Utility of Acculturation in Physical Activity Research in Latina Adults: An Integrative Review of Literature

Abstract

Latina adults in the United States have a disproportionately higher prevalence of chronic diseases related to low physical activity levels than non-Hispanic women. Literature indicates that acculturation may be a contributing factor to being physically active, but the extent of this association remains unclear. An integrative review of literature was conducted on studies that examined acculturation as it relates to physical activity in Latinas in the United States. Our review of 33 studies revealed inconsistent measurement and conceptualization of acculturation and physical activity across studies. Findings from this review reinforce the importance and continued use of acculturation by behavioral researchers; however, acculturation, as conceptualized in the studies reviewed, may not have had an influence on health as much as traditions, life patterns, and resources of Latinas.

Latinas living in the United States have a disproportionately high prevalence of obesity (National Center for Health Statistics, 2011) and chronic diseases, such as diabetes and cardiovascular disease, related to physical inactivity when compared to non-Hispanic White women (Schiller, Lucas, Ward, & Peregoy, 2012). Performing regular physical activity (PA) can help control weight and reduce the risk of premature death and chronic illnesses (e.g., type 2 diabetes, cardiovascular disease, and some cancers; Centers for Disease Control and Prevention, 2012). Despite the benefits of regular performance of PA, inactive lifestyles are disproportionately high among the U.S. Latino population—particularly among Latina women who are even less physically active than Latino men (Schiller et al., 2012).

Latinos are one of the fastest growing ethnic minority groups in the United States (Humes, Jones, & Ramirez, 2011). Accounting for more than half of the population growth in the United States between 2000 and 2010 (Humes et al., 2011), Latinos come from many different countries—each with its own unique cultural traditions. Many of the factors influencing the performance of PA in Latinas appear to be shaped by deeply embedded cultural norms, values, and traditional gender roles (Castaneda et al., 2002; Comas-Dias, 1988; Keller, Coe, & Moore, 2014) that can emphasize self-sacrifice and putting the needs of the family before a woman’s own needs and defines a woman’s self-worth based on what she does for others (D’Alonzo, 2012). Latinas have stressed the importance putting family and caretaking responsibilities first (“mi familia viene primero/my family comes first”); the demands of caring for their children, grandchildren, spouses, and parents have commonly been identified as a barrier to being physically active (Gonzales & Keller, 2004). This devotion to family is so highly valued that performing PA can be perceived as a self-indulgence, as an activity that can lead a woman to be neglectful of her family’s needs (D’Alonzo, 2012). In a number of studies, Latinas reported that despite being aware of the health benefits of PA, family responsibilities regularly took precedence over becoming more physically active (Gonzales & Keller, 2004; S. M. Martinez, Arredondo, Perez, & Baquero, 2009). For Latinas, the health benefits of PA must be understood in the context Latina’s families and their roles as caregivers (Gonzales & Keller, 2004); thus, intervening to reduce health disparities in Latinas requires further understanding of how cultural values influence health behavior, and ultimately health status following migration and presumed acculturation.

Acculturation has commonly been defined as the process of adaptation or assimilation by an ethnic or racial group to that of a new culture (Berry, 2003), which can result in changes in values, attitudes, and behaviors (Cuéllar, Arnold, & Gonzalez, 1995). There exists a substantial body of work that reports convincing and statistically significant associations between an individual’s perceived acculturation and his or her health behaviors, such as PA (Abraído-Lanza, Chao, & Flórez, 2005; Ahluwalia, Ford, Link, & Bolen, 2007; Cobas, Balcazar, Benin, Keith, & Chong, 1996; Coonrod, Bay, & Balcazar, 2004; Coreil, Ray, & Markides, 1991; Daviglus et al., 2012; Evenson, Sarmiento, & Ayala, 2004; Marquez & McAuley, 2006). Existing research, however, has been unable to explain whether or not acculturation (depending on the researcher’s particular definition of acculturation) has a direct, modifying, or mediating effect on health or health behaviors. An understanding of the role that this key mechanism may play in promoting behavioral change is essential for effectively reducing illness and promoting healthy lifestyles in this underserved population. To effectively reduce health disparities, interventions need to be developed based on an in-depth understanding of the wide range of factors—biological, cultural, environmental—that influence health in the Latina population; thus, it is essential to understand the mechanism through which acculturation influences health outcomes for specific behaviors (i.e., PA) in Latina adults in the United States. For instance, migration can result in changes in physical environments and access to resources, which can then influence PA. A shift from residing in a rural to an urban environment may influence participation in PA through increased reliance on motorized transportation (public transportation, access to cars) and a reduction in active transportation, such as walking or riding a bicycle. Latinas have reported that in their native countries they relied more on walking as a means of getting to the grocery store, schools, and workplaces due to the proximity of resources; however, due to farther distances of resources and perception of neighborhood safety in the United States, there is greater reliance on motorized transportation (D’Alonzo, 2012; Sussner, Lindsay, Greaney, & Peterson, 2008). Furthermore, changes in environment might also influence employment, such as changing from working in physically active agricultural environment to a more sedentary work environment.

If traditional lifestyles and values of Latinas are associated with level and type of PA and with health status, then interventions may need to be designed to include and embrace these traits rather than attempting to move individuals along a trajectory of greater acculturation. While values such as putting their children first or lack of time for PA due to precedence for family responsibilities are often reported by Latina mothers as barriers to PA, they can be also used to promote healthy lifestyles. These factors have been incorporated into the design of health promotion interventions with Latinas to increase PA by including time management skills and strategies for walking with family members, children, spouses, and parents (Keller et al., 2014). Thus, including and embracing these traits would mean we would first need to increase our understanding of how deeply embedded cultural norms and values, including traditional gender roles (Castaneda et al., 2002; Stevens, 1973), may influence health behaviors, and how and if these values change with acculturation or assimilation to U.S. lifestyles.

Previous reviews of published literature on acculturation and health have focused on overweight/obesity and acculturation among adult migrant populations (Delavari, Sonderlund, Swinburn, Mellor, & Renzaho, 2013), Hispanic health (Hunt, Schneider, & Comer, 2004), and multiple health behaviors, including PA, in Latinos (Abraído-Lanza et al., 2005; Lara, Gamboa, Kahramanian, Morales, & Bautista, 2005), as well as dietary behaviors (Ayala, Baquero, & Klinger, 2008) and food patterns (Gerchow et al., 2014) in exclusively Latina populations. Although a number of reviews included PA behaviors, no recent reviews of published literature were found that could explain the association between PA and acculturation among Latina adults. Understanding factors that influence performance of PA in this population is important due to the high rates of obesity and chronic illness associated with insufficient PA (Schiller et al., 2012) as well as the central role in family and caretaking responsibilities among Latina adults. Latinas’ strong devotion to their families and central role in attending to their needs not only influences her own ability to be physically active (D’Alonzo, 2012; Gonzales & Keller, 2004), but could potentially affect the health of her family and children. Furthermore, focusing on Latina adults in health promotion is important as certain life events that occur during adulthood (e.g., getting married or starting to live with a partner, having a baby) are associated with a decrease in PA (Brown, Heesch, & Miller, 2009). In this article, we examine the role of acculturation as an influence on PA and challenge the utility of the construct in influencing PA.

Given the complexity of factors influencing PA in Latina adults in the United States, there is a need for greater understanding of the relationship between acculturation and PA in this underserved population. An integrative review of literature, as compared with a systematic review of literature or meta-analysis, is a broad method for reviewing research and gaining deeper understanding about a phenomenon of concern by allowing for simultaneous inclusion of experimental and nonexperimental study designs (Whittemore & Knafl, 2005). Thus, the purpose of this integrative literature review is to (1) identify the state of science on the association of acculturation and PA in Latina adults in the United States and (2) evaluate the utility of using acculturation as a variable to account for PA-related health disparities in Latinas. This review adds to the gap in literature on the association of acculturation and PA in Latina adults.

Method

This search of the literature was conducted using five databases: PubMed, CINAHL, PsycInfo, Web of Science, and JSTOR. Keywords searched were physical activity, exercise, acculturation, Hispanic, Latina, Latino, and women. Inclusion criteria were peer-reviewed articles published in the past 10 years (September 2004 to September 2014) that reported an association between acculturation and PA in Latina adults. The current review was limited to articles published between 2004 and 2014 to include a 10-year span of research. Articles were excluded from the review if the study samples were comprised exclusively of older adults (as defined by author’s use of the term “older” or entire study sample >60 years of age), menopausal or postmenopausal women, or children/adolescents <18 years of age.

Using these search terms, 102 articles were located in the database PubMed, 59 articles were located in CINAHL, 74 in PsycInfo, 184 in the database Web of Science, and 498 were located in JSTOR. After review of titles and abstracts for these articles, 65 full-text articles were reviewed to establish if they met eligibility criteria. Reasons for ineligibility included that findings on the association of acculturation and PA in Latinas were not reported (n = 20), study targeted older adults (n = 3), used the term “English language proficiency” independently of the concept of acculturation (n = 2), were literature review articles (n = 6), or included men only (n = 1). Thirty-three articles met eligibility criteria and were included in this review.

Results

Design and Participants

The majority of studies (n = 24, 73%) used cross-sectional designs, and 4 (12%) used longitudinal designs to assess the association between acculturation and PA. Four (12%) of the studies used qualitative focus group designs (one also had interviews), and 1 (3%) used a mixed methods study design that included the use of cross-sectional surveys and structured interviews (Himmelgreen, Daza, Cooper, & Martinez, 2007). Table 1 illustrates designs of studies reviewed. Sixteen (48%) of the study samples were comprised exclusively of females, while 17 (52%) included both males and females. Two studies included samples of newly arrived Latino immigrants or first-generation Latinos (Evenson et al., 2004; Himmelgreen et al., 2007). Five studies targeted Latinos with specific health conditions such as diabetes (Barrera, Toobert, Strycker, & Osuna, 2012; Mainous, Diaz, & Geesey, 2008) and pregnant or postpartum Latinas (Chasan-Taber et al., 2007; Gollenberg, Pekow, Markenson, Tucker, & Chasan-Taber, 2008; Sussner et al., 2008). This review included a total of 144,432 participants (90,579 Latinos). Study size ranged from 21 to 36,401 participants, with a median sample size of 671.

Table 1.

Study Design and Measures of Acculturation and Physical Activity.

Study	Design	Sample characteristics	Acculturation measures	PA measures	Results
Abraído-Lanza et al. (2005)	Cross-sectional; secondary analysis of data from 1991 National Health Interview	Latino and non-Latino White adults, representative sample of noninstitutionalized U.S. population ≥18 years N = 36,401 Latinos n = 3,154 (57.7% female)	Nativity and length of time residing in United States	Self-reported engagement in recent exercise activity (e.g., walking, jogging, sports activities) during past 2 weeks	Higher acculturated Latinas were more likely (OR = 2.63, 95% Cl = 1.92, 3.60) to have exercised during the past 2 weeks than less acculturated Latinas when controlling for age and SES.
Afable-Munsuz, Ponce, Rodriguez, and Perez-Stable (2010)	Cross-sectional; secondary data analysis of self-report surveys from 2000 U.S. Census to California Health Interview Survey	Mexican (n = 6,008), Chinese (n = 1,313), Filipino (572), and adults ≥18 years responding to telephone survey in California	Generation Language spoken at home Language of interview	Data (leisure time, nonleisure time, and any PA) linked from 2000 U.S. Census to California Health Interview Survey	LTPA increased from 28% among first generation to 37% for third generation Mexican adults, while NLTPA demonstrated a significant decreasing trend from 35% to 28% from first to third generations. Odds are significantly higher for LTPA and significantly lower for NLTPA in third generation than first generation. After adjustment for SES and neighborhood-level variables, odds of PA significantly higher for Mexican who are bilingual in the home (rather than Spanish only).
Arredondo, Elder, Ayala, and Campbell (2005)	Cross-sectional; used a subset of participants from an 18-month randomized field trial.	Spanish-language-dominant Latinas from two San Diego County areas N = 211 (100% female)	ARSMA-II	IPAQ	Significant relationships between 30-minute PA and acculturation (r = .18, p < .01), and 20-minute vigorous PA and acculturation (r = .30, p < .001).
Banna, Kaiser, Drake, and Townsend (2012)	Cross-sectional; secondary analysis of telephone survey data from California Women’s Health Study	Self-identified Latina adults ≥18 years responding to household telephone survey in California N = 1,298 (100% female)	Five questions from Marín acculturation scale (Marín, Sabogal, Marin, Otero-Sabogal, and Perez-Stable, 1987)	Frequency and duration of PA (assessed by modified questions from BRFSS)	No significant differences were found between acculturation levels and MVPA performed weekly (p = .14).
Barrera et al. (2012)	RCT; 2 groups; experimental	Latina adults diagnosed with type 2 diabetes for at least 6 months. Recruited from Kaiser Permanente clinics in Denver, Colorado. N = 280 (100% female)	Short form of the ARSMA-II (Cuéllar, Arnold, & Maldonado, 1995)	IPAQ (Craig et al., 2003)	At baseline, Latina orientation was negatively associated with physical activity (r = −.116, p = .053).
Berrigan, Dodd, Troiano, Reeve, and Ballard-Barbash (2006)	Cross-sectional; data from the 2000 NHIS	N = 4,558 (50.3% female)	Eight questions concerning language	LTPA-based questions on frequency and duration of light, moderate and vigorous intensity PA NLTPA based on distinct questions that assessed three types of activity: (1) transportation-related activity; (2) sitting, standing, or walking; (3) lifting and carrying things	Self-reported adherence to recommendations concerning LTPA increased from 22.6% in the least acculturated tertile to 47% in the most acculturated tertile. Prevalence of walking or bicycling for errands decreased from 25.2% to 18.2%, and prevalence of standing or walking during most of the day decreased from 82.8% to 65.6% as acculturation increased.
Bungum, Thompson-Robinson, Moonie, and Lounsbery (2011)	Cross-sectional; 52-item telephone survey	Hispanic citizens ≥18 years residing in Las Vegas, Nevada N = 331 (52.8% female)	Language of interview	Single question used to assess PA readiness	Being younger, attending at least some college, and having higher acculturation (English language preference) were predictive of being more physically active.
Chasan-Taber et al. (2007) a	Prospective cohort study; nonexperimental, assessments at 2 time points	Pregnant Latina women (≥24 weeks gestation) ages 16–40 years receiving prenatal care at a large tertiary care facility in Western Massachusetts N = 1,231 (100% female)	Birthplace Language preference for speaking, reading, and writing	Modified version of the KPAS, adapted from the Baecke physical activity survey. KPAS questions grouped into 4 sections: household and family care activities, occupational activities, active living habits, and participation in sports and exercise.	Being born in Puerto Rico or a foreign country was associated with an increased likelihood of high occupational activity (OR = 1.5, 95% Cl = 1.1, 2.1), while Spanish language preference was associated with a decreased likelihood (ptrend = .01). Spanish language preference was inversely associated with a high level of sports/exercise participation ptrend = .02).
Creighton, Goldman, Pebley, and Chung (2012)	Cross-sectional; self-report data from the Los Angeles Family and Neighborhood Survey (L.A. FANS-2)	Mexican-origin, U.S.-born Black and White male and female adults ages ≥18 years residing in neighborhoods in Los Angeles County N = 1,610 (n = 1,112 Mexican origin)	Measures of linguistic and social acculturation based on Marín and Gamba (1996) and Marín et al. (1987) bidimensional acculturation scales	NHIS questions on PA (assesses episodes of 20 self-reported vigorous PA in 3 categories: housework, work, and leisure)	Second and 3rd+ generation Mexicans and 3rd+ generation Whites more likely to engage in vigorous leisure-time exercise than recent Mexican immigrants. However, when social and linguistic acculturation was added, these coefficients were smaller and no longer significant (except for 2nd generation Mexicans at p <. 10). The coefficients for social acculturation were significant only for leisure-time exercise and indicated that having more ethnically diverse friends is associated with a greater frequency of exercise. The effect of linguistic acculturation was significant only for housework and for all exercise combined (when SES is held constant).
D’Alonzo (2012)	Qualitative; ethnographic focus group study	Latinas 18–40 years enrolled in a women’s leadership program in an urban community in Central New Jersey N = 28 (100% female)	Data collection and analysis were driven by the use of Spradley’s DRS (Spradley, 1979) through which the researcher learns about the reality of an individual’s experience through a study of the meaning behind the spoken word.	Each session included several questions about physical activity (e.g., importance for women to get enough PA, obstacles to becoming more physically active, possible ways to address obstacles).	Themes: Fast pace of life in United States and the woman is the center of the family. Participants discussed lifestyle changes associated with acculturation (all the women reported participating in more PA in their home country, no female role models for PA, rarely saw mothers engage in LTPA, they adapted to acculturation faster than spouse causing family discord).
Echeverría et al. (2013)	Cross-sectional data from the 2010 NHIS	Latino adults from a nationally representative sample of noninstitutionalized U.S. adults N = 4,929 (52.1% female)	Single question on language most often used (proxy measure of acculturation)	Used the 2008 physical activity guidelines for adults to classify individuals as meeting physical activity guidelines for both aerobic and strengthening, either one, or neither	Individuals with high English-language use were significantly more likely to meet current physical activity guidelines than those with low English-language use (PR 2.29; 95% Cl = 1.98–2.66).
Evenson et al. (2004)	Cross-sectional; face-to-face interviews	First-generation immigrant Latina adults ages 20–50 years. Interviewed at a variety of community-based locations (festivals, bus stops, laundry mats, Latin markets, and malls) N = 671 (100% female)	Four-item language scale developed by Marín et al. (1987), length or residence in United States, age of arrival to United States	2001 BRFSS questions—moderate or vigorous activity, excluding occupational activity, performed in 10-minute bouts	Participants with higher English language acculturation were more likely to be physically active than those with lower English language acculturation. Latinas who arrived to the United States when they were younger than 25 years were more likely to be physically active than women who arrived when they were 25 years or older. Length of U.S. residence not associated with physical activity.
Ghaddar, Brown, Pagán, and Díaz (2010)	Cross-sectional; survey data from Alliance for a Healthy Border	Predominantly Hispanic adults ages ≥18 years residing in largely Hispanic communities along the border between United States and Mexico N = 2,381 (95.3% Hispanic; 78.8% female)	Country of birth Short Acculturation Scale for Hispanics (Marín et al., 1987)	Questions on PA from BRFSS	Individuals with low language acculturation were less likely to engage in physical activity than those with moderate to high acculturation (OR = 0.75; 95% Cl = 0.59–0.95).
Gollenberg et al. (2008) b	Prospective cohort study; nonexperimental; assessments at 2 time points for subgroup of participants	Pregnant Latina women ages 16–40 years receiving prenatal care at a large tertiary care facility in Western Massachusetts N = 1,231 (100% female)	Language preference for speaking, reading, and writing along with place of birth (United States compared with Puerto Rico or foreign born)	Modified version of the KPAS (Sternfeld, Ainsworth, & Quesenberry, 1999)	Participants who preferred Spanish were less likely to meet the physical activity guidelines (OR = 0.6; 95% Cl = 0.3–1.0) compared with those who preferred English.
Guinn, Vincent, Wang, and Villas (2011)	Cross-sectional; self-reported survey	Latino parents and adult relatives ages 20–68 years of elementary school children in U.S./Mexico border communities in Texas with high Latino populations N = 438 (75% female)	Marín and VanOss Marin (1991) language-use acculturation scale	Single question on number of weekly 10-minute bouts moderate-to-vigorous LTPA, based on 2009 National Center for health Statistics method for appraising LTPA	No significant differences in PA found between those who were more acculturated or those who were less acculturated.
Ham, Yore, Kruger, Heath, and Moeti (2007)	Secondary data analysis from 4 national surveillance systems: (1) NHANES, 1999–2002; (2) the BRFSS, 2003; (3) the NHTS, 2001; and (4) the NHIS Cancer Supplement, 2000	Latino adults ≥18 years, noninstitutionalized U.S. civilian population N = 29,361 (57% female)	Varied among the surveys: NHANES and NHTS—country of birth; NHIS—asked for origin or ancestry; NHTS—computed the number of years they had been in the country and their age at arrival	Varied among the surveys: NHANES—LTPA and household PA participation in an activity on at least four occasions during the previous 30 days; BRFSS—respondents were asked whether their activities at work involved mostly sitting or standing, mostly walking, or mostly heavy labor or physically demanding work; NHTS—defined participation in transportation- related PA as either using public transportation at least once per week or walking or bicycling for transportation for at least 10 minutes on a randomly selected day; NHIS—active level of usual daily activity was defined as “usually walking around most of the day” and “usually lifting or carrying moderate or heavy loads” and included all activities except those performed during leisure time.	The prevalence of transportation-related PA declined with increased time spent in the United States. LTPA and household PA increased nonsignificantly with acculturation. The prevalence of an active level of usual daily activity among Latinos in the NHIS did not appear to be associated with their acculturation.
Himmelgreen et al. (2007)	Mixed methods; cross-sectional self-report surveys; qualitative interviews	Latino adults, recently arrived (in United States <2.5 years) N = 21 (n = 10 interviews; 83% female)	Demographic surveys included questions on time in United States, language, lifestyle since arrival in United States; participants were asked open-ended questions about their lives in their native countries (premigration) and in the United States (postmigration)	Quantitative and qualitative data on behavioral change prior to and after immigration were collected.	Quantitative: All participants reported changes in PA since leaving their native countries; 87.5% reported that since arrival to the United States, they had become much more sedentary since arriving to the United States; and 60% reported increased time watching television Qualitative: 80% of participants made reference to the increased pace of life in United States and the effect it had on reducing PA, “I don’t do exercise. I used to exercise I hour a day, and here I don’t have time for that.”
Jurkowski, Mosquera, and Ramos (2010)	Cross-sectional; self-report data collected using 70-item survey created for this study	Latina community members residing in Northeastern New York N = 289 (100% female)	Modified Short Acculturation Scale (Marín et al., 1987) Length of residence	Multipart question from BRFSS	Increasing acculturation associated with greater odds (OR = 1.08; 95% Cl = 1.02–1.13 ,p = .004) of performing PA 2 or more times per week.
Mainous et al. (2008)	Cross-sectional; data from 1999 to 2004 National Health and Nutrition Survey (NHANES)	Latino adults ages ≥20 years previously diagnosed with diabetes, nationally representative sample of noninstitutionalized U.S. population N = 467 (percentage of female not specified)	Country of birth Short Acculturation Scale for Hispanics (Marín et al., 1987)	MVPA (response categorized dichotomously—getting exercise or not getting exercise)	Individual who were more acculturated were more likely to report leisure-time exercise (59.2% vs. 19.3%, p < .001). However, the relationship was not significant after controlling for demographic characteristics.
Marquez and McAuley (2006)	Cross-sectional; self-reported questionnaires and 7 days accelerometer wear	Latino adults 18–60 years, mean age 29.4 years (SD = 7.9), recruited in various town in central Illinois N = 155 (55% female)	12-item Short Acculturation Scale for Hispanics (Marín et al., 1987)	EPAC2 (Wareham et al., 2002) Actigraph accelerometer	More acculturated participants reported less occupational activity (r = −.21) and less overall activity (r = −.23). Correlations between acculturation and household MET hour per week, recreational MET hour per week, and average daily activity counts were not significant. Latina women reported engaging in higher levels of household/domestic activity than Latino men.
J. L. Martinez, Latimer, Rivers, and Salovey (2012)	Qualitative focus groups and key informant interviews	Residents and service providers from a medium-sized, urban community in the northeastern United States Hispanic adults n = 31, mean age = 40.39 years (SD = 10.01; 59% female)	Focus group questions not specified	IPAQ Short form (Craig et al., 2003)	Participants discussed how assimilating to the American lifestyle has brought about negative changes to LTPA habits; participants also discussed changes in LTPA habits after arriving in the United States and expressed concerns that acculturation leads to a more sedentary lifestyle.
S. M. Martinez et al. (2011)	Cross-sectional; telephone survey	Mexican origin Latino adults living in San Diego County N = 672, mean age 39 years (SD = 13; 71 % female)	Country of birth Years living in United States Short Acculturation Scale for Hispanics	IPAQ (Craig et al., 2003)	Only active transportation, NLTW examined in this study. All proxies of acculturation (country of birth, years in United States, and Short Acculturation Scale for Hispanics) were negatively associated with NLTPA; more acculturated participants—U.S. born, living in United States ≥12 years, and those with high acculturation score—were less likely than those with low acculturation to engage in NLTW at recommended levels.
S. M. Martinez et al. (2012)	Cross-sectional; telephone survey	Mexican-American adults living in a U.S.–Mexican border community in San Diego, California N = 672, mean age 39 years (SD = 13; 71 % female)	Country of birth Years living in the United States Short Acculturation Scale for Hispanics (Marín et al., 1987)	IPAQ (Craig et al., 2003)	Language acculturation and LTPA did not show a significant association.
O’Brien, Davey, Alos, and Whitaker (2013)	Cross-sectional; used data from the 2009 CHIS	Adults ages ≥18 years, noninstitutionalized civilian population, excluding women who are pregnant or have a diagnosis of diabetes N = 21,112 (n = 4,321 Latinas; 100% female)	Six-item measure developed and validated for use with the CHIS (Johnson-Kozlow, 2010)	Participants were asked 2 questions about the total number of times per week that they walked for at least 10 minutes for any purpose, and 2 questions on MVPA (other than walking) they participated in for at least 10 minutes	Used binary variables to categorize respondents (least healthy tertile vs. other 2 tertiles). Those with lower levels of acculturation were less likely to be in the unhealthy tertile for walking.
Parra-Medina and Messias (2011)	Qualitative; semistructured interviews and focus groups conducted in 2 geographic locations (Texas and South Carolina)	Latinas, Mexican-origin women residing in either an established Latino settlement area on the U.S.–Mexico border, or an emerging settlement area in South Carolina with a high growth rate of Latino population N = 69 (100% female)	Focus group questions about acculturation not specified	Questions about PA not specified but focus at community level was on community health, social, and economic status and identifying community assets, resources, barriers, and problems related to the ability of individuals and groups becoming more physically active in the community. At the individual level, focus on understanding experiences, opinions, values, and perceived needs and preferences related to PA among Latino women in the community.	Recent immigrants identified how migration had affected their level of regular walking. Immigrant women reported that before coming to the United States, they regularly walked as part of their daily activities (e.g., going to market, church, visiting family). In contrast, they rarely walked in conjunction with these activities in the United States, due to the built environments in the communities where they resided.
Pichon et al. (2007)	Cross-sectional; used baseline data from a randomized community study	Latina adults 21–74 years (median age = 33 years) who are parents and caregivers of children enrolled in a RCT for prevention of childhood obesity in San Diego County. N = 526 (100% female)	30-Item ARSMA (Cuéllar, Arnold, & Maldonado, 1995)	Short version of the IPAQ	Stronger identification with the Mexican culture was associated with less vigorous physical activity, whereas stronger identification with the Anglo culture was associated with both vigorous and moderate levels of physical activity.
Slattery et al. (2006)	Cross-sectional; used data from controls participating in the 4-Corner’s Study of breast cancer	Women in Arizona, Colorado, New Mexico, and Utah selected as controls in a breast cancer study N = 2,039 (Hispanic n = 719, non-Hispanic n = 1,320; 100% female)	Language acculturation was used as an indicator of acculturation and was based on preference for speaking and reading Spanish	Questionnaire adapted from the Cross-Cultural Activity Participation Study (CAPS) Questionnaire (Ainsworth et al., 1999); additional questions about the amount of moderate and intense activity performed over one’s lifetime were asked.	Hispanic women with high language acculturation differed from non-Hispanic women in several domains, including housework, dependent care, and dancing. Hispanic women with lower language acculturation had significantly lower MET-hours of total and vigorous PA compared with Hispanic women with higher language acculturation.
Sussner et al. (2008)	Qualitative study design; 6 focus groups and 20 in-depth interviews	Latina mothers residing in the greater Boston metropolitan area who speak Spanish as primary language and who delivered a live-born baby within last 48 months N = 31 (100% female)	12-item Acculturation Scale (Marín & Gamba, 1996; Marín et al., 1987)	Prompting questions used in focus groups included questions on PA/exercise, such as “How do you compare the exercise you did in your native country to what you do now in the United States?” “How important is exercise in the Latino culture?”	Participants reported dramatic changes in PA and sedentary lifestyles between their native countries and the United States, with more PA promoted in their native countries. Use of more active transportation in their countries was noted (e.g., riding a bicycle or walking in their native countries, as opposed to spending time driving or travelling in buses and trains).
Tovar, Chasan-Taber, Bermudez, Hyatt, and Must (2012)	Single-group prospective cohort study	Pregnant Puerto Rican or Dominican heritage (Caribbean Islanders) women ages 16–40 years recruited from a tertiary care facility in Western Massachusetts N = 952 (100% female)	PAS (Tropp, Erkut, Coll, Alarcon, & García, 1999)	PPAQ (Chasan-Taber et al., 2007)	Physical activity level did not differ across those with low to high acculturation as measured by the PAS.
Van Rompay et al. (2012)	Cross-sectional data from Boston Puerto Rican Health Study (a longitudinal study of Puerto Rican adults)	Puerto Rican adults 45–75 years residing in high Hispanic-density neighborhoods in the Boston area N = 1,219 (72% female)	PAS (Tropp et al., 1999) Bidimensional Acculturation Scale for Hispanics (Marín & Gamba, 1996) Length of stay in the United States	Paffenbarger Harvard Alumni Questionnaire (Paffenbarger et al., 1993)	Puerto Rican adults with greater language acculturation were more likely to be physically active.
Van Wieren, Roberts, Arellano, Feller, and Diaz (2011)	Cross-sectional; self-reported telephone data from the combined 2005 and 2007 CHIS	Noninstitutionalized Latino male and female adults in California households N = 16,000	Six-item scale adapted from a 7-item acculturation scale developed and validated by Johnson-Kozlow (2010); self-reported country or origin	Minutes per week of MVPA converted to weighted METs then categorized to meeting or not meeting 2007 ACSM recommendations of 450-MET-min per week for aerobic PA	Among all Latinos, increased acculturation was associated with an increase in LTPA (p = .0001). Adjusted ORs of meeting ACSM physical activity recommendations were 0.74 (0.66–0.83), 0.99 (0.91–1.07), and 1.08 (0.99–1.18) for the low, moderate, and high acculturation groups, respectively. Increased acculturation was associated with increased odds of meeting ACSM recommendations among Mexicans (p = .0043) and Guatemalans (p = .0183). No associations were found between acculturation and physical activity among Salvadorans, Central Americans, South Americans, and Puerto Ricans.
Vella, Ontiveros, Zubia, and Bader (2011)	Cross-sectional; nonexperi mental	Hispanic women 20–39 years residing in the largest border community in the United States N = 60 (100% female)	Short Acculturation Scale for Hispanics (Marín et al., 1987)	Baecke Questionnaire of Habitual Physical Activity (Baecke, Burema, & Frijters, 1982); ActiGraph GTIM accelerometer worn for 4 days	There were no differences in PA between participants with higher and lower acculturation levels.
Vermeesch and Stommel (2014)	Cross-sectional; secondary data from the 2008–2011 NHIS	Civilian noninstitutionalized Latina adults 18–47 years in U.S. households in 50 states and the District of Columbia N = 7,278 (100% female)	Acculturation measure created based on language and place of birth (U.S. or foreign born)	Five questions about LTPA based on the 2008 Physical Activity Guidelines (U.S. Department of Health and Human Services, 2000)	Less acculturated Latinas (Spanish-preferring) were less likely to engage in PA than English-preferring counterparts (OR = 0.57, p < .01). Spanish-preferring foreign-born Latinas have substantially smaller odds of meeting PA guidelines than U.S.-born English-preferring Latinas (OR = 0.3, p < .001).

Note. PA = physical activity; SES = socioeconomic status; LTPA = leisure-time PA; NLTPA = nonleisure-time PA; ARSMA = Acculturation Rating Scale for Mexican American; IPAQ = International Physical Activity Questionnaire; BRFSS = Behavioral Risk Factor Surveillance System; RCT = randomized control trial; KPAS = Kaiser Physical Activity Survey; DRS = Developmental Research Sequence; NHIS = National Health Interview Survey; PR = prevalence ratio; NHANES = National Health and Nutrition Examination Survey; NHTS = National Household Travel Survey; EPAQ = EPIC Physical Activity Questionnaire; MET = metabolic equivalent; NLTW = nonleisure-time walking; CHIS = California Health Interview Survey; MVPA = moderate-to-vigorous PA; PAS = Psychological Acculturation Scale; PPAQ = Pregnancy Physical Activity Questionnaire; ACSM = American College of Sports Medicine.

^aSame study sample as Gollenberg et al. (2008).

^bSame study sample as Chasan-Taber et al. (2007).

Assessment of Physical Activity

Physical activity was assessed across studies using a wide range of measures. The majority of studies (n = 31, 94%) relied on the use of self-reported PA, while two (Marquez & McAuley, 2006; Vella et al., 2011) included the use of both subjective (self-report) and objective (accelerometer) measures to assess PA. Six studies (18%) reported using the International Physical Activity Questionnaire (IPAQ) or short version of the IPAQ (Craig et al., 2003), six studies (18%) used either a single question or multiple questions from the Behavioral Risk Factor Surveillance System (Centers for Disease Control and Prevention, 2005) or other national surveillance systems, while the EPIC Physical Activity Questionnaire (Wareham et al., 2002) and the Paffenbarger Harvard Alumni Questionnaire (Paffenbarger et al., 1993) were each used in one (3%) study. Other measures of PA were the Baecke Questionnaire of Habitual Physical Activity (Baecke et al., 1982), a modified version of the Kaiser Physical Activity Survey (Sternfeld et al., 1999) use of a single question to assess PA readiness, questions about leisure-time PA (LTPA) based on the Department of Health and Human Service 2008 Physical Activity Guidelines, and questions regarding number of times weekly of walking and moderate and vigorous activities. Table 1 illustrates measures of PA and acculturation.

Assessment of Acculturation

Similar to PA, acculturation measures across studies varied broadly. The most commonly used measure of acculturation was Marín et al.’s (1987) and Marín and Gambas (1996) acculturation scale or modified version of this scale (n = 11, 33%). One study used Marín and VanOss Marin’s (1991) language use acculturation scale (n = 1, 3%). Two studies (6%) used the Acculturation Rating Scale for Mexican Americans (ARSMA-II) or short form of the ARSMA-II (n = 1, 3%; Cuéllar, Arnold, & Maldonado, 1995). Thirteen (39%) of the 33 studies in this review assessed acculturation using a single measure of acculturation, while others included the use of two or more scales and/or proxy measures of acculturation. Proxy measures of acculturation most commonly used were birthplace (n = 5, 20%), length of residence in the United States (n = 4, 16%), and generation (n = 3).

Association of PA and Acculturation

Findings on the association of PA and acculturation indicated mixed results in quantitative studies with 14 studies (42%) reporting that higher acculturation was significantly associated with higher levels of PA (Abraído-Lanza et al., 2005; Bungum et al., 2011; Echeverría et al., 2013; Evenson et al., 2004; Jurkowski et al., 2010; Mainous et al., 2008; van Rompay et al., 2012; Van Wieren et al., 2011) or that low acculturation was associated with lower levels of PA (Ghaddar et al., 2010; Gollenberg et al., 2008; O’Brien et al., 2013; Pichon et al., 2007; Slattery et al., 2006; Vermeesch & Stommel, 2014). Two studies found a negative association between acculturation and PA (Barrera et al., 2012; S. M. Martinez et al., 2011) while five (24%) found no association (Banna et al., 2012; Guinn et al., 2011; S. M. Martinez et al., 2012; Tovar et al., 2012; Vella et al., 2011). Other studies reported that the relationship between acculturation and PA varied based on PA domain (LTPA vs. non-leisure-time PA [NLTPA]; Afable-Munsuz et al., 2010; Berrigan et al., 2006; Creighton et al., 2012; Ham et al., 2007; Marquez & McAuley, 2006) or generation (Afable-Munsuz et al., 2010; Creighton et al., 2012). Among the studies that examined PA by domain, the association between LTPA varied. In several studies, a higher level of acculturation was associated with increased LTPA (Afable-Munsuz, et al., 2010; Berrigan et al., 2006), while in others higher acculturation was associated with less overall activity (Marquez & McAuley, 2006), or the association varied by type of acculturation (social vs. linguistic acculturation; Creighton et al., 2012). In studies that examined the association of NLTPA, findings consistently indicated that as acculturation increased, NLTPA decreased (Afable-Munsuz et al., 2010; Berrigan et al., 2006; Ham et al., 2007; Marquez & McAuley, 2006; S. M. Martinez et al., 2011). In all five studies that used qualitative or mixed method designs, participants reported consistent themes regarding decreases in PA and increases in unhealthy behaviors since migrating to the United States

Discussion

This article reports the results of an integrative review of 33 studies that examined the association of acculturation and PA in Latina adults in the United States. The results of this review of literature indicate that there are significant gaps and inconsistent findings in the published literature on acculturation as a factor that influences PA in Latinas. Findings on the association of acculturation and PA varied greatly across quantitative studies; yet, consistent themes clearly emerged in all qualitative and mixed methods studies about participants being more sedentary and leading more unhealthy lifestyles after living in the United States than they did in their native countries. Similarly to findings from qualitative studies, among the few quantitative studies that examined PA by domain, results consistently indicated that as acculturation increased, the prevalence of NLTPA (e.g., walking for transportation and bicycling) decreased. These findings reflect continued presence of acculturation as a variable in PA behavioral research. However, the significant variability in the conceptualization and measurement of this variable suggests that the reported research does not specify the elements of acculturation (e.g., years in country, family values, and traditions) that might influence PA.

We found that most of the quantitative studies included in our review of literature relied on the use of cross-sectional data to examine the relationship between acculturation and PA, despite the commonly accepted definition of acculturation as a process that occurs over time (Berry, 2003). Although our findings are similar to those reported in previous studies that examined the association of acculturation and health behaviors (Delavari et al., 2013; Yeh, Viladrich, Bruning, & Roye, 2009), the use of cross-sectional data to examine how acculturation influences PA raises several conceptual and methodological concerns.

First, since acculturation is a process, rather than a single event or point in time, there is an assumption that a change in behaviors and/or values has occurred; thus, measurement of behavior change should occur at a starting point and at the end or evaluation point. In other words, changes over time should be documented. For example, Himmelgreen et al.’s (2007) study on dietary acculturation in Latinos included survey questions that asked participants about changes that have occurred in their diets, food habits, and lifestyles since migrating to the United States. Interviews were also conducted about participants’ lives in the native countries and following migration to the United States; these interviews used probes to explore pre- and postmigration changes in participants’ lives that influence their food choices and habits (Himmelgreen et al., 2007). While this study examined changes in diet and food habits over time associated with migration, similar approaches could be used to examine pre- and postmigration changes in PA and exercise behaviors. Another issue regarding the use of cross-sectional measures of acculturation is that they could be subject to variability based on factors such as the respondent’s location, responsibilities, or time of interview that would not accurately reflect their acculturation status when it is measured at a single point in time. For example, a Latina college student responding to acculturation questions (e.g., preferred language, language of television shows, food preference, or friends) when he or she is surrounded by mostly English-speaking friends or working in a primarily English-speaking setting, might be categorized as highly acculturated. This same person could then retake the same acculturation measure days later while spending time with relatives who are primarily Spanish-speaking; in this context, acculturation might be much lower than the previous assessment. The point here is that our social environment influences how we act, with whom we identify, and what we report on questionnaires. Assessing acculturation at a single point in time may not be reflective of actual behaviors.

Many of the measures or proxies of acculturation used in cross-sectional studies may overlook the deeper structural issues that are important in understanding the influences that culture has on health behavior and ultimately whether or not traditions can be used as mechanisms to promote behavioral change in culturally appropriate PA interventions. The studies in this review provide preliminary support to indicate that acculturation, as measured in the studies we reviewed, may not have as much of an influence on health PA as do other factors (e.g., life patterns, traditions, family, and gender roles). In fact, the context of Latina’s lives is a critical factor for engaging in PA and understanding the relationship between PA and acculturation—it was often not addressed in quantitative studies reviewed. This finding is similar to Hunt et al.’s (2004) review, which reported that the discussion of acculturation in health research was focused almost entirely on issues of measurement, and it often neglected the most important conceptual issue about what is actually being measured.

Castro (2007) suggested that scientific answers to the questions can help understand the health effects of acculturation: What constitutes healthy versus unhealthy acculturation, for whom, and under what circumstances (Castro, 2007)? Our review of literature on acculturation and PA in Latinas found that these questions have yet to be answered. Understanding contextual factors might be more informative than a proxy or a measure of acculturation. Although the mechanism through which these factors influence health is not fully understood, researchers in qualitative studies have begun to explore the influence of contextual factors (e.g., life patterns, the culture and changes in lifestyles following migration) on health status. For example, previous studies have shown that Latinas were less likely to work outside of their homes in their native countries, and their primary responsibilities revolved around caring for their families, which included cooking and walking to the market daily to buy food (D’Alonzo, 2012; Sussner et al., 2008). On migration to the United States, participants reported that with the rapid pace of life and need for employment outside of the home, the lack of time, and the distances to the market, they were no longer able to shop daily and had to depended on cars or public transportation to get to the market (D’Alonzo, 2012; Sussner et al., 2008).

Other lifestyle changes associated with migration or acculturation that also can influence performance of PA are the environment where Latinas reside in the United States (Larsen, Pekmezi, Marquez, Benitez, & Marcus, 2013). Parra-Medina and Messias (2011) found that when individuals move from a rural, farming area to a city, the type and frequency of PA is likely to change. Latina immigrants have also reported that prior to coming to the United States, their daily activities involved regular walking (going to market, church, and visiting family), but due to factors such as transportation and safety concerns, they rarely walked in conjunction with these activities in the United States (Parra-Medina & Messias, 2011). These changes may be influenced in part by their neighborhoods, as immigrants and ethnic minorities often live in communities with high proportions of residents of the same ethnicities (Logan, 2003; Suro & Tafoya, 2004). These neighborhoods can offer cultural goods and social networks for non-English speakers (Fernandez Kelly & Schauffler, 1996), but they may also be more likely to be in communities with higher crime and less resources for being physically active since Latinos in the United States report lower household incomes and higher likelihood of living in poverty than non-Hispanic Whites (U.S. Census Bureau, 2012).

Several methodological issues also emerged from this review of literature. Given the inconsistency in measurement of both acculturation and PA among reviewed studies, it is not surprising that marked difference were found across studies. This inconsistency in measurement and conceptualization of acculturation across studies in our review is similar to findings reported in previous literature reviews of studies that examined the association of acculturation and health behaviors (Hunt et al., 2004; Thomson & Hoffman-Goetz, 2009). Some studies used validated measures of acculturation, while others adapted measures that were used in other studies or developed their own tools; however, most researchers did not report reliability and validity for these adapted measures or explain how their measures were adapted to be culturally relevant to Latinos.

A number of studies in this review reported using one or more proxies to measure this variable. Proxy measures of acculturation included birthplace, country of origin, spoken language, length of residence in United States, generation, age at arrival, and language preference for speaking, reading, and writing. This lack of conceptualization and consistent measurement prevents accurate examination of how or whether acculturation might influence health status. Proxy measures of acculturation are neither indicative of a process of change nor predictive of health behaviors. Asking about birthplace, for example, does not provide information about length of time spent in their native countries—whether or not a participant left shortly after birth, was born there during his parent’s short-term residency in that place as they were migrating to the United States, or was born there to foreign residents. It does not imply that any particular cultural practices would have been adopted or used. The lack of consistent reporting of reliability and validity of acculturation measures found in our review of literature further contributes to the inconsistent findings and brings into question the effectiveness of using this concept as a variable in health disparities research with Latinas.

Rather than quantifying an individual’s level of acculturation, a more appropriate measure might involve focusing on how well individuals and family members who have immigrated to a new place have accommodated themselves to their new surroundings; that is, whether or not they have been able to incorporate traditions and practices into their lives in the new place. In Spanish, the word we might use to describe this process is acomodar—to make oneself comfortable. This would mean that we would devise questions that allow us to identify, often in a very positive sense, how and if people have been able to bring comfortable elements of their lives and traditions into the new environment.

Assessment of PA also varied broadly among the studies, with the majority relying on self-reported data and only two studies using objective measures (accelerometers) to assess PA, indicating a lack of consistency across studies regarding the frequency, duration, and type of PA assessed (intensity, leisure time, nonleisure time). For instance, when the type of PA was assessed, it was done using five questions about LTPA based on the 2008 Physical Activity Guidelines (U.S. Department of Health and Human Services, 2000), a single question based on the National Center for Health Statistics 2009 on number of times per week of moderate to vigorous intensity PA for 10 minutes or more, or by using measures with established reliability and validity (e.g., the IPAQ or EPIC Physical Activity Questionnaire). Studies varied in the assessment of type of PA (leisure time vs. nonleisure time). Assessing both LTPA and NLTPA is important as women tend to engage in activities that are more difficult to quantify, such as child care and household activities (Kriska & Caspersen, 1997). It is possible that by not assessing both LTPA and NLTPA or by not using objective measures of PA, PA was not adequately assessed. Furthermore, authors did not consistently report whether or not PA measures were tested for reliability and validity in this population or at all. To understand and come to any conclusions about how acculturation might influence performance of PA in Latinas, it is critical that frequency and duration of PA is consistently assessed. To examine the impact of PA on health outcomes, accurate measurement of PA is necessary. Therefore, a recommendation would be that future studies should include the use of both objective and subjective measures of PA with established reliability and validity.

Limitations

There were several limitations to the current review of literature. Only articles published in the past 10 years and indexed in any of the five databases (PubMed, CINAHL, PsycInfo, Web of Science, and JSTOR) were included in this review; therefore, it is possible that we may have missed relevant articles published on this topic. The lack of intervention studies located for this review may be due in part to the search criteria used; however, a more uniform definition of the concept of acculturation and further understanding of its role in health behaviors is necessary to use in leveraging health promotion in Latinas. Next, of the 33 articles included in this integrative review of literature, the majority (n = 24, 73%) reported cross-sectional data on the association of acculturation and PA, which may not provide an accurate representation of actual PA behaviors. Last, Latinos in the United States come from many different countries and ancestries and comprise a highly heterogeneous group. The Latina population in the United States varies greatly in terms of sociodemographic factors; therefore, findings on the relationship between acculturation and PA from one Latino subgroup may not be generalizable to others. Thus, future studies should examine factors associated with acculturation and PA in specified subgroups of Latinas.

Conclusion

Following decades of research about acculturation, this construct remains a vaguely conceptualized one, and the mechanism through which it affects health outcomes has not been clearly identified. While some discussions are built on the assumption that immigrant acculturation is a measureable mechanism that underlies health inequalities, the research to date has neither been based on consistent and measureable definitions of the term nor can it fully explicate the context of the phenomenon or mechanisms that are hypothesized to produce an effect—poorer or better health (Wallace, Pomery, Latimer, Martinez, & Salovey, 2010). Further research is necessary to standardize the definition of acculturation and to determine how it might best be measured in research, or address the question of whether this concept should even be used as a variable in health research, as was questioned by Hunt et al. (2004).