0.14
0.19, 0.01, 0.21, 0.02,
Ref. 0.23 0.24 0.28 0.11
0.48, 0.55, 0.72, 0.05,
0.07, 0.32 0.19, 0.48
Ref. 0.11 0.21
0.03, 0.25 0.02, 0.40
Ref. 0.12 0.20
0.02, 0.25 0.01, 0.38
0.02, 0.32 0.08, 0.22 0.02, 0.25 0.19, 0.54
Ref. 0.14 0.07 0.05 0.18
0.04, 0.29 0.09, 0.22 0.09, 0.19 0.01, 0.35
Ref. 0.15 0.09 0.07 0.22
0.02, 0.31 0.07, 0.24 0.07, 0.22 0.04, 0.40
0.24,-0.02 0.00, 0.74 0.12, 0.36 0.02, 0.26
Ref. 0.17 0.15 0.23 0.06
0.30,-0.03 0.35, 0.67 0.12, 0.34 0.11, 0.23
Ref. 0.04 0.29 0.21 0.06
0.15, 0.23 0.32, 0.89 0.10, 0.32 0.10, 0.22
0.02, 0.27 0.12, 0.39
Ref. 0.19 0.51
0.02, 0.35 0.36, 0.67
Ref. 0.02 0.31
0.26, 0.22 0.08, 0.54
0.33
0.63,-0.02
0.38 0.43 0.66
0.08, 0.68 0.02, 0.84 0.20, 1.13
0.14 0.38 0.22 0.30
0.01 0.07 0.15 0.27
a
Negative binomial regression, dependent variable: score in the BQREB. Excluding participants who were pregnant at the time of interview. Adjusted for the variables and interactions shown in Table. Unweighted n ¼ 1718. Main effects model, F(20,28) ¼ 13.65, p < .0001. Interaction model F(30,18) ¼ 9.74, p < .0001. c Only interaction categories with p < .05 are shown in Table. b
shape association (Matthiasdottir et al., 2012; Mitchison et al., 2014; Runfola et al., 2013; Tiggemann & Lynch, 2001), but the non-linearity disappeared after adjusting by other predictors in the model, so that it was possibly the result of other characteristics. The association between age and DE was apparent from a younger (30e40 years) age group among participants born in centralsouthern Mexico. As mentioned above, in Mexico there is a north-south gradient of social and economic development, and also of closeness to US cultural elements. It is possible that more traditional gender role expectations protect southern adult women from DE, and although qualitative studies in central and southern rural areas (Kimoto et al., 2014; Perez-Gil & Romero, 2010; Romero rez, Perez Gil-Romo, & De la Rosa Cruz, 2017) show that weight Jua concerns and weight-control practices are widespread, the prevalence of DE is still higher in the northern regions (Palma-Coca et al., 2011). Similarly to Mangweth-Matzek et al. (2014), we observed no significant association between having children and DE. In preliminary analysis we explored associations of DE with number and age of children, and all of them were likewise non-significant. This finding was contrary to our hypothesis, and also to previous
qualitative studies reporting that for Mexican women the role of mother seems to decrease the interest in achieving or conserving a thin figure, and weight gain is naturalized as an inevitable consequence of childbearing (Bojorquez-Chapela et al., 2014). The finding by other authors that having children decreases DE (von Soest & Wichstrom, 2008) might be related to specific childbearing experiences, or be associated with other elements of the life course that confound the association with motherhood. In our study, being employed had a positive association with DE. Qualitative studies have shown that working, especially in professions that involve direct contact with clients, can make women more conscious of their appearance, and a thin shape can be a requisite for some jobs (Anderson-Fye, 2004). A qualitative study of women in Mexico showed that they were aware of the need for a slim figure in the workplace, and described how in that environment skipping meals, worrying about being fat, talking about diets and going on diet with colleagues was relatively common (Bojorquez-Chapela et al., 2014). Another possible explanation has to do with the social construction of female work in the Mexican society. Other studies have shown that in Mexico workingwomen show more depressive symptoms than women who stay at home
212
I. Bojorquez et al. / Appetite 121 (2018) 207e214
Fig. 1. Predicted score in disordered eating, by age group and region of origin.
Fig. 2. Predicted score in disordered eating, by category of body mass index and marital status.
(Bojorquez, Salgado de Snyder, & Casique, 2009), and while in other contexts work can have a positive meaning as the establishment of a career (Tiggemann, 2015), Mexican women could perceive it as conflicting with their expected gender role. The fourth life course indicator we employed, marital status, showed a complex association with DE, in that BMI had a greater effect for single and cohabiting women. Other authors have hypothesized that stable relationships might be protective against DE
(McLean et al., 2010; Tiggemann, 2015), and while other studies have found no association between marital status and DE (Mangweth-Matzek et al., 2014), they have not separated married from cohabiting women, a distinction that might be important in the Mexican context (Doubova Dubova, Pamanes-Gonzalez, Billings, & Torres-Arreola Ldel, 2007). Indicators of higher social position were associated with DE, as described previously in the literature. As for indicators of exposure
I. Bojorquez et al. / Appetite 121 (2018) 207e214
to modernization, being from Northern Mexico was associated with higher scores on the BQREB, and although the association with USorigin was not significant, it could be due to the small sample size in this category. Internet access at home was positively associated with DE in all analyses. While Internet access doesn't automatically imply use, this result is similar to those of other authors reporting that exposure to mass media is associated with DE in middleincome countries (Becker et al., 2011). As both higher social position and exposure to modernization are in theory related to more modern gender-related expectations, these associations support the notion that the sociocultural conditions that accompany modernization could be risk factors for DE. A similar conclusion was reached by Lee and Lee (2000), who also assessed differences in DE within a single upper middle-income country. Those authors compared between young women in three communities in China, varying by degree of socioeconomic development, and found a similar positive association between modernization and DE. As most studies of DE are either not-probability, conducted in special populations, or employ different instruments and case definitions, the results are not easily comparable. In the Mexican National Health and Nutrition Survey 2006, which employed the BQREB, only 1.0% of teenage girls had a score over the cutoff point (Barriguete-Melendez et al., 2009), as compared to 2.0% in our study. Possible explanations for the difference are that the National Health and Nutrition Survey sample included participants from 10 years of age and rural population, in both of which one would expect to find lower DE frequencies. The differences may also be due to sampling variability, since both percentages are relatively low. As for indicators of DE, Machado et al. (2012) reported a 12month prevalence of 26.6% for weight loss attempts among male and female adults in Pelotas, Brazil. In mothers of children at two schools in Mexico City, Lora-Cortez and Saucedo-Molina (2006) found a prevalence of 5% for dieting and 18.3% for weight-related preoccupation, but the difference with our results might be due to specific characteristics of mothers in those schools. Again, as these two studies employed different instruments and reference periods, the comparison is only approximate. The limitations of our study include the cross-sectional design, which does not allow for the evaluation of causality. Moreover, the study of life course warrants a longitudinal treatment, and the cross-sectional comparison between people of different ages is only an approximation to it. Furthermore, the indicators and models employed in this study are a necessary simplification. The complexity of the relationships between life course, social position and sociocultural context should further be explored through quantitative and qualitative studies that undertake an increasingly detailed analysis of these components in diverse populations. Even with these limitations, our study allowed us to explore these relationships in a population-based study with a large sample size, so that the results complement those of more in-depth, qualitative studies. To conclude, life course markers and sociocultural factors were independently associated with DE in this sample, but life course markers also interacted with sociocultural factors and BMI. These interactions point to specific at-risk groups that should be identified and addressed in preventive efforts and public health policies. They also point to the need of future studies that explore the mechanisms linking specific life course markers with DE, in diverse sociocultural contexts.
Acknowledgments Funding for this research was provided by Consejo Nacional de xico (grant no. CB-2010-153536). Ciencia y Tecnología, Me
213
Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.appet.2017.11.085. References Alwan, H., Viswanathan, B., Williams, J., Paccaud, F., & Bovet, P. (2010). Association between weight perception and socioeconomic status among adults in the Seychelles. BMC Public Health, 10, 467. https://doi.org/10.1186/1471-2458-10467. Anderson-Fye, E. P. (2004). A “Coca-Cola” shape: Cultural change, body image, and s, Belize. Culture Medicine and Psychiatry, 28, eating disorders in san andre 561e595. ndez-Barrera, L., Pedroza, A., & RiveraBarquera, S., Campos-Nonato, I., Herna Dommarco, J. (2013). Prevalencia de obesidad en adultos mexicanos, 2000-2012 [Prevalence of obesity in Mexican adults 2000-2012]. Salud Publica de Mexico, 55(supl 2), S151eS160. Barriguete-Melendez, J. A., Unikel-Santoncini, C., Aguilar-Salinas, C., CordobaVillalobos, J. A., Shamah, T., Barquera, S., et al. (2009). Prevalence of abnormal eating behaviors in adolescents in Mexico: Mexican national health and nutrition survey 2006. Salud Publica de Mexico, 51(Suppl 4), S638eS644. Becker, C. B., Diedrichs, P. C., Jankowski, G., & Werchan, C. (2013). I'm not just fat, I'm old: Has the study of body image overlooked “old talk”? Journal of Eating Disorders, 1(6). Becker, A. E., Fay, K. E., Agnew-Blais, J., Khan, A. N., Striegel-Moore, R. H., & Gilman, S. E. (2011). Social network media exposure and adolescent eating pathology in Fiji. British Journal of Psychiatry, 198(1), 43e50. https://doi.org/ 10.1192/bjp.bp.110.078675. Bojorquez-Chapela, I., Unikel, C., Mendoza, M. E., & de Lachica, F. (2014). Another body project: The thin ideal, motherhood, and body dissatisfaction among Mexican women. Journal of Health Psychology, 19(9), 1120e1131. https://doi.org/ 10.1177/1359105313484783. Bojorquez, I., Salgado de Snyder, N., & Casique, I. (2009). International migration of partner, autonomy and depressive symptoms among women from a mexican rural area. International Journal of Social Psychiatry, 55(4), 306e321. https:// doi.org/10.1177/0020764008095117. Bordo, S. (1993). Unbearable weight. Feminism, western culture, and the body. Berkley, CA: University of California Press. del Castillo, G., Peschard-Sverdrup, A., & Fuentes, N. A. (2007). Estudio de puertos de xico-Estados Unidos: An entrada Me alisis de capacidades y recomendaciones para incrementar su eficiencia. Retrieved from http://www.siam.economia.gob. mx/swb/work/models/economia/Resource/406/1/images/Estudio_Completo. pdf . Crossley, N. (2006). Reflexive embodiment in contemporary society. Berkshire, England: Open University Press. Doubova Dubova, S. V., Pamanes-Gonzalez, V., Billings, D. L., & Torres-Arreola Ldel, P. (2007). Partner violence against pregnant women in Mexico City. Revista de Saude Publica, 41(4), 582e590. Fiske, L., Fallon, E. A., Blissmer, B., & Redding, C. A. (2014). Prevalence of body dissatisfaction among United States adults: Review and recommendations for future research. Eating Behaviors, 15(3), 357e365. https://doi.org/10.1016/ j.eatbeh.2014.04.010. Gadalla, T., & Piran, N. (2008). Psychiatric comorbidity in women with disordered eating behavior: A national study. Women & Health, 48(4), 467e484. https:// doi.org/10.1080/03630240802575104. Ganzach, Y. (1997). Misleading interaction and curvilinear terms. Psychological Methods, 2(3), 235e247. Giddens, A. (1991). Modernity and self-identity: Self and society in the late modern age. Stanford, CA: Stanford University Press. rrez, J. P., Rivera-Dommarco, J., Shamah-Levy, T., Villalpando-Herna ndez, S., Gutie n Franco, A., Cuevas-Nasu, I., et al. (2012). Encuesta Nacional de Salud y Nutricio 2012. Resultados Nacionales. Cuernavaca, M exico. Instituto Nacional de Salud Pública. Herpertz-Dahlmann, B., Wille, N., Holling, H., Vloet, T. D., Ravens-Sieberer, U., & BELLA study group. (2008). Disordered eating behaviour and attitudes, associated psychopathology and health-related quality of life: Results of the BELLA study. European Journal of Child and Adolescent Psychiatry, 17(Suppl 1), 82e91. https://doi.org/10.1007/s00787-008-1009-9. International Organization for Migration (IOM). (2015). World migration report 2015: MIgrants and cities, new partnerships to manage mobility. Geneva: IOM. Kally, Z., & Cumella, E. J. (2008). 100 midlife women with eating disorders: A phenomenological analysis of etiology. Journal of General Psychology, 135(4), 359e377. https://doi.org/10.3200/GENP.135.4.359-378. Keel, P. K., Baxter, M. G., Heatherton, T. F., & Joiner, T. E. (2007). A 20-year longitudinal study of body weight, dieting and eating disorder symptoms. Journal of Abnormal Psychology, 116(2), 422e432. Kessler, R. C., Berglund, P. A., Chiu, W. T., Deitz, A. C., Hudson, J. I., Shahly, V., et al. (2013). The prevalence and correlates of binge eating disorder in the world health organization world Mental health surveys. Biological Psychiatry, 73(9), 904e914. https://doi.org/10.1016/j.biopsych.2012.11.020. Kilpela, L. S., Becker, C. B., Wesley, N., & Stewart, T. (2015). Body image in adult Women: Moving beyond the younger years. Advances in Eating Disorders, 3(2),
214
I. Bojorquez et al. / Appetite 121 (2018) 207e214
144e164. https://doi.org/10.1080/21662630.2015.1012728. Kimoto, R., Ronquillo, D., Caamano, M. C., Martinez, G., Schubert, L., Rosado, J. L., et al. (2014). Food, eating and body image in the lives of low socioeconomic status rural Mexican women living in Queretaro State, Mexico. Health & Place, 25, 34e42. https://doi.org/10.1016/j.healthplace.2013.10.004. Kwan, S. (2009). Competing motivational discourses for weight loss: Means to ends and the nexus of beauty and health. Qualitative Health Research, 19(9), 1223e1233. https://doi.org/10.1177/1049732309343952. Lai, B. P., Tang, C. S., & Tse, W. K. (2006). A longitudinal study investigating disordered eating during the transition to motherhood among Chinese women in Hong Kong. International Journal of Eating Disorders, 39(4), 303e311. https:// doi.org/10.1002/eat.20266. Lee, S., & Lee, A. M. (2000). Disordered eating in three communities of China: A comparative study of female high school students in Hong Kong, shenzhen, and rural hunan. International Journal of Eating Disorders, 27(3), 317e327. von Lengerke, T., Mielck, A., & KORA Study Group. (2012). Body weight dissatisfaction by socioeconomic status among obese, preobese and normal weight women and men: Results of the cross-sectional KORA augsburg S4 population survey. BMC Public Health, 12, 342. Lora-Cortez, C. I., & Saucedo-Molina, T. J. (2006). Conductas alimentarias de riesgo e imagen corporal de acuerdo al índice de masa corporal en una muestra de xico. Salud Mental, 29(3), 60e67. mujeres adultas de la ciudad de Me Machado, E. C., Silveira, M. F., & Silveira, V. M. (2012). Prevalence of weight-loss strategies and use of substances for weight-loss among adults: A population study. Cadernos de Saude Publica, 28(8), 1439e1449. Mangweth-Matzek, B., Hoek, H. W., Rupp, C. I., Lackner-Seifert, K., Frey, N., Whitworth, A. B., et al. (2014). Prevalence of eating disorders in middle-aged women. International Journal of Eating Disorders, 47(3), 320e324. https:// doi.org/10.1002/eat.22232. Matthiasdottir, E., Jonsson, S. H., & Kristjansson, A. L. (2012). Body weight dissatisfaction in the icelandic adult population: A normative discontent? European Journal of Public Health, 22(1), 116e121. https://doi.org/10.1093/eurpub/ckq178. McLean, S. A., Paxton, S. J., & Wertheim, E. H. (2010). Factors associated with body dissatisfaction and disordered eating in women in mildlife. International Journal of Eating Disorders, 43(6), 527e536. Mitchison, D., Hay, P., Slewa-Younan, S., & Mond, J. (2014). The changing demographic profile of eating disorder behaviors in the community. BMC Public Health, 14, 943. https://doi.org/10.1186/1471-2458-14-943. Nichter, M. (2001). Fat Talk: What girls and their parents say about dieting. Cambridge, Massachusets: Harvard University Press. de Oliveira, O., & Ariza, M. (2002). Transiciones familiares y trayectorias laborales xico urbano [Family transitions and female work trajectories femeninas en el Me in urban Mexico]. Cadernos Pagu, (17e18), 339e366. n a la adultez de Oliveira, O., & Mora, M. (2008). Desigualdades sociales y transicio xico contempor en el Me aneo [Social inequities and transition to adulthood in n, 14(57), 117e152. contemporary Mexico]. Papeles de Poblacio Palma-Coca, O., Hernandez-Serrato, M. I., Villalobos-Hernandez, A., UnikelSantoncini, C., Olaiz-Fernandez, G., & Bojorquez-Chapela, I. (2011). Association of socioeconomic status, problem behaviors, and disordered eating in Mexican adolescents: Results of the Mexican national health and nutrition survey 2006. Journal of Adolescent Health, 49(4), 400e406. https://doi.org/10.1016/ j.jadohealth.2011.01.019. Paquette, M. C., & Raine, K. (2004). Sociocultural context of women's body image.
Social Science & Medicine, 59(5), 1047e1058. https://doi.org/10.1016/ j.socscimed.2003.12.016. Perez-Gil, R. S., & Romero, J. G. (2010). Imagen corporal en mujeres de tres zonas xico: Percepcio n y deseo [Women's body image in three rural rurales de Me areas in Mexico: perception and desire]. Salud Publica de Mexico, 52(2), 111e118. Rodin, J., Silberstein, L., & Striegel-Moore, R. (1984). Women and weight: A normative discontent. Nebraska Symposium on Motivation, 32, 267e307. rez, A. G., Perez Gil-Romo, S. E., & De la Rosa Cruz, S. A. (2017). Cuerpo y Romero Jua n en una zona rural de Me xico: ¿una preocupacio n so lo de mujeres o percepcio n de hombres? Demetra, 12(1), 21e40. tambie Runfola, C. D., Holle, A. V., Trace, S. E., Brownle, K. A., Hofmeier, S. M., Gagne, D. A., et al. (2013). Body dissatisfaction in women across the Lifespan: Results of the UNC-SELF and Gender and Body Image (GABI) studies. European Eating Disorders Review, 21(1), 52e59. Slevec, J. H., & Tiggemann, M. (2011). Predictors of body dissatisfaction and disordered eating in middle-aged women. Clinical Psychology Review, 31(4), 515e524. https://doi.org/10.1016/j.cpr.2010.12.002. von Soest, T., & Wichstrom, L. (2008). The impact of becoming a mother on eating problems. International Journal of Eating Disorders, 41(3), 215e223. https:// doi.org/10.1002/eat.20493. Solmi, F., Hatch, S. L., Hotopf, M., Treasure, J., & Micali, N. (2014). Prevalence and correlates of disordered eating in a general population sample: The south east london community health (SELCoH) study. Social Psychiatry & Psychiatric Epidemiology, 49, 1335e1346. Stice, E., Marti, C. N., & Durant, S. (2011). Risk factors for onset of eating disorders: Evidence of multiple risk pathways from an 8-year prospective study. Behaviour Research and Therapy, 49(10), 622e627. https://doi.org/10.1016/ j.brat.2011.06.009. Stice, E., Ng, J., & Shaw, H. (2010). Risk factors and prodromal eating pathology. Journal of Child Psychology and Psychiatry, 51(4), 518e525. https://doi.org/ 10.1111/j.1469-7610.2010.02212.x. Tiggemann, M. (2015). Considerations of positive body image across various social identities and special populations. Body Image, 14, 168e176. https://doi.org/ 10.1016/j.bodyim.2015.03.002. Tiggemann, M., & Lynch, J. E. (2001). Body image across the life span in adult women: The role of self-objectification. Developmental Psychology, 37(2), 243e253. Turner, B. (1984). The body and Society: Explorations in social theory. Oxford: Blackwell. ~ o-Garcia, S. (2004). Validation Unikel-Santoncini, C., Bojorquez-Chapela, I., & Carren of a brief questionnaire to measure the risk of abnormal eating behaviors. Salud Publica de Mexico, 46(6), 509e515. n cultural a la Valenzuela, J. M. (2003). Por las fronteras del norte. Una aproximacio frontera M exico-Estados Unidos. Mexico: CONACULTA-FCE. Wade, T. D., Wilksch, S. M., & Lee, C. (2012). A longitudinal investigation of the impact of disordered eating on young women's quality of life. Health Psychology, 31(3), 352e359. https://doi.org/10.1037/a0025956. Warin, M., Turner, K., Moore, V., & Davies, M. (2008). Bodies, mothers and identities: Rethinking obesity and the BMI. Sociology of Health & Illness, 30(1), 97e111. https://doi.org/10.1111/j.1467-9566.2007.01029.x. Williams, L., Germov, J., & Young, A. (2011). The effect of social class on mid-age women's weight control practices and weight gain. Appetite, 56(3), 719e725. https://doi.org/10.1016/j.appet.2011.02.009.
Contents lists available at ScienceDirect
Appetite journal homepage: www.elsevier.com/locate/appet
Life course, sociocultural factors and disordered eating in adult Mexican women nica Valdez c, Claudia Unikel b, * Ietza Bojorquez a, Joyce Bustos b, Vero El Colegio de la Frontera Norte, Mexico, Km. 18.5 Carretera Esc enica Tijuana-Ensenada, San Antonio del Mar, Tijuana CP 22506, BC, Mexico n de la Fuente Mun ~ iz, Mexico, Calz M Instituto Nacional de Psiquiatría Ramo exico-Xochimilco 101, Huipulco, 14370 Ciudad de M exico, Mexico c n, Copilco Universidad, 04510 Ciudad de M noma de M Universidad Nacional Auto exico, Mexico, Avenida Universidad 3004, Coyoaca exico, Mexico a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 7 May 2017 Received in revised form 17 October 2017 Accepted 7 November 2017 Available online 9 November 2017
Disordered eating (DE) can appear in women of all ages and in diverse sociocultural contexts, however most research focuses on younger women in higher income countries. The purpose of this article was to explore the association of life course markers with DE, considering the effects of sociocultural factors, in a sample of adult women in Tijuana, Mexico. We employed data from a household survey (n ¼ 2322) conducted in 2014, to evaluate the associations of DE with age, occupation, marital status and having children (life course markers), and indicators of social position and exposure to modernization (sociocultural factors). The prevalence of weight preoccupation was 69.2% (CI95% 67.3,71.1), the prevalence of dieting 24.8% (CI95% 22.4,27.3), and 2.0% (CI95% 1.4,3.0) had a probable eating disorder according to the questionnaire cutoff score. In the adjusted model, younger age, being employed, higher social position and indicators of exposure to modernization had positive associations with DE. There were interactions between marital status and body mass index, and between age and region of birth. The interaction terms showed that overweight was positively associated with DE among single and cohabiting participants, but not among the married ones; and that the negative association between DE and age was apparent from younger age groups in women born in less developed regions of Mexico. Our results replicate others in showing DE to be present in women through the life course, and point to at-risk groups in the confluences of life course, social position and modernization. © 2017 Elsevier Ltd. All rights reserved.
Keywords: Disordered eating Adult women Life course Social factors
1. Introduction The term “disordered eating” (DE) refers to attitudes and behaviors that are similar to those that appear in eating disorders, without meeting the clinical criteria for the latter. Examples include weight-related preoccupation, dieting, use of medical and nonmedical substances such as laxatives or diuretics, and selfprovoked vomiting. Whether DE is a subclinical form, a prodrome or a risk factor for eating disorders is still unclear (Stice, Ng, & Shaw, 2010), but it can be harmful to health in itself, and people with DE report similar levels of affective disorders, anxiety and suicidality as those with well-established eating disorder diagnoses (Solmi, Hatch, Hotopf, Treasure, & Micali, 2014). DE is also associated
n de Investigaciones Epidemiolo gicas y Psico* Corresponding author. Direccio n de la Fuente Mun ~ iz, Calzada sociales, Instituto Nacional de Psiquiatría Ramo Mexico-Xochimilco 101, Tlalpan, D.F., CP 14370, Mexico. E-mail addresses: [email protected] (I. Bojorquez), [email protected]. mx (C. Unikel). https://doi.org/10.1016/j.appet.2017.11.085 0195-6663/© 2017 Elsevier Ltd. All rights reserved.
with poorer quality of life (Herpertz-Dahlmann et al., 2008; Wade, Wilksch, & Lee, 2012). Risk factors for DE are manifold, including genetics, personality, cognitive style and adverse life events (Kally & Cumella, 2008; Stice, Marti, & Durant, 2011; Stice et al., 2010). At the same time, the increase in the prevalence of DE in the 20th century evidences the importance of sociocultural aspects, and particularly of the social construction of femininity, as DE appears to have increased in parallel with changes in female gender roles (Bordo, 1993; Nichter, 2001; Paquette & Raine, 2004; Rodin, Silberstein, & Striegel-Moore, 1984). These changes are part of a general process of social and cultural modernization, in which the production of a successful identity that includes a certain type of body becomes more and more a central personal project (Giddens, 1991; Turner, 1984). In combination with the thin ideal (the cultural norm that considers slimness as essential for female beauty), the importance attributed to personal cultivation via practices that modify the body (Crossley, 2006) provides the sociocultural context in which DE arises. In more recent times, the public health discourse against obesity has
208
I. Bojorquez et al. / Appetite 121 (2018) 207e214
added to the discourses that promote the thin ideal (Kwan, 2009), increasing the risk of DE. Given the strong association of DE with gender-related social expectations, and the changes in these expectations as women pass through the life course, the manifestations and factors associated with DE in women of different age groups are topics worth of study. While most research on DE focuses on adolescents or young women, an increasing number of studies show that DE behaviors and attitudes are present well into adulthood (Fiske, Fallon, Blissmer, & Redding, 2014; Kilpela, Becker, Wesley, & Stewart, 2015; Slevec & Tiggemann, 2011). In this regard, most research shows a negative association between age and DE (Gadalla & Piran, 2008; Solmi et al., 2014), or an inverted U-shaped relationship that peaks between the ages of 30 and 40 (Matthiasdottir, Jonsson, & Kristjansson, 2012; Mitchison, Hay, Slewa-Younan, & Mond, 2014; Runfola et al., 2013; Tiggemann & Lynch, 2001). However, body shape concerns can also increase as the aging body deviates from the cultural ideal of slimness and youth (Becker, Diedrichs, Jankowski, & Werchan, 2013; Slevec & Tiggemann, 2011) and preoccupation over the body's functioning might also increase (McLean, Paxton, & Wertheim, 2010). At the same time, some studies have found that age decreases the importance attributed to appearance, which might result in less body dissatisfaction and DE (Kilpela et al., 2015). Most of this research has been conducted in upper-income countries (mainly English-speaking), so there is a lack of information about the association of age and DE in other societies. As for the effect of other life course markers, some studies have reported that marriage and motherhood are associated with decrease in DE (Keel, Baxter, Heatherton, & Joiner, 2007), while others find that pregnancy and the postpartum period increase it (Lai, Tang, & Tse, 2006). Some authors suggest that achievement of life course markers such as establishing a career, having children and having a committed long-term relationship might decrease the perceived importance of body shape (Tiggemann, 2015), and qualitative studies have shown that the significance that women attach to their relational identities as mothers can be a protective factor as it helps them to oppose the discourse of self-enhancement through weight control (Bojorquez-Chapela, Unikel, Mendoza, & de Lachica, 2014; Warin, Turner, Moore, & Davies, 2008). While there can be similarities between the experiences of women in different social groups, the life course also interacts with other aspects of the social position to produce experiences of the body, of the importance of weight control and of DE, that are shared with persons in similar positions. Thus, quantitative studies of the association of social position and DE in adults in different regions of the world show that the upper classes are more likely to engage in weight control practices and to be dissatisfied with their appearance even while at a normal weight (Alwan, Viswanathan, Williams, Paccaud, & Bovet, 2010; Machado, Silveira, & Silveira, 2012; Mangweth-Matzek et al., 2014; von Lengerke, Mielck & KORA Study Group, 2012; Williams, Germov, & Young, 2011). The qualitative studies mentioned above also show that upper-class mothers are more conscious of the social expectation of thinness (Bojorquez-Chapela et al., 2014; Warin et al., 2008). On the other hand, while differences by social class in the prevalence of weight-related preoccupation and weight control practices are clear, comparisons between countries show a more complicated picture. If DE is related to modernization, a gradient from lower to higher income countries should be apparent, inasmuch as economic development is associated with other modernizing social processes. However, the World Health Organization worldwide surveys found no clear gradient in the case of
eating disorders (Kessler et al., 2013), and the use of different instruments and special-population samples limits comparisons between countries of the prevalence of DE. The association of exposure to social modernization with DE is still an interesting subject of study, and countries with ample differences in socioeconomic and human development, and varying local and regional cultures, are sites where these associations can be explored (Lee & Lee, 2000). To summarize, the life course, social position, and exposure to modernization, all play a role in DE, but their independent effects and possible interactions are still unclear. Most research in this area has been conducted in high-income countries of Europe and North America, with adolescents or young women, or with special populations such as students. In this article, we assess the association of life course markers with DE in adult women, at the same time exploring the independent effects of indicators of social position and of exposure to modernization, and their interactions with the life course. Following from the literature referenced above, we hypothesized that higher social position and indicators of exposure to modernization would be positively associated with DE. As most literature reports a negative association between life course markers and DE, our second set of hypotheses was that older age, being employed, being married and having children would be associated negatively with DE. Finally, we expected that social position and indicators of exposure to modernization would modify the association between life course markers and DE. 2. Methods 2.1. Site and sample In FebruaryeApril 2014, we conducted a cross-sectional survey on a representative sample of the population of women ages 18 to 65 living in Tijuana, Mexico. Tijuana is a city with over 1.5 million inhabitants on the Mexico-United States (US) border, it has a high rate of internal migration, and is also a transit city for migrants heading to the US(International Organization for Migration (IOM) (2015)). The city has intense dynamics of cross-border relations, with over 40 million annual border crossings (del Castillo, Peschard-Sverdrup, & Fuentes, 2007), and can in many respects be regarded as an intermediate place between Mexican and US culture (Valenzuela, 2003). The survey had a probabilistic, stratified, multi-stage cluster sample design. In the first stage, Basic Statistical Geographical Areas as defined by Mexico's National Institute of Statistics and Geography (INEGI) were divided into strata according to a Marginalization Index reported by Mexico's National Population Council (CONAPO). The Marginalization Index classifies geographical areas by level of social development, and is computed from indicators such as literacy rate, percentage of dwellings with electricity and percentage of population with low income. In the second stage, blocks were randomly selected from each area. All households on the block were visited, and one woman of the required age was randomly selected from each household and invited to participate in the survey. A required sample size of 2500 respondents was predefined. The eligibility criteria were: 1) being 18e65 years old, and 2) agreeing to take part in the survey after a process of informed consent. Potential participants were told of the purpose of the project and the nature of participation, and asked to sign a letter of consent. All procedures were approved by the Ethics Committee of El Colegio de la Frontera Norte. Questionnaires were administered face to face by previously trained interviewers, who also weighed and measured participants using Tanita BF-683W
I. Bojorquez et al. / Appetite 121 (2018) 207e214
scales, and portable SECA 213 stadiometers.
2.2. Variables We measured DE with the Brief Questionnaire of Risky Eating Behaviors (BQREB) (Unikel-Santoncini, Bojorquez-Chapela, & ~ o-Garcia, 2004). The BQREB includes 10 items in a Likert Carren scale, asking about the frequency of restrictive and purging behaviors, binge eating and weight-related concerns over the past three months (see items in Table 2). Response options range from “never or almost never” to “very often (three or more times per week)” and are rated from 0 to 3. Possible scores on the questionnaire thus range from 0 to 30 where a higher score indicates greater frequency. The BQREB was developed and validated in Mexican population and has been used in the National Health rrez Surveys in Mexico (Barriguete-Melendez et al., 2009; Gutie et al., 2012). In a validation study comparing scores in a sample of students and a clinical sample, internal consistency as measured by Cronbach's a was 0.83, and a cutoff >10 had a sensitivity of 0.81 and specificity of 0.78 for the identification of cases of eating disorders (Unikel-Santoncini et al., 2004). While there is no one single accepted definition of DE in the literature, in this article we employed a score >10 in the BQREB as an indicator of a DE level that could be of clinical significance. As indicators of the life course we employed age and life course markers that are important milestones in the passage to adulthood in Mexico (de Oliveira & Ariza, 2002; de Oliveira & Mora, 2008): having children, marital status, and occupation. In exploratory bivariate analysis we observed an inverted-U association of DE with age (continuous variable), but it disappeared after adjustment by other predictors in the model and subsequently we modeled age as a categorical variable (18e29, 30e40 and 41e65 years) for easiness of presentation. As indicators of social position we employed education (in categories by years of education) and an index of socioeconomic level (SEL) computed through principal components analysis of questions on the presence of goods and services in the household. As indicators of exposure to modernization we included region (within Mexico) or country of origin, Internet access at home, and frequency of crossing to the US. We categorized the region of origin following a classification that separates Mexican states in four groups according to their geographic, economic and social characteristics (Palma-Coca et al., 2011), and after exploratory analysis combined these into two groups: the northern (more developed) region, and the central-southern (less developed) region. Out of 36 participants born in other countries, 34 had been born in the US. Therefore, for analysis, we excluded the two participants born elsewhere and considered three categories: northern Mexico, southern Mexico, and US. Since body mass index (BMI) (weight in kilos/height in meters2) is associated with both age and DE, in the final analysis we controlled for this variable as calculated from anthropometric measurements. Values outside the range 10e58 (five cases) were ndez-Bareliminated as invalid (Barquera, Campos-Nonato, Herna rera, Pedroza, & Rivera-Dommarco, 2013), and BMI was recoded into a categorical variable, according to the cutoff points of the World Health Organization (low <18.5, normal 18.5e24.9, overweight 25e29, obesity 30, frequencies shown in Table 1). For multivariate analysis, because of small counts in some combinations of age and BMI, we merged low and normal into one category. The results of analyses separating the low and normal categories were similar to the ones reported, and there were no cases of DE in the low category. Pregnant women (121 cases) were excluded from analyses that considered BMI.
209
2.3. Analysis First, we explored the distribution of the variables of interest, and of their joint distributions. As the BQREB score is a discrete numerical variable, and the over-dispersion parameter (alpha) was statistically significant, we employed a negative binomial regression model for the association analysis. Following a theoretical criterion, variables that were part of our conceptual framework were included in the model regardless of their p-value. We explored nonlinear associations and interactions (Ganzach, 1997), testing for interactions between each of the life course variables (age group, having children, marital status and occupation), and the covariates. Interactions with p < 0.05 were kept in the model. All analyses took into account the sample design, using Taylor linearized series to adjust the standard errors, and were carried out using the “svy” module in Stata, version 13.1. To assess the model fit, we considered the likelihood ratio test comparing it to a Poisson model, comparison of observed vs. predicted counts, and the Akaike Information Criteria (AIC) in unweighted models. 3. Results A total of 2500 women were selected to participate in the study, of whom 2345 (93.8%) answered the questionnaire. Of these, 2322 (99.0%) provided complete answers in the BQREB. The characteristics of the respondents who answered the BQREB are given in Table 1. Most participants had children, were either married or cohabiting, and half were homemakers or stayed at home for other reasons. Over half of participants had less than 10 years of education, and only 13.0% had some college education or more. The percentage born in central-southern Mexico was 16.6%, and 1.6% had been born in the US. Slightly over half of the sample had Internet access at home. About two thirds were either overweight or obese. The 23 respondents who did not give complete answers to the BQREB were similar in respect to observed characteristics (not shown in Table). In Table 1, we also show the percentage scoring above the BQREB cutoff point (i.e. probable clinically significant cases), by covariate levels. Among all participants with complete responses in the BQREB, the percentage scoring above the cutoff point was 2.0% (CI95% 1.4,3.0), and the mean score in the BQREB was 2.8 (CI95% 2.6,3.0). The most frequently reported item of the BQREB was concern about putting on weight, but dieting, exercising and feeling that one had ate too much were also relatively frequent (Table 2). In bivariate analysis (Table 3, second and third columns), DE was negatively associated with age, being separated or widowed, and being from central-southern Mexico; and positively associated with being employed, years of education, SEL, Internet access, border crossing and BMI. In the main effects multivariate model (Table 3, fourth and fifth columns), the negative associations between DE and age and being from central-southern Mexico were still significant, as were the positive associations between DE and being employed, having more than 12 years of education, being in the upper quintile of SEL, having Internet access at home, and higher BMI. In the model including interactions (Table 3, sixth and seventh columns) the effects of employment, education, SEL and Internet access remained significant. The coefficients for interactions between age group and region of origin showed that among participants born in northern Mexico, only those in the 41e65 age group had a significantly lower score as compared with the younger group, while for participants from central-southern Mexico the score was significantly lower also for the 30e40 age group. As the plot of predicted scores in Fig. 1 shows, this means that the negative slope of the association between age and DE started earlier for
210
I. Bojorquez et al. / Appetite 121 (2018) 207e214
Table 1 Characteristics of participants, and percentage with score over questionnaire cutoff.a Variable
Age group 18e29 years 30e40 years 41e65 years Has no children Has children Marital status Single Married Cohabiting Separated/widowed Occupation Home/not employed Student Employed Years of education 0-9 10-12 12þ Socioeconomic level (quintile) 1 (lower) 2 3 4 5 (upper) Born in Northern Mexico Central-southern Mexico United States No Internet access at home Internet access at home No international border crossing International border crossing Body Mass Index:c Low Normal Overweight Obese a b c
Characteristics of participants
Percentage over cutoff, by characteristic
Numberb
Percentageb
Confidence Interval 95%b
Numberb
Percentageb
Confidence Interval 95%b
903 557 862 475 1847
38.8 24.1 37.1 20.3 79.8
38.0,39.6 22.5,25.8 35.2,39.0 18.0,22.7 77.3,82.0
22 18 8 5 43
2.5 3.0 0.9 1.2 2.3
1.5,4.1 1.8,4.8 0.3,3.0 0.4,3.1 1.6,3.2
522 1125 424 251
22.6 49.1 17.5 10.8
21.1,24.2 46.3,51.9 15.7,19.4 9.6,12.2
9 22 14 3
2.3 1.8 3.0 1.0
1.0,5.1 1.2,2.7 1.4,6.3 0.3,3.1
1156 163 1003
50.0 6.9 43.1
49.2,50.8 5.9, 8.0 41.7,44.6
18 2 28
1.6 2.0 2.6
1.0,2.5 0.4,9.0 1.6,4.0
1300 653 271
57.3 29.7 13.0
52.4,62.0 27.1,32.4 9.9,16.9
22 19 4
1.7 2.7 1.6
0.9,3.3 1.7,4.3 0.5,4.7
362 463 542 302 615
14.8 19.4 24.4 13.6 27.9
12.6,17.3 17.0,22.0 21.3,27.7 12.0,15.4 23.1,33.2
7 10 6 8 16
1.7 2.6 1.0 2.0 2.8
0.8,3.6 1.1,5.9 0.4,2.6 0.8,5.0 1.7,4.5
1861 406 32 1154 1164 1466 810
81.8 16.6 1.6 48.2 51.8 62.2 37.8
79.7,83.7 14.8,18.5 1.1, 2.4 43.9,52.5 47.5,56.1 57.5,66.7 33.3,42.5
38 6 3 19 29 32 16
2.0 1.1 12.5 1.6 2.4 2.2 1.9
1.2,3.3 0.5,2.7 4.4,30.5 0.7,3.4 1.7,3.5 1.2,3.9 1.1,3.2
18 506 700 590
1.1 28.4 38.5 32.0
0.7, 1.9 25.8,31.1 35.8,41.2 29.6,34.6
0 8 16 17
0 2.3 2.0 2.6
1.1,4.6 1.1,3.4 1.3,5.5
Participants with complete answers in the Brief Questionnaire of Risky Eating Behaviors (Unikel-Santoncini et al., 2004). Unweighted n ¼ 2322. Number is unweighted n, percentages and means are weighted, confidence intervals consider sampling design. Only for participants who were not pregnant at the time of interview.
Table 2 Prevalence of disordered eating attitudes and behaviors.a In the past three months, how often have you … ?
1/week or more Number
Worried about gaining weight Eaten too much Lost control over eating Vomited to lose weight Fasted to lose weight Dieted to lose weight Exercised to lose weight Taken pills to lose weight Taken diuretics to lose weight Taken laxatives to lose weight a b
1588 617 349 32 140 556 726 193 71 112
b
Percentage
3/week or more b
69.2 27.4 15.2 1.4 6.5 24.8 32.9 8.5 3.4 5.0
b
Confidence Interval 95%
Numberb
Percentageb
Confidence Interval 95%b
67.3,71.1 25.4,30.0 13.4,17.3 1.0,2.1 5.1,8.1 22.4,27.3 29.9,36.1 7.4,9.9 2.6,4.4 4.0,6.3
375 15 9 3 6 51 110 33 12 11
16.5 0.7 0.4 0.1 0.2 2.0 5.1 1.3 0.5 0.6
14.7,18.5 0.4,1.1 0.2,0.7 0.0,0.2 0.1,0.6 1.5,2.7 4.0,6.6 0.8,1.9 0.2,0.9 0.3,1.2
Unweighted n ¼ 2322. Number is unweighted n, percentages and means are weighted, confidence intervals consider sampling design.
those born in the less developed areas of Mexico. The interaction between marital status and BMI showed that for single and cohabiting participants there was a significant positive effect of being in either the overweight or obese categories, while for married participants the effect was only significant for those in the obese category. Besides, the difference in mean scores between the low/normal category and the obese category was more marked for
cohabiting participants (Fig. 2). 4. Discussion As others before (Gadalla & Piran, 2008; Solmi et al., 2014), in this study we found that scores in the BEQRB were lower in the older groups. We also found the previously reported inverted U-
I. Bojorquez et al. / Appetite 121 (2018) 207e214
211
Table 3 Association between life course and other variables, and disordered eating.a Variable
Main effects modelb
Bivariate models Regression Coefficient
Age group 18-29 Ref. 30-40 0.05 41-65 0.33 Occupation Home/not employed Ref. Student 0.17 Employed 0.19 Marital status Single Ref. Married 0.05 Cohabiting 0.13 Separated/Widowed 0.15 Has children 0.00 Years of education 0-9 Ref. 10-12 0.19 12þ 0.34 Socioeconomic level (quintile) 1 (lower) Ref. 2 0.15 3 0.07 4 0.12 5 (upper) 0.36 Born in Northern Mexico Ref. Central-southern Mexico 0.13 United States 0.37 Internet access at home 0.24 Border crossing 0.14 Body Mass Index: Low/Normal Ref. Overweight 0.12 Obese 0.26 Age group * born in 30e40yrs * CS Mexicoc Marital status * BMI Single*Overweight Cohabiting*Overweight Cohabiting*Obese
Model with interaction termsb
Confidence Interval 95%
Regression Coefficient
Confidence Interval 95%
Regression Coefficient
Confidence Interval 95%
0.06, 0.16 0.43,-0.23
Ref. 0.07 0.46
0.22, 0.08 0.63,-0.30
Ref. 0.01 0.41
0.19, 0.16 0.58,-0.23
0.02, 0.36 0.11, 0.28
Ref. 0.23 0.23
0.11, 0.56 0.11, 0.36
Ref. 0.21 0.21
0.11, 0.53 0.09, 0.32
Ref. 0.07, 0.16 0.02, 0.29 0.30,-0.00 0.08, 0.09
Ref. 0.03 0.18 0.01 0.14
0.19, 0.01, 0.21, 0.02,
Ref. 0.23 0.24 0.28 0.11
0.48, 0.55, 0.72, 0.05,
0.07, 0.32 0.19, 0.48
Ref. 0.11 0.21
0.03, 0.25 0.02, 0.40
Ref. 0.12 0.20
0.02, 0.25 0.01, 0.38
0.02, 0.32 0.08, 0.22 0.02, 0.25 0.19, 0.54
Ref. 0.14 0.07 0.05 0.18
0.04, 0.29 0.09, 0.22 0.09, 0.19 0.01, 0.35
Ref. 0.15 0.09 0.07 0.22
0.02, 0.31 0.07, 0.24 0.07, 0.22 0.04, 0.40
0.24,-0.02 0.00, 0.74 0.12, 0.36 0.02, 0.26
Ref. 0.17 0.15 0.23 0.06
0.30,-0.03 0.35, 0.67 0.12, 0.34 0.11, 0.23
Ref. 0.04 0.29 0.21 0.06
0.15, 0.23 0.32, 0.89 0.10, 0.32 0.10, 0.22
0.02, 0.27 0.12, 0.39
Ref. 0.19 0.51
0.02, 0.35 0.36, 0.67
Ref. 0.02 0.31
0.26, 0.22 0.08, 0.54
0.33
0.63,-0.02
0.38 0.43 0.66
0.08, 0.68 0.02, 0.84 0.20, 1.13
0.14 0.38 0.22 0.30
0.01 0.07 0.15 0.27
a
Negative binomial regression, dependent variable: score in the BQREB. Excluding participants who were pregnant at the time of interview. Adjusted for the variables and interactions shown in Table. Unweighted n ¼ 1718. Main effects model, F(20,28) ¼ 13.65, p < .0001. Interaction model F(30,18) ¼ 9.74, p < .0001. c Only interaction categories with p < .05 are shown in Table. b
shape association (Matthiasdottir et al., 2012; Mitchison et al., 2014; Runfola et al., 2013; Tiggemann & Lynch, 2001), but the non-linearity disappeared after adjusting by other predictors in the model, so that it was possibly the result of other characteristics. The association between age and DE was apparent from a younger (30e40 years) age group among participants born in centralsouthern Mexico. As mentioned above, in Mexico there is a north-south gradient of social and economic development, and also of closeness to US cultural elements. It is possible that more traditional gender role expectations protect southern adult women from DE, and although qualitative studies in central and southern rural areas (Kimoto et al., 2014; Perez-Gil & Romero, 2010; Romero rez, Perez Gil-Romo, & De la Rosa Cruz, 2017) show that weight Jua concerns and weight-control practices are widespread, the prevalence of DE is still higher in the northern regions (Palma-Coca et al., 2011). Similarly to Mangweth-Matzek et al. (2014), we observed no significant association between having children and DE. In preliminary analysis we explored associations of DE with number and age of children, and all of them were likewise non-significant. This finding was contrary to our hypothesis, and also to previous
qualitative studies reporting that for Mexican women the role of mother seems to decrease the interest in achieving or conserving a thin figure, and weight gain is naturalized as an inevitable consequence of childbearing (Bojorquez-Chapela et al., 2014). The finding by other authors that having children decreases DE (von Soest & Wichstrom, 2008) might be related to specific childbearing experiences, or be associated with other elements of the life course that confound the association with motherhood. In our study, being employed had a positive association with DE. Qualitative studies have shown that working, especially in professions that involve direct contact with clients, can make women more conscious of their appearance, and a thin shape can be a requisite for some jobs (Anderson-Fye, 2004). A qualitative study of women in Mexico showed that they were aware of the need for a slim figure in the workplace, and described how in that environment skipping meals, worrying about being fat, talking about diets and going on diet with colleagues was relatively common (Bojorquez-Chapela et al., 2014). Another possible explanation has to do with the social construction of female work in the Mexican society. Other studies have shown that in Mexico workingwomen show more depressive symptoms than women who stay at home
212
I. Bojorquez et al. / Appetite 121 (2018) 207e214
Fig. 1. Predicted score in disordered eating, by age group and region of origin.
Fig. 2. Predicted score in disordered eating, by category of body mass index and marital status.
(Bojorquez, Salgado de Snyder, & Casique, 2009), and while in other contexts work can have a positive meaning as the establishment of a career (Tiggemann, 2015), Mexican women could perceive it as conflicting with their expected gender role. The fourth life course indicator we employed, marital status, showed a complex association with DE, in that BMI had a greater effect for single and cohabiting women. Other authors have hypothesized that stable relationships might be protective against DE
(McLean et al., 2010; Tiggemann, 2015), and while other studies have found no association between marital status and DE (Mangweth-Matzek et al., 2014), they have not separated married from cohabiting women, a distinction that might be important in the Mexican context (Doubova Dubova, Pamanes-Gonzalez, Billings, & Torres-Arreola Ldel, 2007). Indicators of higher social position were associated with DE, as described previously in the literature. As for indicators of exposure
I. Bojorquez et al. / Appetite 121 (2018) 207e214
to modernization, being from Northern Mexico was associated with higher scores on the BQREB, and although the association with USorigin was not significant, it could be due to the small sample size in this category. Internet access at home was positively associated with DE in all analyses. While Internet access doesn't automatically imply use, this result is similar to those of other authors reporting that exposure to mass media is associated with DE in middleincome countries (Becker et al., 2011). As both higher social position and exposure to modernization are in theory related to more modern gender-related expectations, these associations support the notion that the sociocultural conditions that accompany modernization could be risk factors for DE. A similar conclusion was reached by Lee and Lee (2000), who also assessed differences in DE within a single upper middle-income country. Those authors compared between young women in three communities in China, varying by degree of socioeconomic development, and found a similar positive association between modernization and DE. As most studies of DE are either not-probability, conducted in special populations, or employ different instruments and case definitions, the results are not easily comparable. In the Mexican National Health and Nutrition Survey 2006, which employed the BQREB, only 1.0% of teenage girls had a score over the cutoff point (Barriguete-Melendez et al., 2009), as compared to 2.0% in our study. Possible explanations for the difference are that the National Health and Nutrition Survey sample included participants from 10 years of age and rural population, in both of which one would expect to find lower DE frequencies. The differences may also be due to sampling variability, since both percentages are relatively low. As for indicators of DE, Machado et al. (2012) reported a 12month prevalence of 26.6% for weight loss attempts among male and female adults in Pelotas, Brazil. In mothers of children at two schools in Mexico City, Lora-Cortez and Saucedo-Molina (2006) found a prevalence of 5% for dieting and 18.3% for weight-related preoccupation, but the difference with our results might be due to specific characteristics of mothers in those schools. Again, as these two studies employed different instruments and reference periods, the comparison is only approximate. The limitations of our study include the cross-sectional design, which does not allow for the evaluation of causality. Moreover, the study of life course warrants a longitudinal treatment, and the cross-sectional comparison between people of different ages is only an approximation to it. Furthermore, the indicators and models employed in this study are a necessary simplification. The complexity of the relationships between life course, social position and sociocultural context should further be explored through quantitative and qualitative studies that undertake an increasingly detailed analysis of these components in diverse populations. Even with these limitations, our study allowed us to explore these relationships in a population-based study with a large sample size, so that the results complement those of more in-depth, qualitative studies. To conclude, life course markers and sociocultural factors were independently associated with DE in this sample, but life course markers also interacted with sociocultural factors and BMI. These interactions point to specific at-risk groups that should be identified and addressed in preventive efforts and public health policies. They also point to the need of future studies that explore the mechanisms linking specific life course markers with DE, in diverse sociocultural contexts.
Acknowledgments Funding for this research was provided by Consejo Nacional de xico (grant no. CB-2010-153536). Ciencia y Tecnología, Me
213
Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.appet.2017.11.085. References Alwan, H., Viswanathan, B., Williams, J., Paccaud, F., & Bovet, P. (2010). Association between weight perception and socioeconomic status among adults in the Seychelles. BMC Public Health, 10, 467. https://doi.org/10.1186/1471-2458-10467. Anderson-Fye, E. P. (2004). A “Coca-Cola” shape: Cultural change, body image, and s, Belize. Culture Medicine and Psychiatry, 28, eating disorders in san andre 561e595. ndez-Barrera, L., Pedroza, A., & RiveraBarquera, S., Campos-Nonato, I., Herna Dommarco, J. (2013). Prevalencia de obesidad en adultos mexicanos, 2000-2012 [Prevalence of obesity in Mexican adults 2000-2012]. Salud Publica de Mexico, 55(supl 2), S151eS160. Barriguete-Melendez, J. A., Unikel-Santoncini, C., Aguilar-Salinas, C., CordobaVillalobos, J. A., Shamah, T., Barquera, S., et al. (2009). Prevalence of abnormal eating behaviors in adolescents in Mexico: Mexican national health and nutrition survey 2006. Salud Publica de Mexico, 51(Suppl 4), S638eS644. Becker, C. B., Diedrichs, P. C., Jankowski, G., & Werchan, C. (2013). I'm not just fat, I'm old: Has the study of body image overlooked “old talk”? Journal of Eating Disorders, 1(6). Becker, A. E., Fay, K. E., Agnew-Blais, J., Khan, A. N., Striegel-Moore, R. H., & Gilman, S. E. (2011). Social network media exposure and adolescent eating pathology in Fiji. British Journal of Psychiatry, 198(1), 43e50. https://doi.org/ 10.1192/bjp.bp.110.078675. Bojorquez-Chapela, I., Unikel, C., Mendoza, M. E., & de Lachica, F. (2014). Another body project: The thin ideal, motherhood, and body dissatisfaction among Mexican women. Journal of Health Psychology, 19(9), 1120e1131. https://doi.org/ 10.1177/1359105313484783. Bojorquez, I., Salgado de Snyder, N., & Casique, I. (2009). International migration of partner, autonomy and depressive symptoms among women from a mexican rural area. International Journal of Social Psychiatry, 55(4), 306e321. https:// doi.org/10.1177/0020764008095117. Bordo, S. (1993). Unbearable weight. Feminism, western culture, and the body. Berkley, CA: University of California Press. del Castillo, G., Peschard-Sverdrup, A., & Fuentes, N. A. (2007). Estudio de puertos de xico-Estados Unidos: An entrada Me alisis de capacidades y recomendaciones para incrementar su eficiencia. Retrieved from http://www.siam.economia.gob. mx/swb/work/models/economia/Resource/406/1/images/Estudio_Completo. pdf . Crossley, N. (2006). Reflexive embodiment in contemporary society. Berkshire, England: Open University Press. Doubova Dubova, S. V., Pamanes-Gonzalez, V., Billings, D. L., & Torres-Arreola Ldel, P. (2007). Partner violence against pregnant women in Mexico City. Revista de Saude Publica, 41(4), 582e590. Fiske, L., Fallon, E. A., Blissmer, B., & Redding, C. A. (2014). Prevalence of body dissatisfaction among United States adults: Review and recommendations for future research. Eating Behaviors, 15(3), 357e365. https://doi.org/10.1016/ j.eatbeh.2014.04.010. Gadalla, T., & Piran, N. (2008). Psychiatric comorbidity in women with disordered eating behavior: A national study. Women & Health, 48(4), 467e484. https:// doi.org/10.1080/03630240802575104. Ganzach, Y. (1997). Misleading interaction and curvilinear terms. Psychological Methods, 2(3), 235e247. Giddens, A. (1991). Modernity and self-identity: Self and society in the late modern age. Stanford, CA: Stanford University Press. rrez, J. P., Rivera-Dommarco, J., Shamah-Levy, T., Villalpando-Herna ndez, S., Gutie n Franco, A., Cuevas-Nasu, I., et al. (2012). Encuesta Nacional de Salud y Nutricio 2012. Resultados Nacionales. Cuernavaca, M exico. Instituto Nacional de Salud Pública. Herpertz-Dahlmann, B., Wille, N., Holling, H., Vloet, T. D., Ravens-Sieberer, U., & BELLA study group. (2008). Disordered eating behaviour and attitudes, associated psychopathology and health-related quality of life: Results of the BELLA study. European Journal of Child and Adolescent Psychiatry, 17(Suppl 1), 82e91. https://doi.org/10.1007/s00787-008-1009-9. International Organization for Migration (IOM). (2015). World migration report 2015: MIgrants and cities, new partnerships to manage mobility. Geneva: IOM. Kally, Z., & Cumella, E. J. (2008). 100 midlife women with eating disorders: A phenomenological analysis of etiology. Journal of General Psychology, 135(4), 359e377. https://doi.org/10.3200/GENP.135.4.359-378. Keel, P. K., Baxter, M. G., Heatherton, T. F., & Joiner, T. E. (2007). A 20-year longitudinal study of body weight, dieting and eating disorder symptoms. Journal of Abnormal Psychology, 116(2), 422e432. Kessler, R. C., Berglund, P. A., Chiu, W. T., Deitz, A. C., Hudson, J. I., Shahly, V., et al. (2013). The prevalence and correlates of binge eating disorder in the world health organization world Mental health surveys. Biological Psychiatry, 73(9), 904e914. https://doi.org/10.1016/j.biopsych.2012.11.020. Kilpela, L. S., Becker, C. B., Wesley, N., & Stewart, T. (2015). Body image in adult Women: Moving beyond the younger years. Advances in Eating Disorders, 3(2),
214
I. Bojorquez et al. / Appetite 121 (2018) 207e214
144e164. https://doi.org/10.1080/21662630.2015.1012728. Kimoto, R., Ronquillo, D., Caamano, M. C., Martinez, G., Schubert, L., Rosado, J. L., et al. (2014). Food, eating and body image in the lives of low socioeconomic status rural Mexican women living in Queretaro State, Mexico. Health & Place, 25, 34e42. https://doi.org/10.1016/j.healthplace.2013.10.004. Kwan, S. (2009). Competing motivational discourses for weight loss: Means to ends and the nexus of beauty and health. Qualitative Health Research, 19(9), 1223e1233. https://doi.org/10.1177/1049732309343952. Lai, B. P., Tang, C. S., & Tse, W. K. (2006). A longitudinal study investigating disordered eating during the transition to motherhood among Chinese women in Hong Kong. International Journal of Eating Disorders, 39(4), 303e311. https:// doi.org/10.1002/eat.20266. Lee, S., & Lee, A. M. (2000). Disordered eating in three communities of China: A comparative study of female high school students in Hong Kong, shenzhen, and rural hunan. International Journal of Eating Disorders, 27(3), 317e327. von Lengerke, T., Mielck, A., & KORA Study Group. (2012). Body weight dissatisfaction by socioeconomic status among obese, preobese and normal weight women and men: Results of the cross-sectional KORA augsburg S4 population survey. BMC Public Health, 12, 342. Lora-Cortez, C. I., & Saucedo-Molina, T. J. (2006). Conductas alimentarias de riesgo e imagen corporal de acuerdo al índice de masa corporal en una muestra de xico. Salud Mental, 29(3), 60e67. mujeres adultas de la ciudad de Me Machado, E. C., Silveira, M. F., & Silveira, V. M. (2012). Prevalence of weight-loss strategies and use of substances for weight-loss among adults: A population study. Cadernos de Saude Publica, 28(8), 1439e1449. Mangweth-Matzek, B., Hoek, H. W., Rupp, C. I., Lackner-Seifert, K., Frey, N., Whitworth, A. B., et al. (2014). Prevalence of eating disorders in middle-aged women. International Journal of Eating Disorders, 47(3), 320e324. https:// doi.org/10.1002/eat.22232. Matthiasdottir, E., Jonsson, S. H., & Kristjansson, A. L. (2012). Body weight dissatisfaction in the icelandic adult population: A normative discontent? European Journal of Public Health, 22(1), 116e121. https://doi.org/10.1093/eurpub/ckq178. McLean, S. A., Paxton, S. J., & Wertheim, E. H. (2010). Factors associated with body dissatisfaction and disordered eating in women in mildlife. International Journal of Eating Disorders, 43(6), 527e536. Mitchison, D., Hay, P., Slewa-Younan, S., & Mond, J. (2014). The changing demographic profile of eating disorder behaviors in the community. BMC Public Health, 14, 943. https://doi.org/10.1186/1471-2458-14-943. Nichter, M. (2001). Fat Talk: What girls and their parents say about dieting. Cambridge, Massachusets: Harvard University Press. de Oliveira, O., & Ariza, M. (2002). Transiciones familiares y trayectorias laborales xico urbano [Family transitions and female work trajectories femeninas en el Me in urban Mexico]. Cadernos Pagu, (17e18), 339e366. n a la adultez de Oliveira, O., & Mora, M. (2008). Desigualdades sociales y transicio xico contempor en el Me aneo [Social inequities and transition to adulthood in n, 14(57), 117e152. contemporary Mexico]. Papeles de Poblacio Palma-Coca, O., Hernandez-Serrato, M. I., Villalobos-Hernandez, A., UnikelSantoncini, C., Olaiz-Fernandez, G., & Bojorquez-Chapela, I. (2011). Association of socioeconomic status, problem behaviors, and disordered eating in Mexican adolescents: Results of the Mexican national health and nutrition survey 2006. Journal of Adolescent Health, 49(4), 400e406. https://doi.org/10.1016/ j.jadohealth.2011.01.019. Paquette, M. C., & Raine, K. (2004). Sociocultural context of women's body image.
Social Science & Medicine, 59(5), 1047e1058. https://doi.org/10.1016/ j.socscimed.2003.12.016. Perez-Gil, R. S., & Romero, J. G. (2010). Imagen corporal en mujeres de tres zonas xico: Percepcio n y deseo [Women's body image in three rural rurales de Me areas in Mexico: perception and desire]. Salud Publica de Mexico, 52(2), 111e118. Rodin, J., Silberstein, L., & Striegel-Moore, R. (1984). Women and weight: A normative discontent. Nebraska Symposium on Motivation, 32, 267e307. rez, A. G., Perez Gil-Romo, S. E., & De la Rosa Cruz, S. A. (2017). Cuerpo y Romero Jua n en una zona rural de Me xico: ¿una preocupacio n so lo de mujeres o percepcio n de hombres? Demetra, 12(1), 21e40. tambie Runfola, C. D., Holle, A. V., Trace, S. E., Brownle, K. A., Hofmeier, S. M., Gagne, D. A., et al. (2013). Body dissatisfaction in women across the Lifespan: Results of the UNC-SELF and Gender and Body Image (GABI) studies. European Eating Disorders Review, 21(1), 52e59. Slevec, J. H., & Tiggemann, M. (2011). Predictors of body dissatisfaction and disordered eating in middle-aged women. Clinical Psychology Review, 31(4), 515e524. https://doi.org/10.1016/j.cpr.2010.12.002. von Soest, T., & Wichstrom, L. (2008). The impact of becoming a mother on eating problems. International Journal of Eating Disorders, 41(3), 215e223. https:// doi.org/10.1002/eat.20493. Solmi, F., Hatch, S. L., Hotopf, M., Treasure, J., & Micali, N. (2014). Prevalence and correlates of disordered eating in a general population sample: The south east london community health (SELCoH) study. Social Psychiatry & Psychiatric Epidemiology, 49, 1335e1346. Stice, E., Marti, C. N., & Durant, S. (2011). Risk factors for onset of eating disorders: Evidence of multiple risk pathways from an 8-year prospective study. Behaviour Research and Therapy, 49(10), 622e627. https://doi.org/10.1016/ j.brat.2011.06.009. Stice, E., Ng, J., & Shaw, H. (2010). Risk factors and prodromal eating pathology. Journal of Child Psychology and Psychiatry, 51(4), 518e525. https://doi.org/ 10.1111/j.1469-7610.2010.02212.x. Tiggemann, M. (2015). Considerations of positive body image across various social identities and special populations. Body Image, 14, 168e176. https://doi.org/ 10.1016/j.bodyim.2015.03.002. Tiggemann, M., & Lynch, J. E. (2001). Body image across the life span in adult women: The role of self-objectification. Developmental Psychology, 37(2), 243e253. Turner, B. (1984). The body and Society: Explorations in social theory. Oxford: Blackwell. ~ o-Garcia, S. (2004). Validation Unikel-Santoncini, C., Bojorquez-Chapela, I., & Carren of a brief questionnaire to measure the risk of abnormal eating behaviors. Salud Publica de Mexico, 46(6), 509e515. n cultural a la Valenzuela, J. M. (2003). Por las fronteras del norte. Una aproximacio frontera M exico-Estados Unidos. Mexico: CONACULTA-FCE. Wade, T. D., Wilksch, S. M., & Lee, C. (2012). A longitudinal investigation of the impact of disordered eating on young women's quality of life. Health Psychology, 31(3), 352e359. https://doi.org/10.1037/a0025956. Warin, M., Turner, K., Moore, V., & Davies, M. (2008). Bodies, mothers and identities: Rethinking obesity and the BMI. Sociology of Health & Illness, 30(1), 97e111. https://doi.org/10.1111/j.1467-9566.2007.01029.x. Williams, L., Germov, J., & Young, A. (2011). The effect of social class on mid-age women's weight control practices and weight gain. Appetite, 56(3), 719e725. https://doi.org/10.1016/j.appet.2011.02.009.