Latina Women, Obesity, and Culture: Design, Rationale, and Participant Characteristics of the ¡Viva la Cultura Latina! Community-Based Study

Abstract

Objectives

Due to the high rates of overweight and obesity in Mexican and Puerto Rican women, interventions that target these groups are in great need. The purpose of this cross-sectional study is to explore the role of four cultural variables (acculturation, acculturative stress, ethnic identity, and cultural values) that may help to explain ethnic disparities in correlates of obesity (diet, physical activity, and body image) among Mexican and Puerto Rican women. This paper describes the study design, measures, and initial analyses of participant characteristics, as well as implications for the development of an obesity intervention.

Methods

A total of 552 participants contacted the study staff, with 368 completing the oral script and initial eligibility interview. From these, 227 completed the written consent, 199 completed the “Health and Culture” interview, and 186 completed all study components; 42 women completed the qualitative interview. Data for the current study is from the 199 participants who completed the “Health and Culture” interview.

Results

Mexican women were younger, more likely to have completed less than a 9th grade education, and more likely to prefer to complete the interview in Spanish (vs. English) compared to Puerto Rican women. In addition, Mexican women were more likely to be in the normal-weight group, be a non-smoker, and not have health insurance compared to Puerto Rican women.

Conclusion

Results revealed key initial differences between Puerto Rican and Mexican women which should be considered when developing interventions that address overweight and obesity.

Introduction

Overweight and obesity are linked to certain cancers [1, 2] and other medical conditions [3, 4]. Moreover, obesity disproportionately affects certain ethnic minority groups. Latinxs, in particular, have a high prevalence of overweight and obesity. For example, in the Hispanic Community Health Study/Study of Latinos (HCHS/SOL), the largest epidemiological study to focus exclusively on individuals of Hispanic/Latino descent in the US, high rates of obesity were found among Hispanic/Latino women (42.6%) and men (36.5%) [5]. The HCHS/SOL also revealed that Hispanic/Latino women were especially likely to have obesity. Women of Puerto Rican descent, for example, had the highest prevalence rate (51.4%) of all Hispanic/Latino groups [5]. Further, although overweight is a risk factor for breast cancer for both Hispanic and non-Hispanic White women, the risk for Hispanic women is independent of menopausal status, while the risk for non-Hispanic White women is apparent only in post-menopausal women [1], suggesting weight reduction is important for pre- and post-menopausal Latinx women. Thus, innovative obesity interventions that are targeted to Latinx women and particularly those with high rates of obesity are needed.

Obesity is affected by dietary intake, physical activity, and body image [6–8]. Several racial or ethnic differences have been found in these variables. In terms of dietary intake, in the Multiethnic Cohort Study (n = 215,251 women and men), Latino women were found to have among the highest caloric and fiber intake than other ethnic groups [9]. Likewise, in a smaller sample (n = 505) of Latin-American and Black women (53.6%), after controlling for body mass index (BMI), education, marital status, and number of children, Latin-American women consumed more carbohydrates (grams) and dietary fiber (total and soluble fiber) than Black women [10]. Further, the HCHS/SOL revealed important dietary variations across Hispanic/Latino backgrounds [11]. Individuals of Mexican background, for example, had higher intake of vitamin C, calcium, and fiber, and persons of Puerto Rican descent had the lowest intake of vitamin C, fiber, and fruits and vegetables [11].

Differences in physical activity across racial/ethnic groups, sex, and BMI status have also been identified. National self-reported data, for instance, indicate that Hispanic adults engage in less physical activity than non-Hispanic White adults [12], while objectively measured data suggests that Mexican Americans engage in more moderate to vigorous physical activity (MVPA) than non-Hispanic White and Black adults [13]. Results from the HCHS/SOL showed that persons of Puerto Rican and Dominican background have the highest levels of MVPA while individuals of Cuban descent had the lowest [14]. Using the same data, Palta et al. [15] found that accelerometer-measured physical activity differed by BMI group, such that those who had obesity were less active compared to those who had a normal weight, and most did not engage in at least 150 MVPA minutes per week. Another study found that Latina women participated in significantly less occupational and overall objective and subjective physical activity than Latino men, but Latina women participated in greater household activity than Latino men [16].

Body image, and body dissatisfaction in particular, have also been examined in relation to obesity [17], and several studies have described ethnic differences in body image [10, 18, 19]. Latin-American women exhibit a cultural preference favoring a large body type [10, 18], which may lessen determination to lose weight among overweight Latin-American women. In one study, in contrast to non-Hispanic White women who reported body dissatisfaction below the criterion for overweight, Hispanic women did not report body dissatisfaction until they were overweight [19]. Other research has not found differences between Latinx and non-Latinx women on body image, which may lend credence to the notion that Latina women may also have a preference for the thin ideals of Western culture [19, 20]. The inconsistency may be due to differences in the populations studied, study methodology, or possible within- and between-group differences such as acculturation level, biculturalism, ethnic identity, or cultural values.

Surprisingly, there is a dearth of research on ethnic differences in correlates of obesity such as diet, physical activity, and body image among Latinx individuals. To address these limitations, the ¡Viva la Cultura Latina! study was conducted to explore the role of four cultural variables (acculturation, acculturative stress, ethnic identity, and cultural values) and their relationship to three overweight/obesity-related variables (dietary intake, physical activity, and body image) among Mexican and Puerto Rican women (funded by NIH grants: R21-CA131433 and R21-CA131433-S). This study focused on Mexican and Puerto Rican women because they are the two largest Hispanic/Latino groups in the nation [21] and in Chicago, Illinois [21], where this study was conducted. In the following sections, we briefly review the migration history of Mexicans and Puerto Ricans in Chicago and outline the role of four culture-related variables and their relationships to diet, physical activity, and body image.

Chicago has long been a preferred destination for Mexican and Puerto Rican immigrants. Drawn by the job prospects in the railroads, steel mills, and packing houses, Mexican migration to Chicago began in the late 1910s [22]. These small waves of Mexican immigrants settled into three neighborhoods—Near West Side, Pilsen, and Back of the Yards—that were inhabited by various racial/ethnic groups, mostly of European origin [22]. The 1960s brought about important changes, especially in the North West Side, as urban renewal began to displace the Mexican population toward Pilsen, a historically Czech neighborhood [23]. The surge of Mexican immigrants to Pilsen quickly changed its demographics, transforming it into the first neighborhood where Mexicans comprised the majority. It was this dramatic growth in Pilsen that allowed Mexican Americans and immigrants to become well established and gain some political power in Chicago [22, 23]. Through the 1970s, the Mexican population in Pilsen continued to grow, expanding into a new neighborhood, Little Village [22]. To date, these two neighborhoods have the largest populations of Mexicans in Chicago, but Little Village has become the main port of entry for new Mexican immigrants. Currently, Mexicans are the largest Latinx group in Chicago, comprising 22% of the city’s population [24].

The first major wave of Puerto Ricans arriving in Chicago began in the 1940s [25]. While both Puerto Rican and Mexican migrants came to Chicago following the same job prospects [26], they had different experiences stemming primarily from their citizenship status [26]. While most Mexican persons were authorized or unauthorized immigrants, Puerto Ricans were US citizens. Upon their arrival, Puerto Ricans concentrated in Lincoln Park, West Town, and Humboldt Park [25]. Initially, they shared these neighborhoods with Polish and Mexican immigrants, but eventually a larger Puerto Rican community established itself in Humboldt Park [25]. A riot against police brutality in 1966 further cemented Puerto Rican roots in Humboldt Park, as several Puerto Rican community organizations were created as a response [25]. Today, Puerto Ricans comprise 3.8% of the Chicago population, making it the second largest Latinx community in the city [24].

Four cultural variables related to race and ethnicity may be associated with diet, physical activity, and body image. Acculturation refers to the changes in behavioral patterns and beliefs that occur from long-term interactions with the dominant culture [27]. Some evidence suggests that acculturation is associated with several healthy behaviors such as greater exercise and leisure-time physical activity [28, 29]. For instance, in one study, Mexican American men who were more acculturated had higher odds of meeting physical activity recommendations [30]. Another study found that Latino adults with a greater degree of linguistic acculturation were more likely to engage in vigorous-intensity leisure time physical activity (VLTPA) than those who were less acculturated [31]. Data from the HCHS/SOL revealed that Hispanic/Latino individuals who were born in the US, had lived longer in the US, or arrived in the US at an early age had the highest prevalence of obesity [32].

Albeit limited, extant evidence suggests that acculturation and body image may be associated. One study, for example, found that Mexican American women who were more acculturated showed a greater preference for thinner figures than women who were less acculturated; conversely, women who were less acculturated showed more tolerance for overweight figures [33]. Even more limited is the research on the relationship between biculturalism—the cognitive and cultural capacity to adapt and operate in two cultures [34]—and diet, physical activity, and body image. Therefore, additional research examining physical activity and body image across Latinx ethnicities and acculturation levels is warranted.

Acculturative stress describes the stress that individuals experience as they undergo the process of cultural change associated with acculturation. Some researchers have argued that cultural conflict, rather than acculturation, may be a more salient correlate of negative body image and maladaptive eating patterns [35, 36]. However, previous studies examining the impact of acculturation on diet, physical activity, or body image have not fully examined the role of acculturative stress.

Ethnic identity is a multidimensional construct that reflects one’s identity or sense of self as a member of an ethnic group [37]. There is limited research, however, that systematically examines the role of ethnic identity in diet, physical activity, and body image in Mexican and Puerto Rican women. Therefore, it is unknown whether one’s selfidentification or the emotional value of belonging to one’s ethnic group may underlie ethnic disparities in correlates of obesity.

Finally, cultural values may also underlie ethnic differences in diet, physical activity, and body image. Commonly cited Latinx cultural values that are relevant to women include familism (family), personalismo (personal relationships), fatalism (fatalistic beliefs), folk illness beliefs, spirituality, and marianismo (a set of cultural beliefs and values related to female gender roles) [38]. Thus, exploring the role of cultural values in diet, physical activity, and body image will aid in the development of interventions that are tailored to Latinx values.

The framework and rationale for this study are derived from the PEN-3 Theoretical Model [39], cultural proficiency and tailoring models [40, 41], and community-based participatory research [42–45]. The PEN-3 model emphasizes culture as the central reason for health behavior and the primary consideration for health promotion programs [46] and has been used in numerous studies to aid in the development of culture-centered interventions [39, 46–48]. This study also utilized a community-based participatory research framework [43–45], whereby the first author—who was the principal investigator (PI) on the research study—actively collaborated with key community informants.

The specific aims of the ¡Viva la Cultura Latina! study were to explore the role of acculturation, acculturative stress, ethnic identity, and cultural values in diet, physical activity, and body image in Mexican and Puerto Rican women. An additional aim was to identify logistical/practical and cultural considerations regarding the development of culturally proficient interventions for overweight Mexican and Puerto Rican women. Qualitative data from a subset of Latina women will be used to identify cultural messages and culture-specific content to be used in a future grant. Overall, data from this study will aid in developing culturally appealing content for an intervention focused on diet, physical activity, and body image among Mexican and Puerto Rican women.

The overall aim of the current paper is to describe the study design, measures, and initial analyses of participant characteristics, as well as discuss implications for the development of an intervention that addresses overweight/obesity. The primary outcomes related to diet, physical activity, and body image and cultural variables will be described in future publications.

Methods

Sample

The sample comprised 199 Latina women (104 Mexican and 95 Puerto Rican) aged 18–65. The initial plan was to recruit 60 participants in each of three weight categories: normal weight (BMI ≥ 18.5), overweight (BMI ≥ 25), and obese (BMI ≥ 30). However, 51 participants with normal weight, 58 with overweight, and 90 with obesity were included in this study.

Inclusion criteria were as follows: (1) self-identify as Mexican or Puerto Rican women (even if they also identified as biracial); (2) between the ages of 18 and 65, given the high rates of obesity among this age group; and (3) agree to provide informed consent. Exclusion criteria included the following: (1) those with a lower BMI limit (i.e., < 18.5 kg/m²), as this study was focused on persons with normal weight, overweight, or obesity; (2) those who had an eating disorder (bulimia nervosa, anorexia nervosa, binge eating disorder); (3) those who had plans to move from Illinois during the study time (i.e., 6 weeks); (4) those not able to speak/write English or Spanish; (5) those who reported being pregnant or planning to become pregnant throughout the course of the study; (6) those planning to breastfeed throughout the course of the study; and (7) those who participated in a study conducted by the first author called ¡Viva la Salud!, due to similarities in the measures in both studies [49].

Sample Size

Statistical power analyses, based on the correlations between acculturation and lifestyle variables, demonstrated that a sample size of 176 would be sufficient to detect a true correlation of r = 0.21, with 80% power. Specifically, we calculated that within Mexican and Puerto Rican women, a sample size of 88 in each group would be sufficient to detect a true correlation of r = 0.30 with 80% power. Since we planned to use multivariable models that control for extraneous factors, we further assessed the statistical power using partial correlations controlling for ten covariates. These power calculations indicated that with a combined sample of 176, we could still detect a partial correlation of 0.20 with 80% power, even if the covariates collectively explained 5% of the variance (total R² = 0.05). This power decreased slightly to detect a partial correlation of 0.15 if the covariates explained a larger portion of the variance (total R² = 0.50). Within each subsample (n = 88), the power analysis indicated sufficient power to detect partial correlations ranging from 0.28 to 0.21, depending on the explanatory power of the covariates (total R² of 0.05 and 0.50, respectively). The sample size also provides adequate power to detect potential differences between the Mexican and Puerto Rican women. With n = 88 per ethnic group, our analysis has an 80% power of identifying differences in correlations and means.

Recruitment Strategies

Participants were recruited using both direct and indirect strategies. Specifically, “intercept sampling” was used [50], which included in-person recruitment efforts in Latinx community organizations and centers, churches, festivals, and health fairs located in predominantly Mexican and/or Puerto Rican communities. Direct recruitment strategies included distribution of flyers in Latinx neighborhoods and university campuses in the Chicagoland area and emails to Latino and/or health organizations, as well as advertisements about the study in newspapers (hard copies and on-line) and other venues. At some of these locations, tables with flyers were set up and information about the study was provided. Indirect strategies included recruitment via advertisements published in newspapers or Latinx periodicals, advertisements in the newsletters of various community organizations and at the University of Illinois Chicago (where the first author was a faculty member), email list-serves and websites of various Latinx health organizations, and through referrals.

Study Procedures

All study procedures took place either at various community locations in Latinx neighborhoods in Chicago, at the University of Illinois Chicago, and/or at another location that was mutually convenient. Research procedures were as follows. First, research assistants obtained oral consent from potential participants and, once obtained, proceeded to conduct an initial 30-min eligibility screening over the phone or in person. This initial screening included self-reported measurements of height, weight, and eating disorder pathology. Participants who did not meet inclusion criteria were offered the opportunity to receive information about other health programs.

Next, an in-person meeting was conducted with individuals who were initially eligible to participate in the study. During the in-person meeting, written informed consent was first obtained, and then, research staff measured participants’ height and weight to confirm their eligibility. Individuals who were deemed not eligible were given a small token of appreciation. Eligible participants then completed a two-and-a-half-hour interview, which included sociodemographic and health and culture questionnaires. Measurements of body fat, hip, and waist were also obtained at this time. As an incentive for completing the health and culture interview, participants were asked to choose either a skillet or a slow cooker with a cookbook.

Participants were also asked to complete a 1-hour diet questionnaire (either in person on the same day as the health and culture interview or on their own within 7–10 days). Participants received a manicure and pedicure nail care set for participating in the 1-hour diet questionnaire. Lastly, participants were provided with accelerometer, which they were asked to use for 7 days and to record daily results and return results in-person or via mail. As an incentive after using the accelerometer for 7 days and returning the results to us, participants received a small food chopper. This study was initially approved at The University of Chicago (IRB: 16087A) and by the Research Review Board at Alivio Medical Center. When the PI changed institutions, this study was then approved by the Institutional Review Board at the University of Illinois Chicago (IRB: 2011–0275).

The initial plan was to conduct additional in-depth qualitative interviews on 20% (n = 36) of the sample (equal numbers of Mexicans and Puerto Ricans) and restricted to overweight and obese participants as a future grant would focus on those most at risk for overweight-related health problems. Participants had the option of choosing either a body weight scale or a DVD player with a Salsa Aerobics DVD for participating in this interview, and they received one of these choices at the end of the qualitative interview.

Community Partners and Hispanic/Latino Health Community Advisory Board

Guided by a Community-Based Participatory Research Approach [46], researchers partnered with a Community Advisory Board (Hispanic Latino Health Community Advisory Board (HLH-CAB)), composed of community partners and leaders, who collaborated on the design and methodology of the study and during the implementation phase. The diversity of HLH-CAB members in terms of educational background, place of birth, immigration status, and level of acculturation brought a wealth of perspectives that added value to the different aspects of the study. Members of the HLH-CAB provided feedback on the following: selection of incentives, recruitment and interview protocols, recruitment and retention strategies, culturally tailoring and translating consent forms and all study questionnaires and objective measures. Further, community partners facilitated the implementation of the study by providing locations for recruitment and to conduct interviews and objective assessments. Community partners have also served as co-authors on prior conference presentations, manuscripts, and grants. Finally, the first author has also been actively involved with several community-based initiatives focused on health and wellness, in which she has interacted with community members in different capacities. Collectively, these experiences allowed the research team to gain the critical skills, knowledge, and expertise in recruitment, retention, and engagement of Latinx community members to make the project successful.

Training

Project staff were trained in adherence to the Guidelines on Multicultural Education, Training, Research, Practice, and Organizational Change for Psychologists [41], which emphasize the awareness, knowledge, and skills needed to conduct culturally competent research. The training provided an overview of culturally competent research, and project staff were encouraged to discuss methods to increase multicultural awareness, knowledge, and skills. The community partnership approach used by the study also ensured the overall cultural sensitivity of this work in all its phases. Finally, the PI and senior staff members were certified to conduct the interviews and to assess anthropometric measures, as we have done in our prior studies [49].

Measures

Following several guidelines [41, 51–53], measures were translated and back translated from English into Spanish using a professional translation company previously used by the PI in prior studies [49]. Select members of the HLH-CAB convened with researchers to examine the translations’ ethnic/cultural validity, conduct pre-translation reviews, and examine Latinx ethnicity-related semantics.

Eligibility Interview

The initial eligibility interview was conducted via an in-person or phone interview. This interview included self-reported height and weight and a modified version of the Eating Disorder Examination Questionnaire (EDE-Q) [54, 55] to assess eating disorders. For participants that met initial eligibility criteria, self-reported measures of height and weight were verified with objective measures in person afterwards.

Anthropometric Measures

To assess levels of obesity, BMI (weight [kg]/height [m]²) [56] was computed. Height was assessed using a stadiometer, and weight was assessed with a Seca company digital scale with participants wearing light clothes and no shoes. Weight was classified according to national guidelines as follows: normal weight: BMI = 18.5–24.99; overweight: BMI = 25–29.99; obese class I: BMI = 30–34.99; obese class II: BMI = 35–39.99; and obese class III: BMI ≥ 40 [57]. Body fat percentage was measured via Bioelectrical Impedance Analysis, using a Tanita BF 682 scale. Fat distribution was measured by waist and hip circumference; and waist-to-hip ratio was also computed.

Health and Culture Interview

The following data were collected using a sociodemographic questionnaire: age (years), ethnicity, race, marital status, number of children, education, occupation and work status, country of birth, parents’ race and ethnicity, preferred language, literacy questions, length of US residence, health insurance, income, religion, spirituality, smoking status, and contact information for tracking purposes. Ethnicity was assessed by asking participants to indicate their ethnic background from a list, which included several Latin American nationalities along with Mexican and Puerto Rican and other races (e.g., African American or Black, Asian). Marital status was recoded as single (i.e., single and never married, separated, divorced, widowed) or partnered (i.e., married or remarried, cohabitating with partner). Participants were asked to indicate their highest level of education, which was recoded into a five-level education category: less than a 9th grade (i.e., never attended school, grade 6 [grammar school] or less, grade 7 to 9 [junior high or middle school]), 9–11th (i.e., some high school), high school graduate/GED, some college/associates degree (i.e., college 1 year to 3 years [some college or technical school], graduated 2-year college), college graduate and above (i.e., graduated 4-year college, part graduate/professional school, completed graduate, professional school), and other/more than one response. Work status was coded as working or not working. Participants reported their country of birth, which was recoded into a four-level variable representing birth in the US excluding Puerto Rico, Mexico, Puerto Rico, and Other. Participants indicated their preferred language for the interview to be in Spanish Only, English Only, Both Spanish and English, or No Preference. Health insurance was coded as insured or not insured. Participants reported their religious affiliation, which was recoded into a four-level variable: Catholic, Christian, more than one religion, or other (i.e., Baptist, Lutheran, Presbyterian, Protestant, Seventh-Day Adventist, Jewish, Buddhist, Hindu, Muslim, Agnostic, Atheist, no preference).

The questionnaire also included the Hollingshead Four Factor Index of Social Status [58], which combines information on sex, marital status, education, and occupation. This index will be used to assess socioeconomic status in future studies. Additionally, social desirability was measured by the Marlowe-Crowne Social Desirability Scale [59].

Stages of change for fruit and vegetable intake [60, 61], fat intake [62], and exercise [63, 64] were assessed using questionnaires which were based on the Transtheoretical Model [65]. Menopausal status, use of oral contraception, hormone replacement therapy, and hysterectomy were also assessed since these affect menstrual cycles. These variables were collected to include as covariates in future studies.

The Social Support and Exercise Survey [66] and the Social Support and Eating Habits Survey [66] were used to measure social support for exercise and eating habits (salt and fat intake), respectively.

The Acculturation Rating Scale for Mexican Americans-II (ARMSA-II) [67] was used to assess acculturation. This measure was selected because of its bi-directional approach, which allows Mexican culture and Anglo culture orientation to be assessed independently. For this study, the word “Mexican” was replaced with the word “Latino” to ensure that both Mexican and Puerto Rican participants could respond appropriately. The ARMSA-II and the Bicultural Identity Integration (BII) Scale-Version 1 [34] were used to assess biculturalism and bicultural identity, respectively.

Multidimensional Acculturative Stress Inventory (MASI) [68] which assesses stress associated with the acculturation process was used to measure acculturative stress.

The Multigroup Ethnic Identity Measure (MEIM) [69], which assesses ethnic identity search and affirmation, belonging, and commitment, was used to measure ethnic identity.

The Multiphasic Assessment of Cultural Constructs [67] was used to measure cultural values, namely, (1) familism, the importance that they placed on the family and their attitudes toward the family [67]; (2) fatalism, the extent to which people feel their destinies are beyond their control [67]; (3) personalism, a warm and personal way of relating to an individual [67]; (4) machismo, “the masculine force, which to one degree or another drives all masculine behavior” [70], and folk illness beliefs. Marianismo, a set of cultural beliefs and values that govern the role of Latina women in the family and encourages them to be modest, virtuous, subordinate, and sexually abstinent [71, 72], was assessed via the Latina Values Scales [73].

Body image was assessed using various tools. First, two sets of body image silhouettes/drawings cards (Card A and Card B) were used to assess current and ideal body image and body image discrepancy. The Figure Rating Scale [74], a visual scale that shows nine schematic figures of women ranging from underweight to overweight, has been widely used among Latinxs and other ethnic groups. A body image discrepancy score was calculated by subtracting ideal body image minus current body image. Card B displayed silhouettes specifically developed for African American participants and has been used in previous studies with African American women [75]. The male version of this questionnaire has also been used by the first author in a study with Latino men [76]. As an initial step to developing a Latinx-centered body image silhouettes/drawings, both cards were shown to participants to assess which drawings they thought looked most like their ethnic group and other questions pertaining to body image preferences. Second, the Body Areas Satisfaction Subscale of the Multidimensional Body-Self Relations Questionnaire (MBSRQ) was also used to assess satisfaction with various body parts [77]. Specifically, the MBSRQ measures personal attitudes toward one’s physical self, including evaluative, cognitive, and behavioral aspects.

Physical activity was assessed in two ways: self-report and accelerometer. The International Physical Activity Questionnaire [78] and the Physical Activity Recall Scale [79] were two self-report measures of physical activity. For the Physical Activity Recall, participants were asked whether they engaged in any regular activity designed to improve or maintain their physical fitness—either on their own or in a class. If they had engaged in an activity to improve their physical fitness during the last week, they were asked to report the frequency and duration of that activity. The Borg Exercise Intensity scale [80] was used to assess exercise intensity. Participants were also asked to report their hours of sedentary behavior, assessed as number of hours per day they watched television [80]. In addition, leisure time and non-leisure time physical activity was assessed using the interviewer-administered long form of the International Physical Activity Questionnaire [81]. Finally, participants were asked to use a piezo-electric accelerometer (New Lifestyles NL-2000) for 7 consecutive days to objectively assess their physical activity.

Dietary intake was assessed using the Block 2005 Food Frequency Questionnaire [82]. The Block 2005 Food Frequency Questionnaire assesses usual intake of nutrients and food groups. It has been used in previous studies among Latinxs and has been shown to be valid and reliable [82]. Group differences in calories, fat, fiber, fruit, and vegetable intake will be explored in future papers. Other measures related to weight and weight perceptions included weight loss patterns and attempts, participants’ personal descriptions of their own weight, and others’ perceptions of the participant’s weight. These items will be used for descriptive purposes.

For the purposes of a future culture-based intervention, logistical (e.g., location and frequency of meetings), practical (e.g., preferences regarding types of physical activity, music, incentives), and cultural questions (e.g., cultural considerations and cultural messages) were assessed using structured and open-ended questions.

Finally, an in-depth qualitative interview, adapted from James [46], was conducted with a subset of the overweight and obese participants to assess cultural aspects and barriers to healthy eating/diet, physical activity, and body image. The qualitative interviews were audiotaped and transcribed and will be analyzed to develop Latinx-specific intervention messages in future studies.

Results

First, results from the enrollment process are presented. Second, analyses of select participant characteristics are described. Tests of the study hypotheses will be reported in future papers.

Participant Study Enrollment

Of the 552 participants who contacted the study staff, 368 eventually completed the oral script and initial eligibility interview (Fig. 1). Of these, 227 participants completed the written consent, and 199 completed the Health and Culture Interview. Of note, 13 participants did not complete the additional study components of the diet questionnaire and accelerometer use.

Fig. 1.

CONSORT flow diagram of enrollment and completion of ¡Viva La Cultura Latina! Study. ^aThese 53 participants either felt they did not meet the initial requirements, they did not like the study incentives, and/or did not have time to participate. ^bCould not reach these participants after contacting them 3–4 times or because of a wrong phone number. NI new institution, PRI previous institution

Data reported in Tables 1, 2, and 3. are from the 199 participants who completed the Health and Culture interview.

Table 1.

Sample demographic characteristics

		Total sample	Mexican	Puerto Rican
		N = 199	n = 104	n = 95	p
Age (years), mean (SD)		41.3 (12.7)	38.8 (12.5)	44.0 (12.4)	0.003
Country of birth, n (%)					< 0.001
	US, except Puerto Rico	95 (47.7)	41 (39.8)	54 (56.8)
	Mexico	62 (31.2)	62 (60.2)	0 (0.0)
	Puerto Rico	40 (20.1)	0 (0.0)	40 (42.1)
	Othera	1 (1.0)	0 (0.0)	1 (1.0)
Education, n (%)					0.036
	Less than 9th	21 (10.6)	18 (17.3)	3 (3.2)
	9th–11th	20 (10.1)	10 (9.6)	10 (10.5)
	High school graduate/GED	39 (19.6)	17 (16.3)	22 (23.2)
	Some college/AA	68 (34.2)	32 (30.8)	36 (37.9)
	College graduate and above	45 (22.6)	23 (22.1)	22 (23.2)
	Other/more than one response	6 (3.0)	4 (3.8)	2 (2.1)
Marital status, n (%)					0.098
	Single	103 (51.8)	48 (46.2)	55 (57.9)
	Partnered	96 (48.2)	56 (53.8)	40 (42.1)
Work status, n (%)					0.815
	Working	108 (54.3)	57 (55.3)	51 (53.7)
	Not working	90 (45.5)	46 (44.7)	44 (46.3)
Religion, n (%)					< 0.001
	Catholic	92 (46.2)	66 (63.5)	26 (27.4)
	Christian	44 (22.1)	14 (13.5)	30 (31.6)
	More than 1	33 (16.6)	10 (9.6)	23 (24.2)
	Otherb	30 (15.1)	14 (13.5)	16 (16.8)
Preferred interview language, n (%)					< 0.001
	English	130 (65.3)	53 (51.0)	77 (81.1)
	Spanish	69 (34.7)	51 (49.0)	18 (18.9)

Ethnic group differences on continuous variables were analyzed using t-tests; categorical variables were analyzed using chi-squared tests

SD standard deviation, GED general education degree, AA associate of arts degree

^aRepresents one person who indicated birth in Guatemala and one person who did not answer the question

^bIncludes Presbyterian, Protestant, Seventh-Day Adventist, Agnostic, Atheist, no preference, and other

Table 2.

Differences in health characteristics between Mexican and Puerto Rican women

		Total sample	Mexican	Puerto Rican
		n = 199	n = 104	n = 95	p
BMI (kg/m2), mean (SD)		30.6 (7.6)	29.2 (6.8)	32.2 (8.1)	0.006
Weight category, n (%)					0.067
	Normala	51 (25.6)	33 (31.7)	18 (18.9)
	Overweightb	58 (29.1)	31 (29.8)	27 (28.4)
	Obesityc	90 (45.2)	40 (38.5)	50 (52.6)
Body fat %, mean (SD)		37.7 (8.3)	35.9 (8.5)	39.6 (7.7)	0.002
Hip (cm), mean (SD)		109.8 (15.9)	106.8 (15.3)	113.0 (16.0)	0.006
Waist (cm), mean (SD)		101.7 (18.0)	98.0 (17.6)	105.7 (17.6)	0.003
Smoking status, n (%)					0.091
	Never	141 (70.9)	77 (74.0)	64 (67.4)
	Current	25 (12.6)	8 (7.7)	17 (17.9)
	Former	33 (16.6)	19 (18.3)	14 (14.7)
Insurance coverage, n (%)		125 (62.8)	52 (50.0)	73 (76.8)	< 0.001

T-tests were conducted for continuous variables and chi-squared tests for categorical variables

SD standard deviation, BMI body mass index

Table 3.

Weight categories of 42 participants randomized into qualitative interview

	Total sample	Mexican	Puerto Rican
	N = 42	n = 25	n = 17	p
Overweight, n (%)	21 (50.0)	14 (56.0)	7 (41.2)	0.346
Obesity, n (%)	21 (50.0)	11 (44.0)	10 (58.8)

Ethnic differences in overweight/obesity analyzed by chi-squared tests

Sample Demographics

As presented in Table 1, the mean age of the 199 Latina women in our sample was 41.3 (SD = 12.7). Most participants (48%; n = 95) were born in the US exclusive of Puerto Rico, followed by Mexico (31%; n = 62), Puerto Rico (20%; n = 40), and another country (1%; n = 1). More than half of the participants had at least some college education (57%; n = 113), were single (52%; n = 103), and were employed (54%; n = 108). Almost half (46%; n = 92) of participants were Catholic and more than half (65%; n = 130) preferred to complete the interview in English.

Several demographic differences between Mexican and Puerto Rican women emerged. Compared to Puerto Rican women, Mexican women were younger and less likely to have been born in the US mainland. In terms of educational attainment, Mexican women were more likely than Puerto Rican women to have obtained less than a 9th grade education. Mexican as compared to Puerto Rican women were more likely to report that their religion was Catholic, less likely to report that their religion was Christian, and less likely to report that they had more than one religion. Finally, Mexican women were less likely to prefer to complete the interview in English than Puerto Rican women.

Health Characteristics

As seen in Table 2, the mean BMI for the entire sample was in the overweight range (M = 30.6; SD = 7.6). Almost half of the sample had weight in the obesity category (45%; n = 90). Similarly, the mean body fat percentage was high (38%; SD = 8.3%). The mean hip measurement was 109.8 cm (SD = 15.9 cm) and the mean waist measurement was 101.7 cm (SD = 18 cm). Finally, 13% (n = 25) women were current smokers and more than half (63%; n = 125) were covered by health insurance.

Mexican and Puerto Rican women differed on three health characteristics. First, Mexican women were more likely to have normal weight compared to Puerto Rican women. Second, Mexican women were less likely to be current smokers. Third, Mexican women were less likely to have health insurance coverage than Puerto Rican women. No ethnic group differences emerged for other health characteristics.

Characteristics of Participants in the Qualitative Interview

As seen in Table 3, 21% (n = 42) of participants completed the qualitative interview. The target of six interviews per weight categories of overweight and obesity were exceeded. No significant ethnic group differences emerged for weight categories.

Discussion

Rates of overweight and obesity are disproportionately high and increasing among Puerto Rican and Mexican women. To combat this pressing public health issue, culturally appropriate, tailored interventions are critical. The NIH-funded ¡Viva la Cultura Latina! study was designed to assess the role of cultural variables underlying obesity to inform the development of tailored interventions for Mexican and Puerto Rican women. The findings of this study provide an overview of selected demographic and background characteristics and highlight significant differences between Mexican and Puerto Rican women. The aforementioned results are preliminary findings but, overall, reveal key initial differences between Latinx ethnic groups which should be borne in mind when developing obesity interventions.

First, results revealed that the development of interventions should take into account the language preferences of women. A greater number of Mexicans versus Puerto Ricans may be more attracted to programs that are offered in Spanish, since 49% of Mexicans versus 19% of Puerto Ricans preferred to complete the interviews in Spanish. However, the fact that slightly over half (51%) of Mexicans preferred to complete the interview in English, while 81% of Puerto Ricans preferred to complete the interviews in English suggests that it may be important to consider the preferred language of the target population to develop effective exercise programs for Mexicans and Puerto Ricans. Further, our findings imply that programs which are tailored to Puerto Ricans, as compared to Mexicans, should consider providing programming in both English and Spanish, as well as advertising programs through English language media sources.

Second, visual inspection of the means revealed that Puerto Rican women had a higher BMI, waist circumference, and percent body fat than their Mexican counterparts, although this did not significantly differ between groups. Previous studies have shown that Puerto Rican women have higher rates of overweight and obesity than other Latinas [5, 83]. In fact, in the HCHS/SOL, about half (51.4%) of women of Puerto Rican background had obesity [5]. Collectively, these findings suggest that interventions to reduce body weight is urgent among both groups and especially among Puerto Rican women.

Overall, about 18% of Puerto Rican women reported smoking, and they were more likely than Mexican women to smoke. This finding aligns with a previous HCHS/SOL study which found that the prevalence of current smoking was highest among women of Puerto Rican descent (32.6%) compared to women of other Hispanic/Latino background [84]. Moreover, smoking has been previously found to be associated with central fat accumulation, especially among women [85], and former smokers are more likely to have obesity than never smokers [86, 87]. This finding highlights the need to develop smoking cessation programs, in addition to obesity interventions for Puerto Rican women.

Finally, the fact that 50% of Mexicans versus 77% of Puerto Ricans were insured emphasizes the potential need to explore methods of payment for healthcare for Mexicans, in particular, who report lower rates of health insurance coverage overall. Future data analyses will reveal additional considerations for the development of an intervention for overweight/obese Mexican and Puerto Rican women, as well as similarities and differences to consider in the design of such an intervention.

There are some limitations that warrant consideration when interpreting our findings. First, participants were recruited using a convenience sampling approach, which limits the generalizability of our findings. To increase the representativeness of our sample, we worked with our community partners to recruit extensively at various venues throughout the Chicagoland area. Second, sample size may not be large enough to detect small effect sizes, especially within a specific ethnic group (e.g., Puerto Ricans). Third, the cross-sectional nature of our study prevents us from drawing causal inferences. Lastly, several measures, including diet and smoking, are self-reported, which could lead to social desirability bias.

Limitations notwithstanding, this study had a number of strengths, principally the community-engaged approach of the ¡Viva la Cultura Latina! study. Community-based approaches were important for at least three main reasons. First, involving key community partners helped us ensure that our target health outcomes aligned with their own priority health areas. Unlike university researchers who may not be as involved in their communities, community partners, especially those who share their culture and language, are more familiar with the specific needs of their community and what is more culturally appropriate. This had a twofold benefit: community partners were more willing to collaborate, and our findings will contribute to their own ongoing health programs. For instance, a few of our community partners have already identified obesity as a priority health area and were already developing or had programs available to increase physical activity and improve diet among Latinx community members.

Second, the support from our community partners to ensure that measures were correctly translated into Spanish and culturally appropriate for Mexicans and Puerto Ricans was of paramount importance. For instance, variations in Spanish wording between different Latinx subgroups is quite common, and, unfortunately, they are often ignored in research. Engaging community partners, who have a deep understanding of the target Latinx population, in translating measures may lead to a greater external validity of findings [50]. In addition to contributing to refining measures, community partners were also instrumental in developing successful recruitment and retention strategies. Importantly, community partners played a significant role in fomenting trust toward the university research team and facilitating participation among community members. Community partners, especially community health workers/promotores de salud or outreach workers, have an enormous capacity to reach high numbers of people in the communities they serve and even outside their communities.

A third important aspect of our community engaged approach was the dissemination of health information to community members. A key tenet of community-based research is reciprocity [88, 89], and one important way in which we reciprocated to the local communities was by disseminating culturally tailored and targeted information related to smoking, obesity, and physical activity in Latinxs. This was particularly important and timely given the high rates of smoking and obesity that were found among women in this and other studies. Health brochures were developed in English and Spanish by the research team and were distributed widely throughout various areas in the Chicagoland area. The health brochures were also valuable to community partners, who continued using them within their own health education/promotion programs after the study ended.

Budding research has uncovered numerous differences in health-related factors between Latinx subgroups. Few studies, however, have focused on the association between cultural factors and correlates of obesity among Latinxs using a community-based approach. Accordingly, the ¡Viva la Cultura Latina! study was designed to identify factors related to acculturation, acculturative stress, ethnic identity, and cultural values that may aid in the development of community-based, culturally appropriate, tailored interventions to address diet, physical activity, and body image in Mexican and Puerto Rican women. The preliminary findings presented in this paper, for example, highlight some important differences between Mexican and Puerto Rican women which require consideration when designing culturally tailored health interventions. Future analyses will explore the independent and interactive influences of various cultural factors on correlates of obesity among Mexicans and Puerto Ricans. The resulting information is expected to expand our knowledge on these populations’ diverse cultural factors, health behaviors, and health outcomes, as well as any potential associations, and to aid in prevention and mitigation of obesity and other chronic diseases.

Data Availability

The study data is available upon reasonable request from the first author via email. The study materials are not publicly available.

Code Availability

The study data is available upon reasonable request from the first author via email.