Social Inequalities in Multiple Chronic Disease Risk Factors among a population-based Sample of Latinas in California

Abstract

Data from 1,248 Latina mothers who participated in the Geographic Research on Wellbeing (GROW) study were used to examine associations between SES, neighborhood-level Latinx concentration, neighborhood-level poverty and having two or more modifiable behavioral risk factors (e.g., smoking, drinking) for chronic disease. Logistic regression models were estimated stratified by nativity and adjusted for age and marital status. Among immigrants, low SES was associated with higher odds of multiple risk factors (Adjusted Odds Ratio [AOR] = 1.66, 95% Confidence Interval [CI] = 1.17–2.38). Among US-born women, low neighborhood-level Latinx concentration was associated with lower odds of multiple risk factors (AOR = 0.43, 95% CI = 0.22–0.84), and high neighborhood-level poverty (AOR = 2.83, 95% CI = 1.61–4.99) and low SES (AOR = 1.72, 95% CI = 1.02–2.92) were associated with higher odds, respectively. Heterogeneous effects between nativity and social factors may produce risk for chronic disease among Latinas.

Introduction

Immigrant Latinas, compared to US-born Latinas, engage in different modifiable behavioral risk factors that may protect or put them at risk of chronic disease. Chronic disease can be defined as a persistent and recurring health problem, such as obesity and diabetes, with a duration measured in months and years, not days and weeks (Thrall, 2005). Latinas are about even (~11%) with Black women for the second highest prevalence rates of diabetes behind American Indian/Alaskan Natives (~14%) and have the second highest (~44%) prevalence rates of obesity behind Black women (~57%) (Centers for Disease Control and Prevention [CDC], 2020a; Fryar, Carroll, & Afful, 2020). Modifiable behavioral risk factors, including not meeting daily physical activity, less than daily fruit and vegetable intake, smoking, and drinking at risk levels, have all been associated with chronic disease, such as diabetes and obesity (Durazo, Mbassa, & Albert, 2016; Lui & Zamboanga, 2018). Depending upon the health behavior, more–compared with less–acculturation can confer advantage or risk for Latinas. For instance, immigrant Latinas tend to engage in lower levels of physical activity, while US-born Latinas tend to have higher rates of smoking (Jurkowski, Mosquera, & Ramos, 2010; Kondo, Rossi, Schwartz, Zamboanga, & Scalf, 2015; Pulvers et al., 2018; Vermeesch & Stommel, 2014). However, less is known about how neighborhood poverty and neighborhood racial/ethnic concentration contribute to differences in modifiable behavioral risk factors for chronic disease between US-born and immigrant Latinas. The scant evidence focused on neighborhood poverty suggests an increased risk of chronic disease and engagement in modifiable behavioral risk factors among Latinas living in neighborhoods with a higher concentration of individuals living in poverty (Chamberlain et al., 2020; Wallace et al., 2021). Evidence on neighborhood racial/ethnic concentration and chronic disease-related health has had mixed findings, with both benefits and disadvantages proposed (Kershaw & Albrecht, 2015; Viruell-Fuentes, Ponce, & Alegría, 2012).

The purpose of the current study was to systematically examine how various sociodemographic factors, including socioeconomic status (SES), language, nativity, neighborhood ethnic density, and neighborhood poverty, impact modifiable behavioral risk factors for chronic disease among Latinas. The modifiable behavioral risk factors of interest include daily physical activity, fruit and vegetable intake, smoking, and drinking at risk levels, due to their established association with chronic disease (Durazo et al., 2016; Lui & Zamboanga, 2018). The research questions centered around associations between neighborhood poverty and neighborhood Hispanic/Latinx concentration and the four risk factors for chronic disease, among Latinas, stratified by nativity. Understanding the social factors contributing to the risk of poor health is important for social workers who partner with other disciplines to intervene and address modifiable behavioral risk factors that can lead to chronic disease among Latinas.

Linking acculturation, nativity, and language to risk factors

The definition of acculturation and its measurement have continued to evolve as more and more nuances are discovered. A commonly accepted definition come from John Berry (2003), who developed an acculturation framework delineating two separate processes: maintenance of the original culture and development of relationships with the new culture. Due to evidence-based acculturation scales not always being available to researchers in datasets, the literature examining the relationship between modifiable behavioral risk factors and acculturation utilizes several proxies such as language use and nativity, both utilized in the current study. Though acculturation scales are preferred, proxies such as language and nativity have proved to be appropriate alternatives. For instance, Lui and Zamboanga’s (2018) meta-analysis examining the extent to which acculturation is associated with alcohol use outcomes found positive relations between measures of acculturation and several alcohol use outcomes, and the effects were relatively stronger among Hispanic women compared to non-Hispanics and men. Lui and Zamboanga’s study examined various measures of acculturation including language use, language preference, and nativity, which were positively correlated with alcohol use status, drinking intensity, and binge drinking (Lui & Zamboanga, 2018). Kondo et al.’s (2015) meta-analysis examining the influence of acculturation on smoking among Latinas found that measures of language use and preference and nativity were consistently positively associated with higher odds of both lifetime and current smoking. Nativity, being US-born, had a particularly strong association with odds of smoking in one’s lifetime (Kondo et al., 2015). It must be acknowledged that nativity and language are crude measures of acculturation and, in theory, we cannot assume that being a primarily English-speaker or being US-born means that one is fully acculturated; however, the literature provides evidence for the practical use of these as proxies.

Two studies using the Acculturation Rating Scale for Mexican Americans, found that, among Latinas, being more acculturated was associated with lower consumption of fruits and vegetables (Gregory-Mercado et al., 2006; Kasirye et al., 2005). Additionally, one study found that US-born and English speaking women consumed fewer fruits and vegetables than Mexican-born women (Montez & Eschbach, 2008); while another found that Mexican-born women consumed significantly more fruits, grains and dairy products per day than US-born women (Harley, Eskenazi, & Block, 2005). Acculturation appears to confer risk when it comes to smoking, drinking, and fruit and vegetable intake; however, the opposite is true for meeting recommended levels of physical activity (Jurkowski et al., 2010; Vermeesch & Stommel, 2014). A systematic review of 33 studies on the relationship between acculturation-related variables and physical activity among women found that English language use and preference measures were consistently positively associated with physical activity, whereas nativity did not show a consistent pattern of associations across studies (Benitez et al., 2016). The literature helps provide evidence of the difference in behaviors among Latinas; however, the literature has only briefly examined acculturation and neighborhood factors together.

Linking neighborhood level factors to risk factors

Evidence on the relationship between the four risk factors and neighborhood level factors is lacking and is inconclusive, among Latina populations. The evidence that does exist often focuses on neighborhood poverty and neighborhood ethnic concentration (Cambron, Kosterman, & Hawkins, 2019; Kimbro, 2009). The literature surrounding these two factors are deeply intertwined with what is defined as an ethnic enclave. Ethnic enclaves are characterized by ethnically dense populations, low-resources, and high poverty (Ludwig et al., 2011; Osypuk, Diez Roux, Hadley, & Kandula, 2009). Due to the connection between neighborhood poverty and neighborhood ethnic density concentration, the current study focuses on literature related to how living in an ethnic enclave can have positive or negative effects for Latinas in the United States. For example, Latinx persons tend to have healthier diets prior to migrating to the United States and the presence of ethnic food stores, which can be found in ethnic enclaves, can help Latinx persons maintain those healthy eating patterns (Durazo et al., 2016; Osypuk et al., 2009). Yet, the relationship is complex and other evidence suggests Latinx density has been associated with unhealthy eating practices and poor physical activity (Durazo et al., 2016; Li, Wen, & Henry, 2017). Evidence also suggests that reduced levels of leisure-time physical activity may stem from the lack of accessible gyms in language of choice, green spaces, perceived safety, walkability in ethnic enclaves (Osypuk et al., 2009).

Other evidence suggests that, for immigrants, living in an ethnic enclave can provide critical financial support that in turn positively impacts physical health (Roy, Hughes, & Yoshikawa, 2013). This idea has been further strengthened by the idea that ethnic enclaves are typically spatially bounded from the main economy allowing it to function as its own economic center (Xie & Gough, 2011). Another study reported a disadvantage of living in an ethnic enclave is that access to high-quality care is often lacking or not in close proximity (Burner et al., 2019). How these possible protective factors impact health is still unclear. Cubbin, Hadden, and Winkleby’s (2001) study suggests strong associations exist between neighborhood poverty and poor health (i.e., being overweight) and modifiable behavioral risk factors (i.e., smoking). Among Hispanic/Latinx persons, there are mixed findings about the risk or protection from living in an ethnic enclave and engaging in problem drinking (Stroope, Martinez, Eschbach, Peek, & Markides, 2015). The inconclusiveness within the literature suggests the risk and protective factors of ethnic enclaves require further examination.

The current study

Research has shown nativity, individual SES, racial/ethnic density, and neighborhood poverty to be important correlates of individual aspects of physical health among Latinx persons. This study adds to the literature by examining several social factors and chronic disease risk factors (i.e., smoking, drinking, physical activity, and fruit/vegetable intake) in the same analysis, to show how these factors might impact a diverse sample of Latinas. California is home to the largest population of Latinx persons in the U.S., with approximately 15 million Latinx persons (Von Behren et al., 2018). With a representative population-based sample, California represents the ideal setting to evaluate the impact of nativity, SES, and neighborhood characteristics on chronic disease risk.

Methods

Data sources

Data for this study are from the Geographic Research on Wellbeing (GROW) study, a follow-up survey of the 2003–2007 Maternal and Infant Health Assessment (MIHA). MIHA is California’s version of CDC’s Pregnancy Risk Assessment Monitoring System (Centers for Disease Control and Prevention [CDC], 2020b) and is an annual, statewide-representative survey of mothers delivering live infants in California during February through May (Cubbin et al., 2002; Rinki, Martin, & Curtis, 2012). MIHA collected data from approximately 3,500 women per year representing approximately 500,000 births using a questionnaire that was administered by mail or phone, in English (71%) or Spanish (29%); response rates exceeded 70% each year. The maternal characteristics of the MIHA sample are weighted to be representative of all eligible births statewide (CDPH, n.d.) and data are routinely geocoded to census tracts via mothers’ residential addresses recorded on birth certificates.

MIHA respondents who agreed to be re-contacted for future studies (95%) were eligible for GROW if they lived in one of six largely urbanized counties at the time of the 2003–2007 surveys (Alameda, Los Angeles, Orange, Sacramento, San Diego, and Santa Clara), which was 55% of all MIHA respondents. Data collection for GROW took place from 2012–2013. The questionnaire comprised approximately 80 questions regarding demographic, socioeconomic, neighborhood, psychosocial, and health-related characteristics. They received a $20 gift card as an incentive and a chance at a raffle for $250. The incentive was enhanced toward the end of data collection to increase response. The GROW study was approved by the Institutional Review Boards at the University of Texas at Austin, the University of California, San Francisco, and the California Department of Public Health (Cubbin, 2015).

Of the 4,026 women who could be located (i.e., the “active” sample), 74.9% responded (N = 3,016). The vast majority of respondents (90.3%) still lived in one of the six GROW counties. Fifty-six percent completed the survey by phone vs. 44% by mail, and 73% completed it in English vs. 27% in Spanish. Missing values for income (9.8%) were imputed using hot-deck methodology with the following variables: age, race/ethnicity, education, employment status, marital status, and neighborhood poverty. Weights were created to produce data that were representative of births in the six GROW counties, and a sampling fraction file was created to make a minor finite population correction to the standard errors for analyses (Cubbin, 2015). The analytic dataset included women whose ethnicity was reported as Latina, lived in California, did not have missing income after imputation, and had an accurate geocode at the census tract level. This resulted in a final analytic sample of 1,248 records, which were subsequently linked to California census tract data from the 2005–2009 American Community Survey, so that the neighborhood variables were measured before the dependent variable.

Variables

Physical activity

Not meeting recommendations for physical activity was based on the Stanford Leisure-Time Activity Categorical Item (L-Cat) – a single item comprised of 6 clinically-relevant categories below, at, and above physical activity guidelines, which has been shown to have excellent reliability, validity, and sensitivity to change in a randomized behavioral weight-loss trial (Kiernan et al., 2013).

Fruit and vegetable consumption

Insufficient fruit and vegetable consumption was defined as consuming fruits and vegetables less often than daily was based on responses to the questions, “During an average week, how often do you eat . . . fruit, including 100% fruit juice,” or “ . . . vegetables, not including French fries.” Response choices were “never or almost never,” “about once or twice a month,” “about once or twice a week,” “about every other day,” and “every day.” Respondents not reporting daily consumption of both fruits and vegetables were coded as consuming fruits and vegetables less often than daily.

Smoking status

A current smoker was defined as having smoked at least 100 cigarettes in the respondent’s lifetime and reporting that she currently smokes some days or every day.

Binge drinking

Binge drinking was operationalized as responding “yes” to the question “During the past 30 days, did you drink 4 or more drinks with alcohol in one sitting (within about two hours)?” Immediately prior to the question, alcohol was defined as “any kind of drink with alcohol in it. A drink is one glass of wine, one wine cooler, one can or bottle of beer, one shot of liquor, or one mixed drink.”

Sociodemograhics

Sociodemograhics included in the current study included socioeconomic status (SES), age, marital status, and nativity. SES was defined as lower (annual family income less than or equal to 200% of the federal poverty level and an educational level less than or equal to a high school diploma/GED) or higher (annual family income more than 200% of the federal poverty level or some college or more education). Age was represented as a three-category variable, (20–29 years, 30–39 years, 40 years and older), as was marital status (previously or never married, married or living together). Nativity was defined as having been born in the U.S. vs. not being U.S. born.

Neighborhood variables

Two neighborhood-level variables were included in the current study. These were the proportion of Hispanic/Latinx persons, and the proportion of persons with income below the poverty level at the census tract level. Each were divided into empirical tertiles based on statewide distributions (Pearson correlation = 0.53).

Data analyses

The primary outcome of interest was odds of endorsing multiple behavioral health risk factors for chronic disease, as a measure of severity, defined as two or more of the following: 1) not meeting recommendations for physical activity, 2) not eating both fruit and vegetables daily, 3) status as a current smoker, and 4) any binge drinking in last 30 days. This was represented as a dichotomous variable wherein “1 = endorses 2 or more,” and “0 = endorses less than 2.” Covariates of interest included SES, neighborhood Hispanic/Latinx concentration, and neighborhood poverty concentration. Higher SES was the referent group for SES, the lowest tertile served as the referent group for neighborhood poverty concentration, and the highest tertile served as the referent group for neighborhood Hispanic/Latinx concentration.

All analyses were stratified by nativity, which was supported by significant interactions between nativity and neighborhood poverty concentration and between nativity and neighborhood Hispanic/Latinx concentration in preliminary analyses (there was not a significant interaction between nativity and SES, however). We first estimated descriptive statistics for the sample. Next, we estimated four logistic regression models for the two nativity groups, each adjusted for age and marital status. Multiple behavioral health risk status was regressed on: neighborhood Hispanic/Latinx concentration (Model 1); neighborhood Hispanic/Latinx concentration and SES (Model 2); neighborhood poverty concentration (Model 3), and; neighborhood poverty concentration and SES (Model 4). Because of the moderate correlation between neighborhood poverty concentration and neighborhood Hispanic/ Latinx concentration, we did not include models with both. Two sensitivity analyses were also conducted, first repeating the analyses stratified by language instead of nativity and second treating the dependent variable as a continuous variable. All analyses were conducted using SAS version 9.4 (SAS Institute Inc, 2013) and incorporated weighting and the complex sample design. We did not use formal multilevel modeling because the data were not highly clustered (90% of tracts contained only 1 or 2 GROW respondents) .

Results

Compared with immigrants, U.S.-born Latinas were younger, had higher SES, and lived in neighborhoods with lower concentrations of Hispanic/Latinx or poor persons, respectively (Table 1). Both groups had low levels of current smoking and binge drinking while also having high levels of less than daily consumption of fruits and vegetables and not meeting physical activity guidelines. Between group differences across the four risk factors were all statistically significant, except for less than daily fruit and vegetable consumption.

Table 1.

Sociodemographic characteristics, GROW, 2012–2013, N = 1,248.

	Latina, immigrant		Latina, US-born		Latina, Spanish		Latina, English
	(n = 809)		(n = 439)		(n = 764)		(n = 484)
	n	%	n	%	n	%	n	%
Maternal age*  20–29 years	144	21.1	133	37.1	146	22.1	131	33.6
30–39 years	426	54.4	222	48.3	400	54.0	248	49.8
40 years or older	239	24.5	84	14.6	218	24.0	105	16.6
Marital status  Married/living together	664	82.9	352	78.5	629	83.0	387	78.7
Single/Separated/Divorced	138	17.1	83	21.5	128	17.0	93	21.3
Socioeconomic status*  ≤200% FPL and ≤high school graduate	609	77.4	104	27.6	611	81.1	102	25.2
200+% FPL and/or some college or more	194	22.6	332	72.3	147	18.9	379	74.8
Neighborhood % Hispanic/Latinx*  Lowest tertile	33	3.5	60	10.6	23	2.7	70	11.6
Middle tertile	176	20.6	131	27.2	153	18.8	154	30.0
Highest tertile	600	75.9	248	62.2	588	78.5	260	58.4
Neighborhood % Poor*  Lowest tertile	82	9.2	128	25.6	65	7.7	145	26.9
Middle tertile	254	31.0	155	34.0	235	30.5	174	34.9
Highest tertile	473	59.8	156	40.4	464	61.8	165	38.3
Risk factors  Current Smoker*	13	1.5	25	6.2	12	1.5	26	5.8
Any Binge drinking*	39	4.4	36	8.3	32	3.9	43	9.0
Less than daily F/V consumption	550	68.1	299	69.2	527	69.2	322	66.8
Not meeting PA guidelines*	692	86.1	331	77.4	667	88.0	356	74.5
Number of risk factors (PA, diet, smoker, binge drinker)*  0	51	6.4	30	5.9	42	5.4	39	7.9
1	257	31.9	158	34.6	239	31.5	176	35.2
2	467	58.1	221	52.1	451	59.4	237	50.2
3	33	3.4	29	7.1	31	3.5	31	6.6
4	1	0.1	1	0.2	1	0.1	1	0.2

FPL = federal poverty level, PA = physical activity. F/V = fruit and vegetable.

^*chi-square test p-values <0.05

Table 2 presents the results of the logistic regression analyses. In Models 2 and 4 for immigrant Latinas, low SES was associated with 70% and 66% higher odds of having two or more risk factors, respectively, while the neighborhood variables were not significantly associated with multiple risk factor status. In a model adjusting for SES and both neighborhood variables, only SES remained significant (not shown, available upon request). Among U.S.-born Latinas, low SES was associated with 72% higher odds of having two or more risk factors, and the neighborhood variables were both significantly associated with having two or more risk factors. Specifically, living in a neighborhood with the lowest concentrations of Hispanic/Latinx people was associated with 57% lower odds of having two or more risk factors compared to living in a neighborhood with the highest concentrations. Living in a neighborhood with the highest concentrations of poor people was associated with nearly three times the odds of having two or more risk factors compared to living in a neighborhood with the lowest concentrations. In a model adjusting for SES and both neighborhood variables, only neighborhood poverty remained significant (not shown, available upon request).

Table 2.

Logistic regression models for more than one risk factor stratified by nativity, GROW, 2012–2013, N = 1,248.

	Latina, immigrant				Latina, US-born
	(n = 809)				(n = 439)
	Model 1	Model 2	Model 3	Model 4	Model 1	Model 2	Model 3	Model 4
	Adjusted Odds Ratio (95% Confidence Interval)				Adjusted Odds Ratio (95% Confidence Interval)
Maternal age
20–29 years	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
30–39 years	1.17 (0.78–1.77)	1.22 (0.80–1.84)	1.17 (0.77–1.76)	1.21 (0.80–1.83)	1.14(0.71–1.85)	1.25 (0.76–2.05)	1.26 (0.77–2.06)	1.43 (0.86–2.39)
40 years or older	0.93 (0.59–1.44)	0.98 (0.63–1.54)	0.94 (0.60–1.46)	0.99 (0.63–1.55)	1.26 (0.69–2.30)	1.39 (0.75–2.58)	1.40 (0.76–2.59)	1.60 (0.85–3.04)
Marital status
Married/living together	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Single/Separated/Divorced	1.45 (0.96–2.20)	1.45 (0.96–2.20)	1.44(0.95–2.17)	1.45 (0.96–2.20)	1.61 (0.93–2.78)	1.51 (0.87–2.63)	1.57 (0.91–2.70)	1.43 (0.82–2.50)
Neighborhood % Hispanic/Latinx
Lowest tertile	1.15 (0.50–2.63)	1.35 (0.59–3.08)			0.38 (0.20–0.73)	0.43 (0.22–0.84)
Middle tertile	1.07 (0.73–1.57)	1.11 (0.75–1.63)			0.69 (0.42–1.11)	0.74 (0.45–1.21)
Highest tertile	1.00	1.00			1.00	1.00
Neighborhood % Poor
Lowest tertile			1.00	1.00			1.00	1.00
Middle tertile			0.89(0.52–1.53)	0.75 (0.43–1.31)			1.52 (0.91–2.55)	1.47 (0.87–2.48)
Highest tertile			1.12 (0.67–1.87)	0.90 (0.53–1.53)			2.98 (1.72–5.18)	2.83 (1.61–4.99)
Socioeconomic status
≤200% FPL and ≤high		1.70 (1.19–2.42)		1.66 (1.17–2.38)		1.61 (0.94–2.76)		1.72 (1.02–2.92)
school graduate		1.00		1.00		1.00		1.00
200+% FPL and/or some college or more

FPL = federal poverty level.

Neighborhood tertiles based on statewide distribution of census tracts.

Sensitivity analyses, stratifying by language (Table 3) or using a continuous dependent variable (Table 4), showed similar results, except that unmarried women had higher odds compared to married women for both groups when modeling the dependent variable as a continuous variable.

Table 3.

Logistic regression models for more than one risk factor stratified by language, 2012–2013, N = 1,248.

	Latina, Spanish				Latina, English
	(n = 809)				(n = 439)
	Model 1	Model 2	Model 3	Model 4	Model 1	Model 2	Model 3	Model 4
	Adjusted Odds Ratio (95% Confidence Interval)				Adjusted Odds Ratio (95% Confidence Interval)
Maternal age
20–29 years	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
30–39 years	1.14 (0.75–1.73)	1.17 (0.77–1.78)	1.14(0.75–1.72)	1.16 (0.77–1.77)	1.10 (0.68–1.77)	1.18 (0.73–1.92)	1.18 (0.73–1.92)	1.29 (0.78–2.14)
40 years or older	0.88 (0.56–1.39)	0.92 (0.58–1.45)	0.89 (0.57–1.41)	0.92 (0.58–1.46)	1.14 (0.65–2.01)	1.25 (0.70–2.22)	1.23 (0.70–2.16)	1.36 (0.76–2.45)
Marital status
Married/living together	1.00	1.00	1.00	1.00	1.00	1.00	1.00	1.00
Single/Separated/Divorced	1.32 (0.86–2.02)	1.32 (0.86–2.01)	1.30 (0.85–1.99)	1.31 (0.85–2.00)	1.97 (1.17–3.31)	1.86 (1.10–3.15)	1.91 (1.14–3.19)	1.78 (1.05–3.01)
Neighborhood % Hispanic/Latinx
Lowest tertile	1.45 (0.52–4.05)	1.58 (0.57–4.43)			0.49 (0.27–0.90)	0.53 (0.29–0.98)
Middle tertile	1.06 (0.70–1.58)	1.07 (0.71–1.61)			0.85 (0.54–1.34)	0.88 (0.56–1.39)
Highest tertile	1.00	1.00			1.00	1.00
Neighborhood % Poor
Lowest tertile			1.00	1.00			1.00	1.00
Middle tertile			0.91 (0.50–1.66)	0.78 (0.42–1.45)			1.25 (0.77–2.03)	1.22 (0.75–1.99)
Highest tertile			1.13 (0.64–2.00)	0.95 (0.52–1.71)			2.17 (1.28–3.68)	2.10 (1.23–3.57)
Socioeconomic status
≤200% FPL and ≤high		1.55 (1.05–2.29)		1.51 (1.03–2.23)		1.46 (0.87–2.46)		1.52 (0.91–2.55)
school graduate
200+% FPL and/or some college or more		1.00		1.00		1.00		1.00

FPL = federal poverty level.

Neighborhood tertiles based on statewide distribution of census tracts.

Table 4.

Linear regression models for number of risk factors stratified by nativity, GROW, 2012–2013, N = 1,248.

	Latina, immigrant				Latina, US-born
	(n = 809)				(n = 439)
	beta (se)	beta (se)	beta (se)	beta (se)	beta (se)	beta (se)	beta (se)	beta (se)
Maternal age  20–29 years	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
30–39 years	0.068 (0.067)	0.079 (0.066)	0.070 (0.066)	0.080 (0.066)	−0.016 (0.079)	0.006 (0.083)	−0.005 (0.081)	0.023 (0.084)
40 years or older	0.074 (0.072)	0.092 (0.072)	0.079 (0.072)	0.089 (0.072)	0.104(0.105)	0.127 (0.108)	0.116 (0.106)	0.144(0.109)
Marital status  Married/living together	Ref	Ref	Ref	Ref	Ref	Ref	Ref	Ref
Single/Separated/Divorced	0.156 (0.058)	0.156 (0.058)	0.149 (0.058)	0.152 (0.057)	0.214 (0.093)	0.196 (0.098)	0.211 (0.091)	0.190 (0.096)
Neighborhood % Hispanic/Latinx  Lowest tertile	−0.068 (0.150)	−0.018 (0.143)			−0.407 (0.116)	−0.373 (0.117)
Middle tertile	−0.011 (0.064)	0.001 (0.064)			−0.142 (0.082)	−0.121 (0.084)
Highest tertile	Ref	Ref			Ref	Ref
Neighborhood % Poor  Lowest tertile			Ref	Ref			Ref	Ref
Middle tertile			−0.020 (0.104)	−0.066 (0.103)			0.187 (0.089)	0.177 (0.090)
Highest tertile			0.095 (0.097)	−0.035 (0.096)			0.345 (0.091)	0.327 (0.093)
Socioeconomic status  ≤200% FPL and ≤high school graduate		0.171 (0.064)		0.156 (0.064)		0.118 (0.092)		0.120 (0.086)
200+% FPL and/or some college or more		Ref		Ref		Ref		Ref

FPL = federal poverty level.

Neighborhood tertiles based on statewide distribution of census tracts.

Bold indicates p < 0.05.

Beta estimates are unstandardized.

Discussion

Among immigrant Latinas, low individual-level SES was associated with greater odds of multiple risk factors while neighborhood-level factors were not. Among US-born Latinas, low neighborhood-level Hispanic/Latinx concentration was associated with lower odds of multiple risk factors, and high neighborhood-level poverty and low-individual SES were associated with higher odds. Immigrant Latinas seem more affected by individual SES; whereas, US-born Latinas seem more affected by neighborhood characteristics as higher concentrations of both poverty and Hispanic/Latinx populations were associated with increased risk. Results point to the idea that there are various factors impacting this population and suggest heterogeneous effects between nativity (or language) and social factors that produce risk for chronic disease among Latinas.

The current study’s results found that immigrant Latinas had lower levels of physical activity, and US-born Latinas had higher levels of smoking and binge drinking. Both groups had high levels of not meeting dietary guidelines with regards to fruits and vegetables. Moreover, both Latina groups had higher odds of modifiable behavioral risk factors for chronic disease as a function of different social factors. These findings support evidence from the literature (Ludwig et al., 2011; Osypuk et al., 2009). The finding that US-born Latinas living in a neighborhood with a low Hispanic/Latinx concentration had lower odds of engaging in multiple risk behaviors indicates that the potential protective factors of living in an ethnic enclave may not exist for US-born Latinas. This is consistent with literature that argues that ethnic enclaves may be detrimental for later generations (Li et al., 2017).

As mentioned above, it has been proposed that living in an ethnic enclave influences the physical health of immigrants by providing critical financial support (Roy et al., 2013). The idea is that financial support comes from the strength of social ties and social support among immigrants (Viruell-Fuentes, Morenoff, Williams, & House, 2013). One of the proposed strengths of the social ties in an ethnic enclave is the ability to find employment. For example, although people in ethnic enclaves typically live in high poverty, there is often a slight diversity in socioeconomic status, allowing a small number of community members to establish businesses and employ their co-ethnics (Xie & Gough, 2011). Additionally, ethnic enclaves can become their own economic centers allowing them to function internally as a labor market, making skills, such as language and cultural knowledge, important for marketability in the internal labor market (Xie & Gough, 2011). There was no evidence that immigrant Latinas were protected by living in an ethnic enclave. One possible explanation for no evidence is the heterogeneity of ethnic enclaves. These benefits may not be universal and can vary based on language ability, gender, and ethnic composition of the neighborhood. The latter meaning that even within the Latinx community, there may be preference given to people who share country/region of origin.

The evidence here indicates that the driving factor behind immigrant Latinas engaging in risky behavior is low individual SES (i.e., low-income, low-education). This study did not provide evidence of this, but other studies’ findings suggest it is possible that effects of individual SES can be overcome by the strengths of the community in an ethnic enclave (Li et al., 2017; Roy et al., 2013; Xie & Gough, 2011). There are other factors that could be preventing people from receiving potential benefits of the community. The community could be overtaken by gang activity hindering positive growth in the community due to safety concerns. Positive community growth that could lead to engaging in less risky behavior could also come from schools or continuing education environments that may or may not exist within an enclave. Organic social ties can be difficult to achieve in the current age of social media that seemingly affects social interaction. One study argues that only people who are able and willing to engage in social media networks are likely to reap social capital benefits, making it increasingly difficult if one is an outsider to this way of communicating (Ellison, Steinfield, & Lampe, 2011). Various face-to-face interactions have moved online (i.e., employment postings) and daily life is becoming more isolated. To overcome this and re-visit a more effective model of interaction, Fitts and McClure (2015) suggest it is beneficial to bring immigrant Latinas together and provide a space so that they can build community and share important knowledge with each other. That knowledge may be about where to find employment, affordable housing through a friend, or an informal child care setting. This in turn could lead to immigrant Latinas building social capital that will help provide stability. Stability, here, is defined as consistent employment, housing, and strong social ties that would affect positive health behavior. Often this is done in community resource settings or churches, which could be a barrier if the enclave does not have the adequate resources for these settings. Targeted interventions efforts could focus on helping immigrant Latinas establish stability within an ethnic enclave (Glaser et al., 2015).

The results suggest that the key factor putting US-born Latinas at risk is living in a high-poverty or high Hispanic/Latinx neighborhood. Whereas, immigrant Latinas may need support to establish themselves within ethnic enclaves, US-born Latinas may need supports to help them expand their network outside of ethnic enclaves. Some researchers argue that living in an ethnic enclave may lead to poor health for US-born racial/ethnic minoritized persons because ethnic enclaves restrict community members to the resources within the enclave, limiting their opportunities for growth beyond the enclave (Li et al., 2017; Roy et al., 2013). High poverty neighborhoods are well known to be associated with environments that may hinder choices to adopt and maintain healthy behaviors, such as through the built environment (e.g., fewer access to healthy affordable food, higher exposure to advertisements for alcohol and tobacco) and social environment (e.g., feeling unsafe for physical activity outdoors). Neighborhood-level investments and engagement among its residents to improve neighborhood conditions are important social policies to consider. Because neighborhoods that are high in the proportions of Hispanic/Latinx persons & poor persons covary (in this study, the correlation between the two neighborhood variables is 0.53), it is difficult to parse the specific mechanism without further research.

At the individual level, A targeted social work approach with the use of Promotoras could be useful in helping US-born Latinas or long-term resident immigrant Latinas in ethnic enclaves overcome challenges with access to more lucrative employment or educational opportunities (Balcázar et al., 2016), as well as with reducing the stressors and regaining the supports they may have lost as they became more acculturated. Immigrant Latinas could benefit from more immediate supports before transitioning to help accessing higher level employment or educational supports. Social work interventions with Latinx communities are increasingly utilizing Promotoras as a way to establish a sense of trust for the client (Arizmendi & Ortiz, 2004; Contreras, Larson, Pierpont, Griffith, & Rocha-Peralta, 2012). Based on the evidence from the current study, a primary goal would be to help both groups find opportunities, whether through education or other employment, to increase their socioeconomic status. This in turn could provide individuals with a feeling of more control over their lives. More control, specifically financial control, could help them make decisions that could improve their surroundings where they were to remain in an ethnic enclave or seek opportunities outside of the enclave with their families. For example, one could choose to stay in an ethnic enclave and use their resources to improve their communities. Others could choose to move out of the enclave and determine if a change of neighborhood is best for them. There are many nuances to these decisions and there would be benefits and disadvantages to each decision but having the option be theirs could promote self-sufficiency and feeling of control over their decisions, including health decisions.

Limitations

The inability to examine differences among Hispanic/Latinx subgroups is a limitation of the current study. The Hispanic Paradox literature suggests the paradox is most relevant for Mexicans but for this reason it is important to examine subgroup differences to the extent possible (Li et al., 2017). However, given that 81.5% of Hispanics in California are Mexican origin, the current study’s results mostly reflect that group (Pew Research Center, 2014; US Census Bureau, 2019). Another limitation is that while measures can vary from study to study, this study used nativity as a crude measure of acculturation, as opposed to a scale or duration in the US. However, similar results were found with language as the measure of acculturation in sensitivity analyses. We did not have information, however, on duration in the US. Additionally, it is possible that the merits of education as a protective factor were lost because of the analytic strategy selected (combining income and education). A small analytic sample limited more extensive measurement of individual SES. SES is often a difficult variable to conceptualize given data limitations but the strategy was that a composite of income and education captured a broader sense of individual SES and is seen as a strength. Another variable that was not examined that could have been a significant covariate was number of children in the home. While marital status was included, a greater number of children could have an effect on a single mother’s health. The sample size across the modifiable behavioral risk factors meant that findings were probably largely driven by physical activity and fruit and vegetable intake. Similarly, the majority of immigrant Latinas lived in the highest tertiles of poverty and Hispanic/Latino concentrations, reflecting societal segregation; this may be why we did not detect neighborhood-level associations for this group. Despite the limitations, the current study had strengths. The GROW study is a representative sample from a large and diverse state. While we are not able to examine Hispanic/Latinx subgroup differences, the dataset still allows for immigrant versus US-born comparisons.

Conclusions

Based on the findings of the present study, it is suggested that social workers in policy arenas look toward ensuring Latina women have opportunities for stability. Providing a venue for achieving social capital and social support would strengthen their ability to practice good health. Social work practitioners who work in medical settings or settings that provide resources for housing and employment can provide tailored supports to immigrant and US-born Latinas. US-born Latinas may be at a stage where they may benefit from assistance with more advanced educational and employment opportunities. For immigrant Latinas, practitioners can help them build environments, including social capital, that promote good health. It is possible that one can find employment and housing but be surrounded by those practicing poor health habits that increase risk of chronic disease, such as diabetes and heart disease. A well-trained social worker can help women identify potentially harmful environments to avoid. Future studies should examine at what point an immigrant becomes “acculturated enough” that there is a change in approach to help them. Social workers are well equipped to help clients achieve good health by helping them become self-sufficient in dealing with social factors that affect their health. Latina women in the U.S. have the difficult task of balancing their many identities to achieve good health. Good social work service will take into account how various factors may interact to help clients assess their strengths to become self-sufficient in managing their health.