There is a paucity of information on the long-term effects of multiple-risk-factor trials, especially in high-risk populations [15
]. The primary goal of this paper was to evaluate whether the effects of a multiple-behavior-change program, ¡Viva Bien
!, adapted for an underserved and high-risk Latina population, sustained improvements on targeted diabetes self-management, psychosocial variables, and biologic outcomes after the intervention faded.
The ¡Viva Bien
! intervention was initially intense, with a 2½-day retreat followed by 6 months of weekly 4-h meetings, but the study was also designed to maximize participant reach and generalizability of the outcomes. Few exclusion criteria were employed, and patients were recruited from two different health systems. The intervention components were delivered by a mix of bilingual clinical staff and community professionals and were evaluated systematically for their appropriateness and appeal to Latina participants in focus groups, a review by Latino/a professionals, and pilot testing prior to the intervention trial [40
]. Cultural appropriateness was assured by systematically giving participants opportunities to inform and influence intervention components to ensure cultural fit while maintaining fidelity to the original intervention. The finding that attrition rates were not associated with language preference or acculturation suggests that the program was culturally relevant to most Latinas in the sample.
While the 38.6% 24-month dropout rate was disappointing, attrition did not differ significantly by treatment condition or key participant characteristics, and attrition analyses indicated that those unavailable for the 24-month follow-up were generally similar to those who continued participation. Compared with our previous lifestyle-change study with mostly Anglo women [55
], the Mediterranean Lifestyle Program (MLP), from which ¡Viva Bien
! was adapted, program attendance was higher (65% for 0–6 months and 47% for 6–24 months in ¡Viva Bien
! vs. 54% for 0–6 months and 31% for 6–24 months). The ¡Viva Bien
! attendance rates are in line with those reported in similar studies with this population, such as those reported in a review of multifactorial lifestyle interventions to prevent chronic illness (diabetes and coronary heart disease) by Angermayr et al. [56
]. All studies in the review contained a stress-management component. In eight of the studies that explicitly reported attendance rates, most reported that participants attended more than 60% of the scheduled sessions.
The 24-month retention rate in ¡Viva Bien
! was lower than in the MLP (61.4% in ¡Viva Bien
! vs. 85.3% in the MLP). There are few studies with which to compare the ¡Viva Bien
! retention rates directly (i.e., multiple-risk-factor programs with Latinas having type 2 diabetes, at 24-month follow-up). Somewhat comparable studies report retention rates ranging from 56% to 100%, although mostly for much shorter follow-up times (with a notable exception [57
]). Silberman et al. [58
] reported that 78.1% of the participants remained enrolled in the program at the end of 1 year. In the Angermayr et al. review [56
], the proportion of participants terminating study interventions prematurely varied from 0% to 44%, depending on length of follow-up. At 24 months in the Diabetes Prevention Program Outcome Study [57
], 81% of participants were still enrolled. In a program by Eakin et al. [59
] targeting urban Latinos with multiple chronic conditions, the retention rate at 6 months was 81%. However, this intervention was much lower in intensity than either the MLP or ¡Viva Bien
!. A dietary intervention targeting Latinas [60
] found that at 12 months 79% of the original 357 study participants were available for follow-up analyses. Brown, Garcia, Kouzekanani, and Hanis [61
] reported a 12-month retention rate of 90% in their study involving a culturally competent lifestyle intervention targeting Mexican-Americans with type 2 diabetes. Poston et al. [62
] reported a 12-month retention rate of 66% in a study of 108 Mexican-American women testing a culturally tailored lifestyle modification intervention.
¡Viva Bien! employed a number of methods to maximize attendance and retention. These included family member involvement, expert presentations, practical, and relevant skill-based intervention components, friendly competitions, and active involvement. Social connection with staff and peers was strongly encouraged. Staff and peers made phone calls and sent cards to participants who missed sessions. Health benefits experienced in the program also exerted a strong influence to remain in the study. Transportation barriers were minimized by the provision of taxi service and by locating meetings in convenient community settings. Of those giving a reason for non-attendance, the most frequent was illness (9.9%), followed by work-related conflicts (4.5%), being on vacation (4.2%), or having a social conflict (3.7%). Anecdotal evidence from exit interviews suggested that as the in-person intervention sessions faded, intervention condition participants felt a diminished responsibility to themselves and to the program. Future studies in this population might improve adherence/retention by introducing maintenance strategies from the beginning of the program, maintaining the level of intensity throughout the intervention period and offering greater incentives for follow-up assessment participation.
Despite missing data arising from attrition, results of generalized estimating equation analyses using complete-case and imputed data were similar. ¡Viva Bien! participants compared with usual care significantly improved by the 6-month follow-up on the psychosocial measures of problem solving, self-efficacy, and perceived support, and these gains were maintained at 12- and 24-month follow-ups. Between-condition effects were found across 24 months on dietary patterns (percent calories from saturated fat) and use of social–environmental resources to support chronic disease self-management. Improvements in stress management and days per week exercised were sustained by ¡Viva Bien! participants; however, there were improvements in these factors by usual care participants by the end of 24 months as well, thus diminishing the longitudinal intervention effect. Effects of ¡Viva Bien! on biological outcomes were initially promising. These effects were not maintained, however, as participants in both conditions approached baseline levels by 24 months.
Although the treatment condition improved across the 24-month intervention, the challenge in sustaining physical activity is well documented [63
], and sustaining stress management over long periods is virtually undocumented. For behaviors that are not currently part of daily living, such as physical activity or stress management (as opposed to eating) to become habitual, it may be that the intervention must address not only individual behaviors, but also the home, neighborhood, job, and social environments, which may increase stress and promote sedentary lifestyles. While these environments may also encourage unhealthful eating, it is possible that changing eating habits (e.g., modifying food preparation techniques to reduce saturated fat) requires less effort to sustain, once mastered, than a daily walk or stress-management practices [65
]. Recent evidence suggests that people with diabetes experience greater perceived exertion than those without diabetes, necessitating greater motivational and supportive resources than the general population to maintain physical activity [66
]. And, in general, participation in regular physical activity among Hispanics tends to be lower than in non-Hispanic whites [67
]. Thus, an ecological approach that includes personal, intrapersonal, organizational, and environment/policy changes to support healthful lifestyles, particularly among populations with increased risk and disproportionate burden for chronic conditions such as diabetes, may be required for sustained change in multiple health behaviors.
This practical trial has both methodological strengths and weaknesses, as well as implications for practice, policy, and research [68
]. Strengths include an intervention that addressed a multiple-risk-factor problem with a previously tested multiple-behavior-change intervention, cultural adaptation for a Latina population, a reasonably large and diverse Latina sample, use of multiple measures, multiple imputation procedures for handling missing data, and generalized estimating equation analyses across 24 months.
A possible limitation of our study was the use of primarily self-report measures for behavioral outcomes. Most of these measures have been validated against more objective standards in previous studies, but it is not known whether possible self-reporting inaccuracies influenced the observed results. The correlation (r
0.34) in this study between the Modified International Physical Activity Questionnaire and accelerometer data, though moderate, is similar to findings of other studies correlating self-reports of activity and more objective indicators (e.g., our study [70
] correlating 7-day pedometer step counts 0.31 with a 7-day diary physical activity in a sample of older women). Such results suggest that the two modes of physical activity measurement (self-report and pedometer/accelerometers) provide both common and unique information. Another limitation was that the comprehensive lifestyle intervention cannot be easily disentangled to understand the contribution of discrete elements. The efficacy of the core components of the lifestyle intervention (diet, physical activity, stress management, and social support) have been well established in the literature; less understood is how they work in concert.
The short-term effectiveness of multibehavioral interventions, such as ¡Viva Bien
!, is encouraging, but the inability of behavioral interventions to sustain behavior change or improvements in biologic or quality-of-life outcomes over time continues to be a challenge [71
]. More research is needed to develop and test novel and cost-effective interventions that can sustain the motivation and support needed to accomplish successful, lifelong diabetes self-management. Research that investigates characteristics of social and physical environments associated with sustaining multiple-health-behavior change is sorely needed [72
The ¡Viva Bien! program succeeded in changing multiple key health behaviors. As contact faded after 6 months, so did the intervention effects. For people with chronic conditions, maintenance interventions that provide ongoing social, motivational, and problem-solving support may be as important as addressing long-term medication adherence. Such interventions may be best delivered through community organizations (e.g., faith-based, worksites, schools), neighborhoods (e.g., infrastructures that support activity; neighborhood groups that support norms for health), and technology (e.g., phone, social media, internet).