“Comparative effectiveness of behavioral interventions for cardiovascular risk reduction in Latinos: A systematic review”

Abstract

Background:

Latinos, the fastest growing ethnic minority group in the United States (U.S.), are at high risk for cardiovascular disease (CVD). However, little is known about effective strategies to reduce CVD risk in this population.

Objective:

To systematically review and synthesize evidence from randomized controlled trials that examined the effectiveness of behavioral interventions to reduce CVD risk in Latinos living in the U.S.

Methods:

Four electronic databases were searched for relevant peer-reviewed English- and Spanish-language articles published between January 1, 2000 and December 31, 2019. Four reviewers independently completed article screening, data abstraction, and quality appraisal. At least two reviewers completed data abstraction and quality appraisal for each article and a third reviewer was assigned to settle disagreements. Data on study characteristics and outcomes were abstracted.

Results:

We retrieved 1,939 articles. After applying inclusion/exclusion criteria, 17 articles were included. The majority of interventions were led by community health workers (n = 10); two family-based interventions were identified. None of the included studies were nurse-led. Behavioral factors were assessed across all included studies, while only four studies reported on psychosocial outcomes. Improvements were observed in dietary habits and psychosocial outcomes. Findings for physical activity and biological outcomes were mixed. We identified no differences in outcomes based on intervention modalities used or the role of those who led the interventions.

Conclusion:

Existing evidence is mixed. Future research should assess the effectiveness of understudied treatment modalities (including nurse-led, mobile health, and family-based interventions) in reducing CVD risk in Latinos.

INTRODUCTION

Latinos are the largest and fastest-growing ethnic group in the United States (U.S.), representing approximately 18% of the population.¹ It is estimated that over 40% of Latino adults in the U.S. have some form of cardiovascular disease (CVD), such as coronary artery disease, stroke, or heart failure.² In 2017, CVD was the second leading cause of death among Latinos after cancer, accounting for approximately 20% of deaths among Latinos.³

Analyses of population-based data indicate that over 70% of Latino adults have at least one major risk factor for CVD.⁴ Data from the National Health and Nutrition Examination Survey indicate that Latino adults are less likely than non-Latino White and Asian adults to meet at least five of the seven criteria of ideal cardiovascular health, as defined by the American Heart Association.² For example, compared to non-Latino Whites, Latino adults have a higher prevalence of obesity.^5,6 Latinos are also less likely to meet aerobic physical activity recommendations and more likely to engage in sedentary behaviors than non-Latino White and Black adults.^7,8 In addition to physical inactivity, much of the elevated CVD risk in Latinos may be attributed to high rates of uncontrolled CVD risk factors.¹ Although Latinos report the lowest rates of high cholesterol screening,⁹ Latino men have the highest prevalence of elevated total cholesterol.² Latinos have lower mortality attributed to hypertension,¹⁰ however between 2007-2017 the mortality rate due to hypertension increased by 17.7% in Latinos versus 1.8% in Black adults.¹¹

Social determinants of health, including cultural factors, may further contribute to cardiovascular health disparities in Latino adults. Latinos experience barriers to health care,¹² including lower access to preventive health care and primary care providers¹³ and lower rates of health insurance compared to other racial/ethnic groups in the U.S.¹ For instance, less than 30% of Latinos with high cholesterol receive recommended medication treatment.¹⁴ Further, both U.S. and foreign-born Latino adults have inadequate health literacy, which can impact the quality of health care received and ability to comprehend health information.^15,16 In 2015, over 20% of Latino adults in the U.S lived in poverty.¹⁷ Latinos often live in disadvantaged communities that might lack access to healthy foods.^18,19 Moreover, the role of culture (including traditions, beliefs, values, and practices existing within a community) has been identified as a social determinant of health.²⁰ Higher acculturation among Latinos in the U.S. has been associated with a greater likelihood of eating food prepared outside of the home, which generally have higher sugar and fat content.^21,22 Paradoxically, traditional food preparation and the socialization of family with food is also associated with poor diet quality among Latinos.¹³

Given the growing number of Latinos in the U.S. and the increased prevalence of CVD risk factors observed in Latinos, reducing cardiovascular health disparities in this population represents a major public health concern. To address cardiovascular health disparities in Latinos, the American Heart Association has recommended that researchers design interventions that take into account social determinants of health, such as cultural factors.²³ Therefore, the purpose of this systematic review was to review and synthesize evidence from randomized controlled trials (RCTs) that examined the effectiveness of behavioral interventions for CVD risk reduction among Latinos living in the U.S and provide recommendations for future CVD research with this population.

METHODS

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guided the conduct and reporting of this systematic review.²⁴ The search strategy was developed in consultation with an informationist at the Columbia University health sciences library. We conducted a search of four electronic databases: Cumulative Index to Nursing and Allied Health Literature, Embase, PubMed, and Scopus. Table 1 shows the search strategy and filters used. We combined search terms in three categories: 1) Latinos, 2) CVD, and 3) intervention research. Subject headings were used when available. Further, we conducted an ascendency and descendancy search of the preliminary list of included articles to identify additional records.

Table 1.

Search Strategy

Operator	Definition
Keywords
Latino keywords	Latino OR Latinas OR Latinx OR Hispanic American OR Hispanic
Cardiovascular disease keywords	Cardiovascular disease OR heart disease OR angina OR angina pectoris OR stroke OR cerebrovascular accident OR heart failure OR coronary heart disease OR coronary artery disease OR ischemic heart disease
Intervention keywords	Intervention OR behavioral intervention OR randomized clinical trial OR randomized controlled trial OR clinical trial
Age filters	Adults, 19+ years
Language filters	English and Spanish only
Year filters	January 1, 2000-December 19, 2019
Location filters	United States

Inclusion and Exclusion Criteria

All peer-reviewed English and Spanish-language articles that were conducted in the U.S. and published between January 1, 2000 and December 31, 2019 were eligible for inclusion. Articles were limited to studies conducted in the U.S. to understand the unique cultural and social considerations of Latinos in the U.S. We included all studies with an RCT design that assessed the comparative effectiveness of behavioral interventions for CVD risk reduction in Latino adults (over the age of 18). Articles were included if at least 90% of participants were Latino. We excluded studies with quasi-experimental designs, literature reviews, case studies, and non-empirical publications. Because our focus was on primary and secondary prevention of CVD, we excluded RCTs that sought to improve self-management in individuals with an existing diagnosis of CVD. Given extensive research on interventions for diabetes self-management and prevention among Latinos,^25,26 we excluded articles that solely focused on improving diabetes management rather than targeting diabetes management with the intent to reduce CVD risk in this population.

Data abstraction

Four authors participated in data abstraction (LVE, JS, MRT, and BAC). Data abstraction consisted of data reduction, data display, data comparison, conclusion drawing, and verification.²⁷ Data from each study was abstracted by one author and verified by a second author. The lead author (LVE) developed themes based on the data presented in summary tables. An additional author (BAC) reviewed and further refined themes. The remaining authors then reviewed tables to confirm findings.

Quality Appraisal

We appraised the quality of included articles using the Downs and Black checklist, which is a 27-item tool used to critically appraise the methodological quality of randomized and non-randomized health care interventions.^28,29 This checklist is divided into four categories: 1) study reporting; 2) external validity; 3) internal validity; and 4) power. Each item, except question five, is scored as a binary indicator (i.e., Yes = 1; No = 0) indicating whether or not data was reported related to that item. Question five, which assesses the distribution of confounders across intervention arms, is weighted higher (i.e., Yes = 2; Partially = 1; No = 0). Scores range from 0-28, with a higher score indicating higher quality (≤ 14 = poor; 15-19 = fair; 20-25 = good; 26-28 = excellent). Each study was independently appraised by two reviewers. A third reviewer was consulted to resolve appraisal discrepancies, as needed.

RESULTS

Search Results

A total of 1,939 citations were retrieved from our database search (Figure 1). After excluding duplicates, we reviewed the titles and abstracts of 942 articles. After applying inclusion and exclusion criteria, the full texts of 44 articles were reviewed of which 15 met inclusion criteria. Two additional articles^30,31 were identified through ascendency and descendancy searches for a total of 17 included articles, representing 15 unique studies.

Figure 1.

PRISMA Flow Diagram

Quality Appraisal

Table 2 presents quality appraisal scores for the included studies. Quality appraisal scores ranged from 16 (fair) to 27 (excellent). Although the included studies had generally good methodological quality, common limitations identified were short follow-up (less than 12 months), high attrition rates (greater than 20%), use of self-report to measure behavioral outcomes, potential contamination between intervention and control group participants, and lack of attention-control groups.

Table 2.

Individual Study Quality Appraisal Scoring

Study	Downs and Black Checklist Domainsa
	Study Reporting (11)	External Validity (3)	Internal Validity (12)	Power (1)	Overall (range = 0-27)
Balcazar et al. 2009	9	1	9	0	19
Balcazar et al. 2010	11	2	9	0	22
Carrasquillo et al., 2017	11	2	12	1	27
Coleman et al., 2012	10	3	9	0	22
deHeer et al., 2017	7	3	7	0	17
Elder et al., 2000	7	3	10	0	20
Hayashi et al., 2010	11	1	8	0	20
Keller & Cantue, 2008	7	2	7	0	16
Khare et al., 2014	10	0	10	0	20
Koniak-Griffin et al., 2015	7	1	10	1	19
Otilingam et al., 2015	9	1	9	0	19
Poston et al., 2001	9	0	11	0	20
Rosas et al., 2015	9	1	11	1	22
Skapinksy et al., 2018	10	2	10	0	22
Soto Mas et al., 2018	11	1	9	0	21
Toobert et al., 2011a	7	1	12	1	21
Toobert et al., 2011b	11	2	10	0	23

^aDowns SH, Black N. The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions. J Epidemiol Community Health. 1998;52(6):377-384. doi:https://10.1136/jech.52.6.377.

Study Characteristics

Table 3 presents study characteristics. All included articles were published in English between 2000 and 2018. A total of seven studies were conducted in Texas, ^30,32–37 six in California,^38–43 two in Colorado,^44,45 one in Florida,³¹ and one in Illinois.⁴⁶ Sample sizes ranged from 18 to 1,093. The mean age of participants across studies ranged from 31 to 60 years. Most articles (n = 9) focused solely on women.^{30,33,38–40,42,44–46} Ten articles had a majority of participants who were of Mexican descent, while the remaining articles did not specify participants’ countries of origin.

Table 3.

Study Characteristics

Author, Year	Location	Sample size	Participant characteristics	Methodology
Balcazar et al., 2009	Lower Valley of El Paso, Texas	N = 98 Intervention group = 58 Control group = 40	Mean age = 54.9 – 49.7 Female = 87.9% Country of origin = 82% born in Mexico	Community health worker-led: Group educational classes in the community, telephone calls Cultural adaptation: Community-based participatory research (CBPR)
Balcazar et al., 2010	Lower Valley of El Paso, Texas	N = 328 Intervention group = 192 Control group = 136	Mean age = 54 Female = 70% Country of origin = 94% born in Mexico	Community health worker-led: Group educational classes in the community Cultural adaptation: CBPR
Carrasquillo et al., 2017	Miami, Florida	N = 300 Intervention group = 150 Control group = 150	Mean age = 54 % female in intervention group = 55.3% % female in control group = 54.7% Country of origin = 38.3% born in Cuba	Community health worker-led: Home visits, telephone calls, and group-level activities Cultural adaptation: CBPR
Coleman et al., 2012	San Diego and Los Angeles counties, California	N = 1,093 Intervention group = 552 Control group = 541	Mean age = 52 Female = 100% Country of origin = 65% Mexican or Mexican American	Community health worker-led: One-on-one counseling sessions in community clinics Cultural adaptation: Expert panel
deHeer et al., 2017	Harris County, Texas	N = 162 families Adults = 320 Children = 1,081	Adult mean age = 49.1 Child mean age = 23.3 Adult female = 55.0% Child female = 50% Country of origin = 80.9% born in Mexico	Mailed bilingual pamphlets to families with graphical representation of participants’ family health history, modified to an 8th grade reading level Cultural adaptation: Not reported
Elder et al., 2000	San Diego, California	N = 732	Mean age = 31.0 years Female = 59% Country of origin = Not specified	Group educational classes in community colleges Cultural adaptation: Not reported
Hayashi et al., 2010	Los Angeles and San Diego, California	N = 869 Intervention group = 433 Control group = 436	Mean age intervention group = 51.8 years Mean age control group = 52.1 years Female = 100% Country of origin = Not specified	Community health worker-led: Group educational workshops at four health centers Culturally adaptation: Intervention led by bilingual, bicultural community health workers
Keller & Cantue, 2008	San Antonio, Texas	N = 18 Intervention group 1 = 11 Intervention group 2 = 7	Mean age intervention group 1 = 56.5 years Mean age intervention group 2 = 53.5 years Female = 100% Country of origin = Not specified	Community health worker-led: Weekly walking sessions in participants’ neighborhood/community center Cultural adaptation: Utilized culture-specific moderators to target physical activity (e.g., participant partnerships)
Khare et al., 2014	Chicago, Illinois	N =180 Enhanced intervention group = 90 Minimum intervention group = 90	Mean age = 50.9 Female = 100% Country of origin = Not specified	Group educational workshops in the community Cultural adaptation: Literacy expert, translator, pilot testing in Spanish and English, orientation session for family members of Spanish intervention
Koniak-Griffin et al., 2015	Los Angeles, California	N = 223 Intervention group = 111 Control group = 112	Mean age = 44.6 Female = 100% Country of origin = 82.8% born in Mexico	Community health worker-led: Group educational workshops in the community Cultural adaptation: CBPR
Otilingam et al., 2015	Los Angeles County, California	N = 100 Intervention group 1 = 32 Intervention group 2 = 33 Control group 1 = 17 Control group 2 = 18	Mean age intervention group 1 = 57.1 Mean age intervention group 2 = 60.1 Mean age control group 1 = 61.1 Mean age control group 2 = 58.1 Female = 100% Country of origin = not stated	Community health worker-led: Group educational workshops in community clinics Cultural adaptation: Literature review, community engagement, pilot testing
Poston et al. 2001	Southern Texas	N = 379 Intervention group = 194 Control group = 185	Mean age intervention = 39.2 Mean age control = 40.0 Female = 100% Country of origin = Mexico	Community health worker-led: Group educational workshops in the community and walking clubs Cultural adaptation: Bilingual materials and instructors, modified and culturally tailored rationales for diet and activity modifications, and use of established social support networks
Rosas et al., 2015	North Fair Oaks, California	N = 207 Intervention group 1 = 84 Intervention group 2 = 82 Control group = 41	Mean age = 48 Female = 76.8% Country of origin = 76.8% born in Mexico	Community health worker-led: Group educational workshops in the community and home visits Cultural adaptation: Community engagement
Skapinksy et al., 2018	Southwestern metropolitan city and surrounding area in Texas	N = 162 multigenerational households 137 females (wives) 126 males (husbands)	Mean age for wives = 47.9 years Mean age for husbands = 49.7 years Female = 58.1% Country of origin = Mexico	Mailed bilingual pamphlets to spouses of their family health history, modified to an 8th grade reading level Cultural adaptation: Not reported
Soto Mas et al., 2018	El Paso, Texas	N = 181 Intervention group = 77 Control group = 78	Mean age = 52.3% over the age of 46 Female = 80.6%% Country of origin = 91.6% born in Mexico	Group educational classes in the community Cultural adaptation: Bilingual educational materials and classes
Toobert et al., 2011a	Commerce City, Colorado	N = 280 Intervention group = 138 Control group = 142	Mean Age = 57.0 years Female = 100% Country of origin = 79.6% born in the USA; 15.8% born in Mexico	Group retreat, educational meetings in the community Cultural adaptation: Literature review and focus groups, pilot study with participants
Toobert et al., 2011b	Commerce City, Colorado	N = 280 Intervention group = 142 Control group = 138	Mean age = 57.1 Female = 100% Country of origin = Not specified	Group educational meetings or classes in the community or at home Cultural adaptation: Literature and focus groups, pilot study with participants

Table 4 summarizes intervention characteristics of the included articles. Ten (58.8%) interventions were led by promotoras/promotores de salud, hereafter referred to as community health workers (CHWs).^{30,31,33,34,38–40,42,43,47} None of the included studies were nurse-led interventions. Eight articles (47.1%) described interventions delivered via group educational workshops or classes in the community or at home, some of which were led by CHWs.^{32,34,37,39,41,44–46} The CHW-led interventions had follow-ups ranging from 6.5 to 24 months. Interventions that consisted of in-person group workshops or classes had follow-up assessments ranging from one to 24 months. The authors of two articles reported on separate analyses from the same intervention. The intervention consisted of mailing customized family health history pamphlets to Latino families with 10-month follow-up and included either children or spouses.^35,36

Table 4.

Summary of Interventions

Authors	Intervention Characteristics	Results	Limitations
Balcazar et al., 2009	Duration: 9 weeks Intervention group: -2-hour module delivered by CHWs in Spanish. -Additional module for hypertension control. -Follow-up telephone calls to discuss lifestyle changes following modules. Control group: -Educational materials related to overall health at baseline.	Behavioral outcomes: -Healthy dietary intake was higher in the intervention group vs. control group (p = .04). Biological outcomes: -No significant difference between groups. Knowledge and attitudes: -Intervention group had significantly higher perceived benefit that behaviors will help them control their BP (p = .02).	-Short follow-up of 9 weeks. -Group contamination likely. -Lack of assessment and or monitoring of medication use in both groups. -No attention control group. -Did not report on participant compliance or staff fidelity.
Balcazar et al., 2010	Duration: 4 months Intervention group: -8 weekly of Su Corazon, Su Vida curriculum conducted by CHWs -Follow-up telephone calls and one small group session. Control group: -Basic educational materials from curriculum.	Behavioral Outcomes: -At follow-up intervention group had improvements in: weight control practices (p = .01), salt intake (p < .001), and cholesterol and fat intake (p = .01). Biological outcomes: -Intervention group had significantly higher DBP at follow-up (p < .01) -Framingham score was 5% lower for both groups (p < .001). Knowledge and attitudes: -Perceived susceptibility to CVD (p = .01) and perceived benefits of behaviors of control of CVD risk factors (p = .01) were higher in the intervention group.	-Short follow-up of 4 months. - Significant baseline differences between groups were present. -Reduced statistical power. -No attention control group. -No information provided about the 44 (13%) participants lost to follow-up. -Did not report on participant compliance or staff fidelity.
Carrasquillo et al., 2017	Duration: 12 months Intervention group: -Personalized CHW intervention via tailored home visits and telephone calls, support of non-medical services to address social determinants of health. Control group: - Usual care plus diabetes education materials. -Follow-up phone calls to confirm receipt of information.	Behavioral outcomes: -No difference in physical activity or fruit and vegetable intake between groups. -No difference in intensification of medication regimen or medication adherence between groups. Biological outcomes: -HbA1c levels in the intervention group were lower than control group participants (p < .05). -Intervention group participants had significant reduction in SBP (p < .05). -No differences in LDL or BMI found between groups.	-Greater than 20% attrition in intervention and control groups. -Missing data at 12-month follow-up for 28.3%. -Approximately half already at target SBP at baseline. Participants were already linked to diabetes care. -Did not report on participant compliance.
Coleman et al., 2012	Duration: between 9-14 months (average 12 months) Intervention group: -CHW delivered one-on-one counseling sessions adapted from a previously developed curriculum. Control group: -Usual care for elevated BP or high cholesterol. -Some received education and referral to healthy lifestyle resources.	Behavioral outcomes: -The intervention group was twice as likely to report being in the action/maintenance stage for vigorous physical activity at follow-up. -Greater improvements in the intervention group to initiate new physical activity, incorporate physical activity into daily life, and perform daily habits more briskly. -Intervention group was more likely to self-report engaging in moderate and vigorous physical activity at follow-up (p < .001) compared to the control group.	-Physical activity was self-reported. -No attention control group. -Follow-up data was collected by the same interveners. -Being recruited by mail limits samples to those who have permeant housing. -225 (21%) of participants were lost to follow-up. -Did not report participant adherence.
deHeer et al., 2017	Duration: 10 months Families were randomized into one of four interventions (described below) where they received pamphlets within one week of assessment. Intervention Group 1: All adults received risk assessment, with personalized recommendations added. Intervention Group 2: One adult received risk assessment, with personalized recommendations added. Intervention Group 3: All adults received risk assessment, no personalized recommendations. Intervention Group 4: One adult received risk assessment, no personalized recommendations.	Behavioral outcomes: -Across all groups, encouragement and co-engagement of physical activity increased significantly between baseline and the 10-month follow-up for parents to children and children to parents (p < .001). -Parents were more likely (OR = 2.02, 95% CI = 1.17-3.50) to develop new ways to encourage their children to be physically active and maintain a healthy weight if all adults in household received family health history-based risk information, compared to only one adult. -The personalized behavioral recommendations did not have an impact on outcomes. -New encouragement from parents resulted in increased odds of co-engagement of physical activity at 10-month follow-up (OR = 3.75, 95% CI = 1.01-13.92). -New encouragement from children also resulted in new parent-child physical activity co-engagement at 10 months follow-up (OR = 7.19, 95% CI = 1.95-26.48).	-Limited data on adult health literacy and language preferences -SES based only on home and car ownership. -Encouragement questions were not tailored to parents’ or children’s baseline behaviors. -Study findings are based on parents’ perceptions alone. -Social desirability bias could have influenced study results. -Did not report on participant compliance or staff fidelity.
Elder et al., 2000	Duration: post-test at 3 months and follow-up until 6 months Intervention group: -Nutrition and heart health information integrated into ESL classes Control group: -Stress management integrated into ESL classes.	Behavioral outcomes: -Fat avoidance scores increased more in the intervention group. Biological outcomes: -Intervention group showed greater reductions in the total cholesterol to HDL cholesterol ratio and SBP at 3-month post-test. No differences at 6-month follow-up found. Knowledge and attitudes: -Increased nutrition-related knowledge for intervention group.	-Attrition was a concern with only 72% of participants providing longitudinal survey data. -Randomization resulted in more women in the control group than intervention group. -Evidence of contamination present: 14% of ESL instructors provided control group participants with some information on heart disease and nutrition. -Did not report on participant compliance or staff fidelity.
Hayashi et al., 2010	Duration: 12 +/− 2.5 months> Intervention group: -Lifestyle intervention with face-to-face sessions consisting of assessment and counseling for nutrition and physical activity by CHW. This is a structured program intended for CVD risk reduction. Control group: -Educational pamphlets related to high BP and high cholesterol.	Behavioral outcomes: -The intervention group was more likely to make improvements in health behaviors including eating habits (p < .001) and physical activity (p < .001). Biological outcomes: -Both groups had significantly lower 10-year CHD risk between baseline and follow-up (p < .05). There was no difference between groups. -Only SBP significantly decreased for intervention group participants compared to control group (p = .04).	-Social desirability bias from self-reported data on health behavior outcomes. -Follow-up period not long enough to determine maintenance of health behaviors. -Timing of behavior changes not assessed in relation to CVD risk factors. -Did not report on participant compliance or staff fidelity.
Keller & Cantue, 2008	Duration: 36 weeks Intervention group: -CHW-led weekly walking sessions at a community center. -Heart health information provided. Intervention group 1: -Walk 3 days per week for 30 minutes at a 20-minute mile pace. Intervention group 2: -Walk 5 days per week for 30 minutes at a 20-minute mile pace.	Behavioral outcomes: -Neither group met physical activity goals. Biological outcomes: -BMI decreased significantly between groups over time (p = .001); Intervention 1 differences observed up to follow-up (p < .001). -BMI and weight significantly decreased in both intervention group 1 (p < .05) and intervention group 2 (p < .001) from baseline to 36 weeks. -No significant change in serum cholesterol levels. -LDL levels decreased in intervention 2 group, but not intervention 1 group. -Triglycerides decreased in intervention 1 group, but not in intervention 2 group.	-Self-reported behavioral outcomes may introduce bias. -Neither group met expectations for minutes walked per week. -High attrition in both groups: 63.6% in intervention group 1 and 42.8% in intervention group 2. -Small sample size. -Did not report on staff fidelity.
Khare et al., 2014	Duration: 12 months MI group: -CVD risk factor screening. -CVD-related educational handouts -referrals for physician care as needed; -follow-up assessments -postcards and newsletters. EI group: -Minimum intervention plus a 12-week nutrition and physical activity lifestyle change intervention	Behavioral outcomes: -EI group had significant improvements only in the dietary fiber summary score (p = .03) at 12-month follow-up. -EI group had increases in both total physical activity (p = .01) and moderate intensity physical activity (p = .01) from baseline to post-intervention. These increases in physical activity were attenuated at 12-month follow-up. Biological outcomes: -No differences in total cholesterol and LDL cholesterol between the MI and EI groups at 12 months. -At 12-month follow-up, BMI for the EI group was lower than that the MI group (p = .03).	-High attrition in the enhanced intervention group and minimum intervention group (40.0% and 26.7% respectively). -Self-reported behavioral outcomes. -Participants had healthy nutrition behaviors at baseline, which may limit generalizability of results. -Did not report on participant compliance or staff fidelity.
Koniak-Griffin et al., 2015	Duration: 6 months Intervention group: -Lifestyle behavior intervention group education plus individual teaching and coaching -Weekly classes -Individual teaching and coaching by CHWs to promote healthy lifestyle behaviors (diet and physical activity) Control group: -Safety/disaster preparedness educational program	Behavioral outcomes: -Scores for dietary habits improved for women who received the intervention. -No difference in minutes of moderate physical activity between groups from baseline to 9 months measured using accelerometer data. Biological outcomes: -Intervention group significantly decreased waist circumference over the follow-up period. Weight and cholesterol changes were not significant. Knowledge and attitudes: -Heart Disease Knowledge questionnaire scores increased in the intervention group (p < .001).	-Study only addressed short-term change in behavior. -Control group may not be the “best” control group match. -Self-recall bias, exact measurement of diet intake not possible with dietary intake questionnaire. -Small sample size.
Otilingam et al., 2015	Duration: 1 month Intervention group 1: Heart health only -Two workshops used education techniques including cooking workshops, fotonovelas, and game show formats. Intervention group 2: Brain and heart health -Included information on the link between metabolic syndrome and increased risk for dementia, visual representations of a non-pathological brain versus the brain of someone with AD, and research findings about the relationship between dietary fat intake and increased risk of cardiovascular disease and dementia. Wait-list control group: -Complete pre- and post-test. Post-test only waist-list control group: -Completed only the post-test to examine the impact of pre-testing on outcomes.	Behavioral outcomes: -Intervention group 2 decreased fat consumption compared to control group at follow-up. Knowledge and attitudes: -Intervention group 2 had increased dietary fat knowledge from pretest to posttest that was maintained at follow-up.	-Small sample sizes in each group. -Self-reported behavioral outcomes may introduce bias. -No long-term follow-up of participants. -Did not report on participant compliance.
Poston et al., 2001	Duration: 12 months Intervention group: -Weekly 90-minute meetings led by Mexican American health care professionals for six months followed by weekly peer-led maintenance groups for another 6 months. -Walking groups for physical activity component. Control group: -Waitlist control assessed but did not receive the intervention until after the 12-month follow-up.	Behavioral outcomes: -No significant differences in changes in physical activity at 12 months. Biological outcomes: -No differences between groups.	-Greater attrition in the intervention group compared to control group (47.0% vs. 28.0%, p < .001). -Process measures to evaluate effectiveness of the maintenance phase were not assessed. -Significant differences identified between groups at baseline. -Did not report on participant compliance or staff fidelity.
Rosas et al., 2015	Duration: 24 months Intervention group 1: -CM only, group sessions and individual sessions. Intervention group 2: -CM plus CHW support by CHW home visits in the intensive phase and two CHW home visits in the maintenance phase. Control group: -Usual care and potential for referral to lifestyle counseling.	Behavioral outcomes: -Findings for physical activity not reported. Biological outcomes: -At 6 months, intervention 2 group had significant weight loss (p = .05), which was greater weight loss among men than women. All weight lost was regained by 24 months. -No significant difference in waist circumference, BP, lipid levels, and indicators of glycemic status between groups.	-Not a representative sample of Latinos as participants had high prevalence of socioeconomic vulnerability. -Did not report on participant compliance or staff fidelity.
Skapinksy et al., 2018	Duration: 10 months Families were randomized into one of four interventions (described below) where they received pamphlets within one week of assessment. Intervention group 1: One spouse received risk assessment, no personalized recommendations. Intervention group 2: One spouse received risk assessment, with personalized recommendations Intervention group 3: Both spouses received risk assessment, no personalized recommendations Intervention group 4: Both spouses received risk assessments, with personalized recommendations.	Behavioral outcomes: -No significant interaction between personal and spousal personalized risk feedback on physical activity levels. -Wives of husbands who were identified as having an increased risk on baseline assessment with engaged in higher levels of physical activity than other women at 3 months. -No significant effect of personal risk feedback on husbands’ physical activity levels at 3-month or 10-month follow-up.	-Self-reported data for physical activity can lead to bias. -Small sample size. -Limited generalizability to other Latino populations. -Did not report on participant compliance or staff fidelity.
Soto Mas et al., 2018	Duration: 6 weeks Intervention group: -Health literacy and ESL curriculum with CVD specific content. Control group: -Conventional ESL curriculum.	Behavioral outcomes: -Significantly greater improvement in cardiovascular health behaviors scores in the intervention group from pretest to 6-week posttest (p = .05) compared to control group participants. -Findings for diet and physical activity not reported separately but as part of the Cardiovascular Health Questionnaire.	-Short follow-up period of 6 weeks. -Significant differences at baseline between groups. -Did not report on participant compliance or staff fidelity.
Toobert et al., 2011a	Duration: 12 months Intervention group: -Culturally adapted program that promoted Mediterranean diet, physical activity, supportive resources, problem solving, stress management practices, and smoking cessation. -2.5-day retreat and consecutive meetings Control group: -Usual care	Psychosocial outcomes: -Improvements for intervention group compared to control group; maintained up to 12-month follow-up. Behavioral outcomes: -Intervention group had improvements in percent of calories from saturated fat, practice of stress management, and physical activity at 6-month. These were not maintained at 12-month follow-up. Biological outcomes: -The intervention group had significant reductions in BMI, but the early improvement in HbA1c in the intervention group not maintained at 12 months.	-Self-reported behavioral outcomes may introduce bias. -Assessors were aware of treatment assignments after randomization. -Attendance was inconsistent, during the first 6 months it averaged 58% and declined to 48% for meetings between 6 and 12 months. -Did not report on staff fidelity.
Toobert et al., 2011b	Duration: 24 months Intervention group: -Culturally adapted program that promoted Mediterranean diet, physical activity, supportive resources, problem solving, stress management practices, and smoking cessation. -2.5-day retreat and consecutive meetings Control group: -Enhanced usual diabetes care (included one free class covering targeted areas of the intervention).	Psychosocial outcomes: -Improvements in psychosocial outcomes observed at 6-month follow-up were maintained at 24-month follow-up. Behavioral outcomes: -The intervention group had improvements in percent calories from saturated fats at 24-month follow-up. -Initial improvements in physical activity and stress management that were observed in intervention group participants at 6-month follow-up were not maintained. Biological outcomes: -Improvements decreased across the 24-month follow-up. BMI remained lower for the intervention group. The early decrease in HbA1c observed in the intervention group was not maintained. -10-year coronary heart disease risk was not improved in the intervention group.	-Self-reported behavioral outcomes may introduce bias. -Efficacy of the individual intervention elements are unknown. -More than 30% of participants in the intervention and control groups were lost to follow-up at 24 months. -Did not report on participant compliance or staff fidelity.

Abbreviations: CHW, community health worker; BP, blood pressure; DBP, diastolic blood pressure; CVD, cardiovascular disease; SBP, systolic blood pressure; LDL, low-density lipoprotein; BMI, body mass index; OR, odds ratio; CI, confidence interval; SES, socioeconomic status; ESL, English as a Second Language; CHD, coronary heart disease; MI, minimum intervention; EI, enhanced intervention; AD, Alzheimer’s disease; CM, case management; HbA1c, glycated hemoglobin.

All included articles examined more than one outcome. Psychosocial outcomes were measured in four (23.5%) articles.^34,41,44,45 Behavioral outcomes (e.g., diet and physical activity) were examined across all included studies. Biological outcomes (e.g., blood pressure and body mass index [BMI]) were assessed in 13 (76.5%) articles.^{30,31,33,34,36,38,41–47} Changes in nutrition knowledge^39,41 and heart disease knowledge³⁸ were assessed in fewer articles. Only three (17.6%) articles described changes in attitudes towards CVD risk factors as outcomes.^34,41,47 Supplementary Table 1 presents detailed information on outcome measures.

Cultural Adaptation

Cultural adaptation, which incorporates the target audience to purposefully integrate culture and language into health interventions, has been recommended to improve the cardiovascular health of Latinos.²³ The majority of included articles (n=14) reported some form of cultural adaptation in the study design and/or intervention; however there was large variability in the methodologies used for cultural adaptation.^{30,31,32–34,37–40,42–46} The majority of articles (n = 7) reported using at least one of the following methods for cultural adaptation: literature reviews, expert panels, focus groups, and pilot testing with members of the community for cultural adaptation.^{37,39,40,43–46} Authors of four additional articles reported their interventions were led by bilingual CHWs,^33,42,43,46 though the specific methods used for cultural adaptation were not provided. Four articles reported using community-based participatory research methods in developing interventions.^31,32,34,38 Investigators in the remaining studies did not specify methods used as part of cultural adaptation of their interventions.^30,36,41,48 Notably, we did not find a link between the methodology of cultural adaptation used and study outcomes.

Psychosocial Outcomes

Few studies (n = 4) reported findings for psychosocial outcomes such as self-efficacy, perceived support, and problem-solving skills.^34,41,44,45 The majority of these articles reported significant improvements in self-efficacy (n = 4) and social support (n = 2) among intervention group participants. Self-efficacy and social support were assessed with validated measures (Supplementary Table 1). Measures used to assess problem solving (n = 2) were not described.^44,45 One study consisting of problem-solving support groups found higher self-efficacy, social support, and problem-solving scores in the intervention group compared to the control group; improvements in all three outcomes were maintained at 12- and 24-month follow-up.^44,45 Similarly, Elder et al.⁴¹ found Latino adults who participated in a heart health workshop combined with English as a Second Language classes had greater self-efficacy at 6-month follow-up than did control group participants. In contrast, Balcazar et al.³⁴ found no difference in self-efficacy between intervention and control group participants at 4-month follow-up in their CHW-led educational intervention.

Behavioral Outcomes

Dietary intake and habits.

Self-reported changes in diet intake and habits were assessed in 11 (64.7%) articles of which seven were CHW-led.^{31,32,34,37–39,41,42,44–46} There was considerable variation in the measurement of diet across the included studies (Supplementary Table 1). While Elder and colleagues⁴¹ only assessed avoidance of fatty foods, Soto Mas et al.³⁷ measured multiple aspects of diet (including sodium and fat intake). Most studies assessed only some components of diet rather than using comprehensive measures of dietary intake (e.g., the Food Frequency Questionnaire).⁴⁹

Eight lifestyle interventions consisting of nutrition education workshops reported that at least one aspect of diet improved for intervention versus control group participants;^{32,34,37–39,41,42,46} five were CHW-led interventions. Improvements in dietary habits were maintained at follow-up among intervention group participants in most of these studies. Participants who received heart health information integrated into English as a Second Language classes had greater improvements in dietary habits than control group participants.^37,41 Although Khare et al.⁴⁶ found their 12-week lifestyle intervention led to improved dietary fat and fiber scores post-intervention (p = .01), only dietary fiber scores remained higher at follow-up (p = .03). They found no improvements in fruit and vegetable intake.⁴⁶ Using data from the Food Frequency Questionnaire,⁴⁹ Toobert et al.⁴⁵ found that a culturally adapted program to promote a Mediterranean diet in Latinos significantly reduced the percent of calories from saturated fat at 12-month follow-up (p < .01); these changes were maintained at 24-month follow-up (p < .05).⁴⁴ On the contrary, a CHW-led intervention conducted by Carrasquillo et al.³¹ found no differences in fruit and vegetable intake at 12-month follow-up between groups (measured using six items from the Behavioral Risk Factor Surveillance System).⁵⁰

Physical activity.

A total of 13 (76.5%) articles assessed physical activity as an outcome, however, only three reported significant improvements in physical activity among intervention group participants.^40,42,48 There was considerable variation in the measurement of physical activity across the included studies (Supplementary Table 1). Most (n = 9) assessed physical activity using self-reported measures (e.g., Stanford 7-Day Physical Activity Recall).⁵¹ Two studies used objective measures of physical activity by pedometry or accelerometry, but neither study included assessment of self-reported physical activity.^38,43

The impact of interventions on physical activity varied, with only two CHW-led interventions resulting in higher moderate or vigorous physical activity levels in Latinas throughout the study and at follow-up.^40,42 Of three workshop-based interventions,^37,42,46 only one significantly improved physical activity levels.⁴² In contrast, a 12-week lifestyle intervention had significant increases in physical activity levels immediately post-intervention (p < 0.01), but levels were similar to baseline at 12-month follow-up.⁴⁶ Two articles reporting findings of a family co-engagement intervention that included family health history risk information found that family co-engagement in physical activity increased from 11.4% at baseline to 15.7% at 10-month follow-up in the full sample (p < .001).^35,48 Husbands’ increased family health history risk was a significant motivator among women to increase their own physical activity at 3-month follow-up (p < .01), but this was not significant at 10-month follow-up.³⁵ Among men, their wives’ family health history risk had no effect on their physical activity.³⁵

Biological Outcomes

Anthropometric outcomes.

The main anthropometric measures assessed were BMI and changes in weight (in pounds). Twelve (70.6%) articles reported BMI as an intervention outcome,^{30,31,33,34,36,38,42–47} however, only four found significant decreases in BMI at follow-up.^33,44–46 Effective interventions included one CHW-led walking intervention and three lifestyle educational interventions.^44–46 Five studies (29.4%) reported results for changes in weight;^{33,34,38,41,43} yet only one study reported sustained weight loss for enhanced intervention versus minimal intervention participants after a 36-week walking intervention (mean 35.4 vs. 22.0, p <.001).³³

Blood pressure.

Among the nine (52.9%) articles that examined blood pressure, seven were CHW-led interventions.^{30–32,34,38,41–43,46} Regardless of intervention modality, few differences in blood pressure were observed between intervention and control group participants at follow-up. Intervention group participants in three studies, two of which were CHW-led, had significant reductions in systolic blood pressure compared to control group participants at 12-month follow-up.^31,42,46 Although Elder and colleagues⁴¹ found intervention group participants had greater reductions in systolic blood pressure at 3-month follow-up compared to controls (p <.001), this was not maintained at 6-month follow-up. Balcazar and colleagues³⁴ found that following an 8-week CHW-led intervention, the intervention group had higher diastolic blood pressure than the control group at 4-month follow-up (p <.001).

Metabolic outcomes.

Total cholesterol was assessed in eight (47.1%) articles.^{30,33,34,38,41–43,46} Differences in total cholesterol were found post-intervention between intervention and control group participants in three of these studies,^38,41,46 but those differences were not sustained at follow-up. The six (35.2%) articles, five of which were CHW-led, that assessed low-density lipoprotein cholesterol found no significant differences between intervention and control groups.^{31,33,34,38,43,46} Few differences in high-density lipid cholesterol were also identified.^{33,34,38,41–43} Compared to control group participants, intervention group participants significantly improved high-density lipid cholesterol in one study (p < .001).⁴¹ Although two CHW-led lifestyle interventions found improvements in high-density lipid cholesterol in both intervention and control participants, these differences were not statistically significant.^38,42

Glycemic status was assessed with glycosylated hemoglobin (HbA1c) or fasting blood glucose. Findings for HbA1c were reported in five (29.4%) articles^{31,34,43–45} and significant decreases among intervention group participants were reported in three of these studies.^31,44,45 In the three (17.6%) articles that assessed fasting blood glucose, no differences were observed between intervention and control group participants.^34,38,43

10-year CVD risk.

Four articles examined differences in 10-year CVD risk between intervention and control group participants.^34,42,44,45 Two of these articles reported findings of the same intervention at 12-month and 24-month follow-up.^44,45 Investigators found that 10-year CVD risk in intervention versus control group participants was lower at 12-month, but not 24-month follow-up.^44,45 Another lifestyle intervention that consisted of three CHW-led, face-to-face nutrition counseling sessions found a non-significant decrease (p = .26) in intervention group participants’ Framingham 10-year CVD risk score at 4-month follow-up compared to the control group that only received educational pamphlets about blood pressure and high cholesterol.³⁴

Knowledge and Attitudes

The three (17.6%) articles that reported on changes in knowledge found improvements among intervention group participants that were sustained at follow-up.^38,39,41 Elder and colleagues⁴¹ measured participants’ self-reported nutrition knowledge using a validated 12-item nutrition knowledge test. They found significant improvements (p <.001) in nutrition knowledge for intervention group participants with medium to high Spanish literacy following a heart health workshop combined with English as a Second Language classes. Similarly, a CHW-led intervention that combined information of heart and brain health resulted in a significant increase in knowledge about dietary fat (p =.03) at 1-month follow-up.³⁹ In addition, Koniak-Griffin et al.³⁸ found their CHW-led intervention led to significant improvements in knowledge of heart disease in the intervention group compared to the control group at 6-month follow-up (p <.001).

Furthermore, three (17.6%) studies examined differences in attitudes towards behavior change.^32,34,41 Using validated items from the Salud Para Su Corazón intervention,⁵² Balcazar and colleagues³⁴ CHW-led intervention resulted in significantly higher perceived susceptibility for CVD (p = .01) and perceived benefits of behaviors that would control CVD risk factors (p = .01) in intervention group participants (Supplementary Table 1). The same team of investigators found that another CHW-led intervention consisting of group workshops led to improvements in attitudes towards CVD risk control.³² However, control group participants who only received health education materials had greater improvements (p <.01) in perceived benefit of health behaviors on blood pressure control compared to the intervention group.³² Although two (11.8%) articles measured changes in attitudes about nutrition and blood pressure, neither reported these results.^34,41

Adherence and intervention fidelity

Participant adherence.

Only three articles, representing two studies, reported on measures used to assess participants’ adherence with intervention procedures.^38,44,45 Investigators in these studies used class attendance records to measure participant adherence.^38,44,45 Of these three, two reported on strategies to increase participant adherence and attendance at program sessions, including involvement of family members, providing travel, and sending reminders to participants who missed sessions.^44,45

Intervention fidelity.

Four studies reported on fidelity checks to ensure intervention staff delivered interventions as intended.^31,38–40 Carrasquillo et al.³¹ incorporated observation of intervention sessions and reviewed data from CHWs’ home visits and phone calls. Two studies had comprehensive plans to enhance fidelity.^38,40 These included, but were not limited to, providing CHWs with extensive training on the study protocol and intervention content, regular meetings with CHWs, observing CHWs’ participant visits, and collecting data on CHWs’ experiences delivering the intervention. Another study indicated fidelity checks were performed after every educational workshop during the study period, but did not provide details.³⁹

DISCUSSION

This is one of the first systematic reviews to synthesize evidence from RCTs that examined the effectiveness of behavioral interventions for CVD risk reduction among Latinos living in the U.S. Our findings can inform the development of future interventions for CVD risk reduction among this population. Overall, we found that behavioral interventions for CVD risk reduction in Latinos led to improvements in diet intake and eating habits. Although assessed in a smaller number of studies, we found significant improvements in psychosocial outcomes and knowledge of nutrition and heart disease. Most studies found no significant improvements in physical activity among intervention group participants. Similarly, with the exception of HbA1c, most studies found no significant improvements for other biological outcomes.

Methodological differences across the included studies did not appear to influence outcomes. Most studies in this review reported some degree of cultural adaptation in their methodology. Although the importance of cultural adaptation for CVD risk interventions for Latinos has been highlighted previously,²³ we did not find cultural adaptation influenced study outcomes. Our findings are consistent with a systematic review of 44 studies on diabetes self-management interventions that found cultural adaptation yielded mixed results.⁵³ Most interventions in our review were CHW-led, a method of cultural adaptation; yet few differences in outcomes were identified between CHW-led interventions and those that employed other methods, with the exception of changes in dietary habits. It is possible that other methodological considerations such as participant adherence, length of the intervention, fidelity, attrition, and follow-up period more strongly influenced outcomes. More research is needed that examines best methods for cultural adaptation of interventions to promote CVD risk reduction in Latinos in the U.S.

Although few included studies assessed psychosocial outcomes, our findings suggest that behavioral interventions for CVD risk reduction in Latinos in the U.S. may improve psychosocial outcomes. Targeting psychosocial factors in behavioral interventions has led to CVD risk reduction among adults. For instance, an RCT of 616 adults conducted in the United Kingdom found that greater social support was associated with an increase in heart healthy behaviors.⁵⁴ Further, a recent cluster, nurse-led RCT found that cardiovascular health promotion was associated with improvements in self-efficacy for physical activity among rural Black adults in the U.S.⁵⁵ We did not identify behavioral interventions that have targeted self-efficacy or social support to improve CVD risk among Latinos in the U.S. However, prior observational research has found that greater self-efficacy is positively correlated with healthy lifestyle behaviors (e.g., physical activity and weight loss) in this population.⁵⁶ Future studies should investigate how intervention modalities that target psychosocial factors can improve CVD risk among Latinos.

The majority of studies (72.7%) that examined diet found significant improvements among intervention participants. We identified no differences in dietary outcomes by intervention modalities or measures. Given that few studies in this review used comprehensive diet measures (e.g., 24-hour recall), it is possible that findings for diet were influenced by social desirability or recall bias. Research on measurement error in self-report diet measures among Latinos is limited, but previous work suggests that Latinos underestimate report of sodium intake by approximately 20% compared to 24-hour excretion of sodium.⁵⁷ Therefore, it is possible participants in this review may have overestimated improvements in diet. Future interventions should incorporate comprehensive self-reported measures of diet or objective measures (e.g., urinary excretion of sodium) in their study procedures to accurately assess changes in diet among Latino adults.

Less than 25% of the studies that examined physical activity reported any differences between intervention and control group participants. This was true regardless of intervention modalities used or the type of measurement. Further, analyses of population-based data in the U.S. indicate Latino adults are more likely to underreport their physical activity levels than non-Latino adults.^58,59 However, investigators also found Latinos had higher physical activity than non-Latino adults when objective measures of physical activity (e.g., actigraphy) were examined.^58,59 As all but two of the included studies used subjective measures to assess physical activity, it is likely that participants across included studies may have underreported their physical activity levels. Future behavioral interventions that target physical activity in Latinos should incorporate both subjective and objective measurements.

With the exception of HbA1c, the included studies reported generally null findings for biological outcomes. Less than half of the included studies found improvements in BMI (33.3%), changes in weight (20.0%), blood pressure (33.3%), low-density lipoprotein cholesterol (0.0%), high density lipid cholesterol, and fasting blood glucose (0.0%). Although only five studies assessed HbA1c, 60% found significant reductions in HbA1c in intervention relative to control group participants. Overall, findings for biological outcomes were not influenced by intervention methodology. For instance, the three studies that found significant improvements in HbA1c among intervention group participants differed in intervention methodology and cultural adaptation design. Moreover, most studies did not incorporate composite measures of biological CVD risk (e.g., Framingham risk score) or subclinical markers of CVD (e.g., pulse wave velocity and coronary artery calcification). Composite CVD risk scores and subclinical measures have been shown to improve CVD risk prediction beyond traditional risk factors alone.^60,61 Additional research is needed to determine whether behavioral interventions for CVD risk reduction lead to sustained changes in biological CVD risk factors among Latinos.

None of the included studies were nurse-led interventions, which have had beneficial effects on CVD risk factors in other populations.^62–64 Multiple RCTs of nurse-led interventions focused on Black adults have been effective at improving hypertension and diabetes outcomes.^65–68 Although nurse-led interventions have been shown to improve diabetes self-management among Latinos,⁵³ they remain underutilized to address other CVD risk factors in Latinos. As the largest group of health care providers worldwide, nurses play an important role in primary and secondary prevention of CVD. As the most-trusted health profession, nurses have a responsibility to enhance the well-being of individual patients and the society. Thus, nurses are uniquely positioned to design and lead culturally-adapted interventions that can improve the cardiovascular health of individuals and populations. Future studies should assess the effectiveness of nurse-led interventions to reduce CVD risk among Latinos.

None of the included studies used mobile health (mHealth) for CVD risk reduction among Latinos. mHealth is generally defined as the delivery of health information and interventions using smartphones and other mobile platforms.⁶⁹ A recent systematic review found that mHealth significantly improved CVD risk factors in the general population, including physical activity, diet, and weight loss.⁷⁰ Another review reported that use of mHealth had positive effects on management of CVD risk factors, particularly in low- and middle-income countries including Mexico and Honduras.⁷¹ mHealth may provide an opportunity to deliver behavioral interventions for CVD risk reduction as nearly 80% of Latino adults in the U.S. own a smartphone and 57% own a desktop/laptop computer.⁷² Yet, few mHealth resources are tailored to Latinos. For example, very few portals are available in languages other than English.⁷³ This prevents Spanish-speaking Latinos from accessing personal health and clinical data, including laboratory results and educational health resources, which we found was a motivator to increase physical activity. mHealth access disparities may exacerbate cardiovascular health disparities among Latinos. Testing mHealth interventions and methodologies adapted to the cultural, linguistic, and technical preferences of Latinos in the U.S. should be a research priority.

Our findings suggest that although cultural adaptation is important to consider, additional factors have yet to be accounted in interventions for CVD risk reduction among Latinos. Although Keller and Cantue³³ assessed how neighborhood characteristics influenced their intervention’s outcomes, few of the included studies incorporated other social determinants of health (e.g., poverty and health care access) as part of intervention strategies. Despite recognition that social determinants influence cardiovascular health in Latinos, interventions that directly address these factors are lacking. Future interventions should investigate how targeting social determinants impacts cardiovascular health among Latinos. Addressing social determinants of health has the potential for improving the cardiovascular health of Latinos. Similarly, most of the RCTs included in this review focused on individual-level interventions. Since the two family co-management interventions led to improvements in physical activity (especially among women),^35,36 future work should further examine the effectiveness of family co-management on CVD risk reduction in Latino adults.

Limitations

This systematic review has several weaknesses. First, there might be publication bias as we only included peer-reviewed articles. Second, we excluded RCTs that focused on self-management of CVD among Latinos with existing CVD diagnoses. Although many health behaviors to reduce CVD risk and improve self-management of CVD overlap, given the higher rates of CVD risk factors in Latinos, we deemed it was necessary to examine behavioral interventions for CVD risk reduction separately from interventions focused on self-management of CVD. Third, Latinos are a heterogenous group with various cultural beliefs, traditions, and experiences; yet most included studies (52.9%) had Latinos primarily of Mexican descent. This limits the generalizability of our findings to other Latino subgroups. Analyses of population-based data indicate that Caribbean Latinos have the highest rates of hypertension.⁴ Similarly, CVD risk among Latinos differs by age group and sex.⁷⁴ Future interventions should account for potential variations based on country of origin, sex, and other demographic characteristics.

Additional limitations related to the methodological quality of included studies were identified. Outcome measurements varied (validated, self-reported, etc.) across studies, which may impact results. Most RCTs included in this systematic review had less than 12-month follow-up. A previous behavioral intervention among non-Latina women found that at 48 months, participants in the intervention maintained lower weight, but there were no significant changes in lipid levels, blood pressure, glucose, or subclinical CVD measures.⁷⁵ Longer follow-up assessment using validated, objective measures is needed among Latinos samples to determine effective strategies to improve long-term cardiovascular health.

CONCLUSION

This systematic review of RCTs that examined the effectiveness of behavioral interventions for CVD risk reduction in Latinos living in the U.S. found that the majority of existing studies have been CHW-led interventions, aimed primarily at women of Mexican descent. Our results indicate that behavioral interventions were associated with improvements in diet, knowledge of CVD risk, and attitudes towards behavior change among Latinos. Future interventions aimed at CVD risk reduction in Latinos should also target psychosocial factors and assess composite biological measures associated with CVD risk in this population. We highlight the need to study the effectiveness of nurse-led and mHealth interventions to improve the cardiovascular health of Latinos. Furthermore, there is a need for research that addresses social determinants of health to identify factors other than culture that may improve and sustain cardiovascular health among Latinos.