1. Introduction
Obesity contributes to poor health outcomes [1], a lower quality of life [2], and higher healthcare costs [3]; the direct medical costs due to obesity in adults in the United States are more than USD 260.6 billion per year [3]. The successful behavior change interventions for obesity typically include increased physical activity, improved diet quality, goal setting, and self-monitoring [4]. These types of intervention can result in weight loss averaging 8 kg or 5–10% of their body weight [4,5]. Participants also often improve their cardiovascular disease risk factors (e.g., hemoglobin A1c [HbA1c], blood pressure, and triglycerides) and quality-of-life measures (e.g., improvements in physical function, self-esteem, and psychosocial function and reduced symptoms of depression) [4].
Due to the complex causes of obesity, multilevel interventions that aim to change behaviors and address health outcomes by affecting more than one level of influence within the socioecological model are recommended [6]. Over the last two decades, multilevel interventions to address obesity have expanded greatly, and the results are promising, with most multilevel interventions improving the measures of adiposity (e.g., body weight, body mass index [BMI], and waist circumference) and/or obesity-related behaviors (e.g., physical activity and diet) [6]. Despite the significant benefits of behavior change interventions on obesity and other health outcomes, many studies result in the significant heterogeneity in outcomes [7]. The heterogeneity in treatment responses refers to the idea that the true effect of a treatment on one individual will be different than the effect on a different individual [8].
There are many potential sources of the heterogeneity in treatment responses beyond adherence to the intervention, including complex biological reasons (e.g., resting energy requirements based on metabolic programming and the impact of gastrointestinal peptides and the gut biome); the bi-directional relationship between mental health and weight loss; and the social determinants of health (e.g., having had access to high-quality education; economic stability; the social context; and the built environment). Detailing these factors will allow for more accurate predictions of intervention success and more effectively tailored strategies [9]. The analysis of 32 adult obesity studies found that 12 of them showed evidence that treatment heterogeneity was present, and of those 12, 3 demonstrated an estimated proportion of >5% of the sampled population having an outcome opposite to the mean effect (e.g., weight gain instead of weight loss) [8].
In a recent review of multilevel interventions targeting obesity, fewer than one in ten interventions reported differences in the outcomes across sociodemographic factors, and those that did showed differences in effectiveness by race, ethnicity, obesity status, age, and sex [6]. Much remains to be understood about which individuals benefit from lifestyle behavior interventions, with critical implications for dissemination and addressing health equity. Understanding who benefits the most (or least) from an intervention is explored through moderator analyses [10]. Strong Hearts, Healthy Communities-2.0 (SHHC-2.0) is a multilevel intervention that aimed to reduce the cardiovascular disease risk among rural women through targeting behavioral determinants at the individual, social, and environmental levels. SHHC-2.0 has demonstrated improvements in cardiovascular disease risk, [11] that were maintained at a 6-month follow-up [12], while increasing physical activity levels [13] and improving diet quality [13,14]. This report’s objective is to explore the moderation effects of the sociodemographic and baseline physical and mental health characteristics of SHHC-2.0 participants on selected health outcomes (weight, systolic blood pressure, HbA1c, and triglycerides).
2. Methods
This community-randomized control trial was conducted between January 2017 and August 2018. Eleven rural, medically underserved communities in New York were pair-matched based on the Rural–Urban Commuting Area (RUCA) code and population size, and then randomized into the SHHC-2.0 intervention or a delayed-intervention control condition, in which intervention delivery followed a 24-week primary outcome period. The participants were female and >40 years old at enrollment and were required to have a BMI > 30, or a BMI 25–30 plus < 1 weekly occurrence of 30 min of leisure-time physical activity. A full description of the study design and related effectiveness reports have been published elsewhere [11,12].
2.1. Intervention
The SHHC-2.0 intervention included 60-min classes delivered two times per week for 24 weeks, led by local health educators. The intervention was built around goals and behavioral aims for both diet and physical activity. Specifically, the physical activity behavioral aims included increasing strength training to two or more times per week, increasing aerobic activity to at least five times per week, and to minimize sedentary time. The dietary aims included eating fewer calories, saturated and trans fats, processed foods, desserts, and sugar-sweetened beverages and less sodium and increasing fruit, vegetable, and whole grain consumption. The social environment’s influence on diet and exercise behavior was engaged and reinforced through concepts such as social support for healthy behaviors; healthy eating plans for friends and family; social influences on sugar-sweetened beverage consumption; and what to do when loved ones are unsupportive. The intervention’s Change Clubs engaged participants to identify an issue relevant to them in their community context and facilitate the implementation of an action plan that would affect the social, cultural, environmental, and political factors within each of their communities.
2.2. Measures
At baseline and the end of the intervention (24 weeks), the participants completed a data collection visit that included anthropometric and biometric measurements (venipuncture, weight, height, and blood pressure), functional fitness tests, and a survey of self-reported behaviors and psychosocial variables.
2.2.1. Outcomes
The primary outcome for this study was change in body weight; the secondary outcomes included herein due to an observed intervention effect were systolic blood pressure, triglycerides, and HbA1c. The equipment used included Omron (Kyoto, Japan) HBF-510W scales for body weight and Omron automated devices for blood pressure. Fasting blood specimens were collected by phlebotomists and registered nurses to measure triglycerides and HbA1c.
2.2.2. Moderators
Baseline measures were used as the potential moderators for sociodemographic characteristics, physical health, and mental health. The sociodemographic characteristics included the following: age, employment status, household income (<USD 50,000 versus ≥ USD 50,000), education (some college or higher versus high school or less), and relationship status (yes or no). The SHHC-2.0 participants were predominantly White, non-Hispanic (97.6%); therefore, no moderation was explored by race or ethnicity. The potential baseline physical health moderators included measured BMI; the Short Form 36 (SF36) health-related quality-of-life survey scale [15]; and a self-reported history of arthritis, cancer, diabetes, heart disease, high blood cholesterol levels, high blood sugar levels, or hypertension. The potential baseline mental health moderators included the Generalized Anxiety Disorder Scale (7 items) [16]; the Brief Resilience Scale (6 items) [17]; the Perceived Stress Scale (10 items) [18]; the Patient Health Questionnaire (8 items, maximum 24-point score) and the moderate-or-higher-depression-severity classification (>10 on Patient Health Questionnaire) [19].
2.3. Statistical Analyses
The effect of moderation was tested with an interaction term between the intervention and potential moderator on the difference in change from 0 to 24 weeks between groups using linear, mixed-effects, multilevel models. These models included random cluster (community) effects to account for community-level randomization and correlation between participants in the same community. The covariates in the models determined a priori included age and education. Intention-to-treat analyses that included all the participants as randomized were conducted, regardless of the number of assessments performed or intervention attendance. For this analysis, one-sided tests were used to explore the impact of moderators on improvements in each outcome (e.g., weight loss) with a 95% confidence interval (95%CI). While two-sided tests are often automatically used, one-sided tests are warranted when the direction of the outcome is predetermined (e.g., improvements) and is more liberal in identifying potential moderators, which is of importance for not missing the potential moderators that could impact intervention refinement and dissemination [20]. Complete case analysis was used with a restricted maximum likelihood and incorporation of all the available data. For the outcome variables, no data were missing at the baseline; 50 participants (27%) withdrew or did not complete the data collection visit at 24 weeks. For the moderators, the degree of missingness was 5% at the baseline and 37% at 24 weeks. All analyses were conducted in 2023–2024 using SAS, version 9.4. Dr. Rebecca A. Seguin-Fowler, as this study’s Principal Investigator, takes responsibility for integrity of the data and data analysis.
3. Results
Five communities were randomized into the intervention group (n = 87 participants), and six communities were randomized to the delayed intervention control group (n = 95 participants) (see Figure 1). The baseline characteristics for the total sample and each group are presented in Table 1.
Moderation Effects
The moderation effects of the sociodemographic information and the baseline physical and mental health on weight loss, systolic blood pressure reduction, HbA1c reduction, and triglyceride reduction are presented in Table 2.
Weight loss. Greater weight loss was observed for the participants who had some college education compared to those who were educated at a high school level or less (−10.04 kg, 95%CI from −13.63 to −6.44); there were no significant moderation effects for any other sociodemographic potential moderator. Depression was also associated with differential weight loss, as those with higher baseline depression symptomatology demonstrated greater weight loss (−0.52 kg, 95%CI from −0.82 to −0.22), and those with moderate or higher depression severity had a weight loss of 6.45 kg greater than those with mild-to-no depression (95%CI from −10.30 to −2.60).
Systolic blood pressure. The participants who had a history of hypertension had a greater reduction in systolic blood pressure compared to those who did not (−7.71 mmHg, 95%CI from −13.95 to −1.47). No other moderator had a significant effect on systolic blood pressure.
Hemoglobin A1c. For the reduction in HbA1c, greater improvements were seen for the participants who had a history of cancer compared to those who did not (−0.28%, 95%CI from −0.55 to −0.02) and baseline moderate or higher depression severity compared to those who did not (−0.31%, 95%CI from −0.57 to −0.05).
Triglycerides. The sociodemographic characteristics associated with greater reductions in triglycerides included the participants who were employed (−53.49 mg/dL, 95%CI from −72.93 to −34.04) and those in a relationship (−32.49 mg/dL, 95%CI from −50.76 to −14.22). Across the health status moderators, the participants with a history of cancer also demonstrated a greater reduction in triglycerides (−48.72 mg/dL, 95%CI from −80.61 to −16.83). Higher health-related quality-of-life (−0.67 mg/dL, 95%CI from −1.12 to −0.22) and resilience scores (−36.00 mg/dL, 95%CI from −50.90 to −21.09) were associated with greater improvements in the level of triglycerides.
4. Discussion
In our analysis, the majority of the moderators evaluated did not have a significant effect on the health outcomes. As noted, most published papers on multilevel interventions did not evaluate the potential moderators of intervention effectiveness; thus, our findings have the potential to not only improve the future dissemination of the SHHC-2.0 intervention, but can also more broadly inform the development and evaluation of other multilevel interventions. While many sociodemographic characteristics were not associated with differential outcomes, we did find that those with higher educational attainment experienced greater weight loss; our data also showed that being employed and being in a relationship were associated with greater improvements in the level of triglycerides. Depression status was found to be a significant moderator for the primary outcome of weight loss, as the participants with worse depressive symptoms demonstrated greater weight loss. Higher depressive symptomatology was also associated with greater improvements in the HbA1c levels.
Women with moderate or higher depression severity at the baseline saw greater improvements in weight loss and reductions in the HbA1c level compared to those of their counterparts. Furthermore, similar trends were observed for systolic blood pressure and triglycerides, though these did not reach significance (p = 0.06). This finding is particularly promising given the complex and bi-directional relationship between weight and mental health. In a meta-analysis of longitudinal studies, there was an increased risk of depression for people who are overweight (27% increased risk) or obese (55%), while baseline depression increased the risk of the development of obesity by 58% [21]. The effect of weight loss interventions among persons with depression is unclear. Legenbauer et al. reported persons with depression had no differences in weight loss following a behavioral weight loss intervention [22]. However, other studies have reported poor weight loss outcomes among those with depression. For example, in a study of rural black women, depression was associated with less weight loss following a behavioral weight loss program [23]. The results of the current analysis suggests that the effectiveness of the SHHC 2.0 intervention is not impacted by depression status.
The prior studies have also identified depression status as negatively impacting behavior changes and the adherence to behavior change interventions [24,25,26,27,28]. To that end, we explored if women with elevated depressive symptoms at the baseline engaged with the intervention differently via attendance and intervention tool use (e.g., health journal, participant binder, Fitbit, digital scale, or exercise DVDs) and found no significant differences (results not reported). We could posit that there may not be a single facet of the intervention that was most impactful for these women, but in fact the multi-component, multilevel approach may have addressed the various needs of these women. However, these exploratory analyses were limited by a small sample size of women with baseline moderate or higher depression severity at the baseline (n = 9). One relevant finding for this subgroup is that 77% (n = 7) of the women who were classified as having a moderate or higher depression severity at baseline no longer met that categorization at the end of the intervention. This is in line with prior research indicating that weight loss can treat depressive symptoms, whether weight loss is achieved from surgery or through behavioral changes [29,30,31], and that changes in physical activity and diet have also shown to improve depressive symptoms [32,33].
Our results also indicated that the participants with higher educational attainment saw greater weight loss, while being employed was associated with a greater improvement in triglycerides. The findings from a systematic review on how different aspects of inequality, including education and employment, impact uptake, adherence, and the effectiveness of trials of behavioral weight management interventions, suggest that the education can have varying impacts on weight loss [34]. Thus, it is possible that these factors reflect structural barriers experienced by those with fewer resources (e.g., participants with less education may need the program to be tailored and/or additional social or environmental supports), highlighting a potential opportunity to address and mitigate the social determinants of health for participants in the context of these types of program.
The participants that reported being in a relationship also demonstrated greater improvements in the level of triglycerides. While the SHHC 2.0 intervention aimed to increase social support both among the participants and with friends and family, it is possible that more focused strategies to increase social support within and beyond the class setting and from a range of social network members (e.g., coworkers and faith-based community members) might be particularly beneficial for those not in a relationship.
While the moderators discussed above warrant further research, we note that most of the factors evaluated did not have a significant effect, or the effects were not consistent across all the health outcomes. The lack of heterogeneity in treatment responses across the subgroups may be due to the strong evidence base of the program [35], extensive tailoring via multiple rounds of formative research, and subsequent intervention refinement via process evaluations [36,37]. Additionally, health improvement is not one-size-fits-all, and having a multi-component (focused on diet and physical activity), multilevel (focused on individual behavior change, social support, and environmental changes) intervention creates multiple avenues for behavior changes and health improvements. This likely minimized the heterogeneity of intervention impact across the participant characteristics.
There are limitations to our inferences due to the exploratory nature of this analysis. The original study was not powered for moderation analysis; thus, null findings may be due to a lack of power resulting from small subgroups, while significant findings could be the result of small comparison groups. In part, that is why we interpreted our findings with one-sided testing with established improvements in the outcomes (e.g., weight loss) to be more liberal in identifying the potential moderators that could impact intervention refinement and dissemination. Even so, the consistency and lack of moderation in the outcomes may also be due to the rural, medically underserved towns and participants, reflecting their predominantly White populations. Future studies in more racially and ethnically diverse communities are needed to truly assess the consistency in outcomes observed with SHHC 2.0. Furthermore, it is possible that the other factors not assessed in this study (e.g., community-level factors, sleep, and health beliefs/attitudes) could moderate the intervention effects. The rates of missing data may also impact our findings. At 24 weeks, 50 participants (27%) withdrew or did not complete the data collection visit, so it is possible that the missing data could impact the reliability of our results. The primary concern was that data may not be missing at random and that participants with worse health might be more likely to drop out or not report. To explore this potential bias, the baseline characteristics of the respondents and the non-respondents were compared at 24 weeks. No significant differences were observed for age, income, education, race, BMI, weight, physical activity levels, or self-reported perceived overall health. These findings do not indicate systemic bias resulting from missing data, though we note that unobserved differences between the groups may exist. We chose to report this complete case analysis rather than conduct multiple imputation (MI) given the exploratory nature of this report, which did not require more representative parameter estimates (a strength of MI). Future studies with larger samples would benefit from exploring non-linear moderation effects as well. The strengths include exploring multiple health outcomes (e.g., weight loss, reduction in systolic blood pressure, reduction in HbA1c, and reduction in triglycerides) and a variety of moderators, specifically the inclusion of baseline mental health.
5. Conclusions
SHHC is effective [11,12] and cost-effective [38], and based on these findings, likely beneficial for a wide range of participants, making it well poised for wider dissemination and impacts. These findings are critical for intervention-specific dissemination and as a demonstration for future interventions to critically examine the impacts across participant characteristics to ensure interventions are optimizing the opportunities to help address health equity. Further efforts to broaden the dissemination of the SHHC-2.0 intervention could also include adaptations for other populations (i.e., women living in urban/suburban areas), or alternative delivery strategies (i.e., digital or hybrid intervention delivery). Another area for future research is related to the maintenance of the observed effects. We note that the SHHC-2.0 intervention effects were maintained over a 6-month follow-up period [12], but studies with a longer follow-up period are needed to demonstrate the long-term maintenance of the intervention effects. Finally, the SHHC-2.0 intervention may have positive health impacts beyond those evaluated in the current analysis (i.e., mental health improvements) that can be evaluated in future studies.
Conceptualization, R.A.S.-F., C.D.R., M.M.D., M.L.G.; methodology, M.M.D., S.H.; formal analysis, S.H.; writing—original draft preparation, M.M.D., C.D.R.; writing—review and editing, R.A.S.-F., M.L.G., C.D.R., M.M.D., S.C.F., G.D.E.; funding acquisition, R.A.S.-F., S.C.F. All authors have read and agreed to the published version of the manuscript.
This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Cornell University (IRB protocol code 1402004505; approved on 22 May 2015) and Bassett Healthcare (IRB protocol code: 2022; approved on 28 October 2015).
Informed consent was obtained from all the subjects involved in this study.
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to use of human subject data.
The authors wish to thank the Strong Hearts, Healthy Communities leaders, participants, and communities.
Rebecca A. Seguin-Fowler is a co-founder of strongpeopleprogram.org. No other authors have conflicts of interest to declare. The funders had no role in the design and conduct of this study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Strong Hearts, Healthy Communities-2.0 participant characteristics at baseline.
| Total | Control | Intervention | |
|---|---|---|---|
| Age (n = 182), years ±SD | 57.2 ± 9.0 | 55.9 ± 8.5 | 58.5 ± 9.3 |
| Employment, working (n = 130), n (%) | |||
| Working | 120 (92.3) | 62 (92.5) | 58 (92.1) |
| Not working | 10 (7.7) | 5 (7.5) | 5 (7.9) |
| Income, household annual (n = 162), n (%) | |||
| <USD 50,000 | 66 (40.7) | 32 (19.7) | 34 (21.0) |
| ≥USD 50,000 | 96 (59.3) | 130 (80.3) | 128 (79.0) |
| Education, some college or more (n = 172), n (%) | |||
| HS or less | 35 (20.3) | 17 (19.5) | 18 (21.2) |
| Some college or more | 137 (79.7) | 70 (80.5) | 67 (78.8) |
| Relationship status (n = 171), n (%) | |||
| In a relationship | 116 (67.8) | 62 (71.3) | 54 (64.3) |
| Not in a relationship | 55 (32.2) | 25 (28.7) | 30 (35.7) |
| Body Mass Index, kg/m2 | 36.7 ± 7.8 | 37.9 ± 8.5 | 35.4 ± 6.8 |
| History of arthritis (n = 170), n (%) | 70 (41.2) | 39 (44.8) | 31 (37.3) |
| History of cancer (n = 170), n (%) | 12 (7.1) | 5 (5.7) | 7 (8.4) |
| History of diabetes (n = 170), n (%) | 25 (14.7) | 17 (19.5) | 8 (9.6) |
| History of heart disease (n = 170), n (%) | 10 (5.9) | 5 (5.7) | 5 (6.0) |
| History of high blood cholesterol (n = 170), n (%) | 71 (41.8) | 33 (38.4) | 38 (45.2) |
| History of high blood sugar (n = 170), n (%) | 37 (21.8) | 16 (19.3) | 21 (24.1) |
| History of hypertension (n = 170), n (%) | 71 (41.8) | 41 (47.7) | 30 (35.7) |
| Short Form health-related quality of life (SF36) (n = 171), 36 items | 81.6 ± 20.1 | 81.0 ± 20.8 | 82.3 ± 19.4 |
| Generalized Anxiety Disorder Scale (n = 171), 7 items | 2.6 ± 3.3 | 2.8 ± 3.3 | 2.4 ± 3.3 |
| Brief Resilience Scale (n = 171), 6 items | 3.8 ± 0.6 | 3.9 ± 0.6 | 3.7 ± 0.7 |
| Perceived Stress Scale (n = 170), 10 items | 4.8 ± 3.0 | 4.9 ± 3.2 | 4.8 ± 3.2 |
| Patient Health Questionnaire (Depression) (n = 171), 8 items | 4.4 ± 3.9 | 5.0 ± 4.0 | 3.9 ± 3.8 |
| Moderate or higher depression severity (n = 171), (≥10 on Patient Health Questionnaire) | 19 (11.1) | 11 (12.6) | 8 (9.5) |
Notes: For dichotomous variables, count and % of each sample are reported for level listed (e.g., employment, working). For continuous variables, mean ± SD is reported.
Moderation effects of between-group changes in outcomes from baseline to intervention end point (24 weeks).
| Weight Loss (kg) | Reduction in Systolic Blood Pressure (mmHg) | Reduction in Hemoglobin A1c (%) | Reduction in Triglycerides (mg/dL) | |||||
|---|---|---|---|---|---|---|---|---|
| Potential Moderators | n | Est (95% CI) p-Value | n | Est (95% CI) p-Value | n | Est (95% CI) p-Value | n | Est (95% CI) p-Value |
| Age, years | 131 | −0.08 (−0.20,0.03) p = 0.121 | 130 | 0.05 (−0.31,0.40) p = 0.587 | 130 | 0.00 (0.00,0.01) p = 0.799 | 129 | 1.45 (0.49,2.41) a p = 0.993 |
| Employment, working | 97 | 3.37 (0.97,5.76)a p = 0.989 | 96 | −0.83 (−8.07,6.41) p = 0.425 | 98 | −0.10 (−0.26,0.06) p = 0.145 | 98 | −53.49 (−72.93,−34.04) p < 0.001 |
| Income, ≤USD 50 k | 121 | 1.04 (−3.51,5.60) p = 0.354 | 120 | 0.55 (−0.96,2.06) p = 0.724 | 120 | 0.02 (−0.26,0.29) p = 0.539 | 119 | −13.48 (−46.33,19.38) p = 0.251 |
| Education, some college or more | 129 | −10.04 (−13.63,−6.44) p < 0.001 | 128 | −4.85 (−16.49,6.79) p = 0.248 | 128 | 0.07 (−0.18,0.32) p = 0.676 | 127 | 21.12 (−9.57,51.80) p = 0.870 |
| In a relationship status | 128 | 2.52 (0.27,4.77)a p = 0.966 | 127 | 2.23 (−4.49,8.95) p = 0.707 | 127 | 0.11 (−0.04,0.25) p = 0.880 | 126 | −32.49 (−50.76,−14.22) p = 0.002 |
| Body Mass Index, kg/m2 | 131 | 0.03 (−0.11,0.17) p = 0.623 | 130 | 0.13 (−0.29,0.55) p = 0.692 | 130 | −0.01 (−0.02,0.00) p = 0.092 | 129 | 1.15 (0.01,2.30)* p = 0.950 |
| History of arthritis | 127 | 1.41 (−0.84,3.67) p = 0.848 | 126 | −5.09 (−11.61,1.43) p = 0.101 | 126 | −0.12 (−0.27,0.03) p = 0.089 | 125 | 0.20 (−17.88,18.27) p = 0.507 |
| History of cancer | 127 | 2.32 (−1.77,6.40) p = 0.824 | 126 | −4.76 (−16.31,6.78) p = 0.250 | 126 | −0.28 (−0.55,−0.02) p = 0.041 | 125 | −48.72 (−80.61,−16.83) p = 0.006 |
| History of diabetes | 127 | −1.51 (−5.21,2.18) p = 0.252 | 126 | 7.59 (−3.01,18.19) p = 0.879 | 126 | 0.76 (0.54,0.99)a p = 1.000 | 125 | 22.89 (−4.78,50.56) p = 0.912 |
| History of heart disease | 127 | −1.60 (−6.50,3.31) p = 0.297 | 126 | −12.53 (−26.53,1.46) p = 0.072 | 126 | 0.30 (−0.03,0.62) p = 0.934 | 125 | 21.35 (−17.43,60.13) p = 0.816 |
| History of high blood cholesterol | 127 | 2.00 (−0.21,4.22) p = 0.930 | 126 | 3.85 (−2.43,10.13) p = 0.842 | 126 | 0.01 (−0.14,0.16) p = 0.545 | 125 | −1.69 (−20.01,16.63) p = 0.440 |
| History of high blood sugar | 127 | −2.31 (−4.96,0.35) p = 0.078 | 126 | 6.9 (−0.81,14.62) p = 0.928 | 126 | 0.25 (0.08,0.43)a p = 0.991 | 125 | −17.30 (−38.21,3.61) p = 0.088 |
| History of hypertension | 127 | 0.17 (−2.06,2.39) p = 0.549 | 126 | −7.71 (−13.95,−1.47) p = 0.022 | 126 | 0.16 (0.01,0.31)a p = 0.958 | 125 | −11.54 (−30.00,6.93) p = 0.153 |
| Short Form health-related quality of life, 36 items | 128 | −0.02 (−0.07,0.03) p = 0.286 | 127 | −0.04 (−0.19,0.11) p = 0.335 | 127 | 0.00 (0.00,0.00) p = 0.284 | 126 | −0.67 (−1.12,−0.22) p = 0.008 |
| Generalized Anxiety Scale, 7 items | 128 | −0.22 (−0.56,0.12) p = 0.142 | 127 | 0.07 (−0.87,1.01) p = 0.549 | 127 | −0.02 (−0.04,0.00) p = 0.085 | 126 | 0.73 (−1.98,3.43) p = 0.670 |
| Brief Resilience Scale, 6 items | 128 | 0.07 (−1.78,1.92) p = 0.525 | 127 | 4.12 (−1.24,9.47) p = 0.896 | 127 | −0.07 (−0.19,0.05) p = 0.181 | 126 | −36.00 (−50.90,−21.09) p < 0.001 |
| Perceived Stress Scale, 10 items | 127 | −0.04 (−0.24,0.16) p = 0.376 | 126 | −0.16 (−0.74,0.42) p = 0.326 | 126 | 0.00 (−0.02,0.01) p = 0.394 | 125 | 2.33 (0.71,3.94) a p = 0.991 |
| Patient Health Questionnaire—8 items | 128 | −0.52 (−0.82,−0.22) p = 0.002 | 127 | −2.75 (−7.48,1.99) p = 0.171 | 127 | −0.02 (−0.04,0) p = 0.105 | 126 | 2.06 (−0.46,4.59) p = 0.909 |
| Moderate or higher depression severity (≥10 on Patient Health Questionnaire) | 128 | −6.45 (−10.30,−2.60) p = 0.003 | 127 | −10.19 (−21.41,1.03) p = 0.069 | 127 | −0.31 (−0.57,−0.05) p = 0.024 | 126 | −8.27 (−17.09,0.55) p = 0.063 |
Notes: All estimates are from complete case models and adjusted for random cluster (community) effects, assignment group, age, and education. Confidence intervals and p-values are adjusted for one-sided tests on effect of moderator on outcomes of weight loss, reduction in systolic blood pressure, reduction in hemoglobin A1c, and reduction in triglycerides. a indicates 95%CI that is in opposite direction of one-sided test. Bold indicates p value < 0.05.
References
1. Ansari, S.; Haboubi, H.; Haboubi, N. Adult obesity complications: Challenges and clinical impact. Ther. Adv. Endocrinol. Metab.; 2020; 11, 2042018820934955. [DOI: https://dx.doi.org/10.1177/2042018820934955] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32612803]
2. Kolotkin, R.L.; Andersen, J.R. A systematic review of reviews: Exploring the relationship between obesity, weight loss and health-related quality of life. Clin. Obes.; 2017; 7, pp. 273-289. [DOI: https://dx.doi.org/10.1111/cob.12203] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28695722]
3. Cawley, J.; Biener, A.; Meyerhoefer, C.; Ding, Y.; Zvenyach, T.; Smolarz, B.G.; Ramasamy, A. Direct medical costs of obesity in the United States and the most populous states. J. Manag. Care Spec. Pharm.; 2021; 27, pp. 354-366. [DOI: https://dx.doi.org/10.18553/jmcp.2021.20410]
4. Wadden, T.A.; Tronieri, J.S.; Butryn, M.L. Lifestyle modification approaches for the treatment of obesity in adults. Am. Psychol.; 2020; 75, 235. [DOI: https://dx.doi.org/10.1037/amp0000517]
5. Elmaleh-Sachs, A.; Schwartz, J.L.; Bramante, C.T.; Nicklas, J.M.; Gudzune, K.A.; Jay, M. Obesity management in adults: A review. JAMA; 2023; 330, pp. 2000-2015. [DOI: https://dx.doi.org/10.1001/jama.2023.19897]
6. Seguin-Fowler, R.A.; Graham, M.L.; Demment, M.; Uribe, A.L.M.; Rethorst, C.D.; Szeszulski, J. Multilevel interventions targeting obesity: State of the science and future directions. Annu. Rev. Nutr.; 2024; 44, pp. 357-381. [DOI: https://dx.doi.org/10.1146/annurev-nutr-122123-020340]
7. Zoh, R.S.; Esteves, B.H.; Yu, X.; Fairchild, A.J.; Vazquez, A.I.; Chapple, A.G.; Brown, A.W.; George, B.; Gordon, D.; Landsittel, D. et al. Design, analysis, and interpretation of treatment response heterogeneity in personalized nutrition and obesity treatment research. Obes. Rev.; 2023; 24, e13635. [DOI: https://dx.doi.org/10.1111/obr.13635]
8. Kaiser, K.A.; Gadbury, G.L. Estimating the range of obesity treatment response variability in humans: Methods and illustrations. Hum. Hered.; 2013; 75, pp. 127-135. [DOI: https://dx.doi.org/10.1159/000351738]
9. Dent, R.; McPherson, R.; Harper, M.-E. Factors affecting weight loss variability in obesity. Metabolism; 2020; 113, 154388. [DOI: https://dx.doi.org/10.1016/j.metabol.2020.154388]
10. Kraemer, H.C.; Wilson, G.T.; Fairburn, C.G.; Agras, W.S. Mediators and moderators of treatment effects in randomized clinical trials. Arch. Gen. Psychiatry; 2002; 59, pp. 877-883. [DOI: https://dx.doi.org/10.1001/archpsyc.59.10.877]
11. Seguin-Fowler, R.A.; Strogatz, D.; Graham, M.L.; Eldridge, G.D.; Marshall, G.A.; Folta, S.C.; Pullyblank, K.; Nelson, M.E.; Paul, L. The Strong Hearts, Healthy Communities Program 2.0: An RCT examining effects on Simple 7. Am. J. Prev. Med.; 2020; 59, pp. 32-40. [DOI: https://dx.doi.org/10.1016/j.amepre.2020.01.027] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32389532]
12. Seguin-Fowler, R.A.; Eldridge, G.D.; Rethorst, C.D.; Graham, M.L.; Demment, M.; Strogatz, D.; Folta, S.C.; Maddock, J.E.; Nelson, M.E.; Ha, S. Improvements and maintenance of clinical and functional measures among rural women: Strong Hearts, Healthy Communities-2.0 cluster randomized trial. Circ. Cardiovasc. Qual. Outcomes; 2022; 15, e009333. [DOI: https://dx.doi.org/10.1161/CIRCOUTCOMES.122.009333] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36378768]
13. Maddock, J.E.; Demment, M.; Graham, M.; Folta, S.; Strogatz, D.; Nelson, M.; Ha, S.Y.; Eldridge, G.D.; Seguin-Fowler, R.A. Changes in physical activity outcomes in the Strong Hearts, Healthy Communities (SHHC-2.0) community-based randomized trial. Int. J. Behav. Nutr. Phys. Act.; 2022; 19, 159. [DOI: https://dx.doi.org/10.1186/s12966-022-01401-1]
14. Uribe, A.L.M.; Demment, M.; Graham, M.L.; Szeszulski, J.; Rethorst, C.D.; Githinji, P.; Nelson, M.E.; Strogatz, D.; Folta, S.C.; Bailey, R.L. et al. Improvements in dietary intake, behaviors, and psychosocial measures in a community-randomized cardiovascular disease risk reduction intervention: Strong Hearts, Healthy Communities 2.0. Am. J. Clin. Nutr.; 2023; 118, pp. 1055-1066. [DOI: https://dx.doi.org/10.1016/j.ajcnut.2023.09.003]
15. Ware, J.E., Jr.; Sherbourne, C.D. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med. Care; 1992; 30, pp. 473-483. [DOI: https://dx.doi.org/10.1097/00005650-199206000-00002]
16. Spitzer, R.L.; Kroenke, K.; Williams, J.B.W.; Lowe, B. A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch. Intern. Med.; 2006; 166, pp. 1092-1097. [DOI: https://dx.doi.org/10.1001/archinte.166.10.1092]
17. Smith, B.W.; Dalen, J.; Wiggins, K.; Tooley, E.; Christopher, P.; Bernard, J. The Brief Resilience Scale: Assessing the ability to bounce back. Int. J. Behav. Med.; 2008; 15, pp. 194-200. [DOI: https://dx.doi.org/10.1080/10705500802222972]
18. Cohen, S.; Kamarck, T.; Mermelstein, R. A global measure of perceived stress. J. Health Soc. Behav.; 1983; 24, pp. 385-396. [DOI: https://dx.doi.org/10.2307/2136404]
19. Spitzer, R.L.; Kroenke, K.; Williams, J.B. Validation and utility of a self-report version of PRIME-MD: The PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA; 1999; 282, pp. 1737-1744. [DOI: https://dx.doi.org/10.1001/jama.282.18.1737]
20. Murphy, R. On the use of one-sided statistical tests in biomedical research. Clin. Exp. Pharmacol. Physiol.; 2018; 45, pp. 109-114. [DOI: https://dx.doi.org/10.1111/1440-1681.12754]
21. Luppino, F.S.; de Wit, L.M.; Bouvy, P.F.; Stijnen, T.; Cuijpers, P.; Penninx, B.W.; Zitman, F.G. Overweight, obesity, and depression: A systematic review and meta-analysis of longitudinal studies. Arch. Gen. Psychiatry; 2010; 67, pp. 220-229. [DOI: https://dx.doi.org/10.1001/archgenpsychiatry.2010.2] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20194822]
22. Legenbauer, T.; De Zwaan, M.; Benecke, A.; Mühlhans, B.; Petrak, F.; Herpertz, S. Depression and anxiety: Their predictive function for weight loss in obese individuals. Obes. Facts; 2009; 2, pp. 227-234. [DOI: https://dx.doi.org/10.1159/000226278] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20054228]
23. Carson, T.L.; Jackson, B.E.; Nolan, T.S.; Williams, A.; Baskin, M.L. Lower depression scores associated with greater weight loss among rural black women in a behavioral weight loss program. Transl. Behav. Med.; 2017; 7, pp. 320-329. [DOI: https://dx.doi.org/10.1007/s13142-016-0452-2]
24. Aggarwal, B.; Liao, M.; Allegrante, J.P.; Mosca, L. Low social support level is associated with non-adherence to diet at 1 year in the Family Intervention Trial for Heart Health (FIT Heart). J. Nutr. Educ. Behav.; 2010; 42, pp. 380-388. [DOI: https://dx.doi.org/10.1016/j.jneb.2009.08.006]
25. Wang, J.B.; Pierce, J.P.; Ayala, G.X.; Cadmus-Bertram, L.A.; Flatt, S.W.; Madanat, H.; Newman, V.A.; Nichols, J.F.; Natarajan, L. Baseline depressive symptoms, completion of study assessments, and behavior change in a long-term dietary intervention among breast cancer survivors. Ann. Behav. Med.; 2015; 49, pp. 819-827. [DOI: https://dx.doi.org/10.1007/s12160-015-9716-1]
26. Susin, N.; de Melo Boff, R.; Ludwig, M.W.B.; Feoli, A.M.P.; da Silva, A.G.; Macagnan, F.E.; da Silva Oliveira, M. Predictors of adherence in a prevention program for patients with metabolic syndrome. J. Health Psychol.; 2016; 21, pp. 2156-2167. [DOI: https://dx.doi.org/10.1177/1359105315572451]
27. Teixeira, P.J.; Going, S.B.; Houtkooper, L.B.; Cussler, E.C.; Metcalfe, L.L.; Blew, R.M.; Sardinha, L.B.; Lohman, T.G. Pretreatment predictors of attrition and successful weight management in women. Int. J. Obes.; 2004; 28, pp. 1124-1133. [DOI: https://dx.doi.org/10.1038/sj.ijo.0802727]
28. Ahnis, A.; Riedl, A.; Figura, A.; Steinhagen-Thiessen, E.; Liebl, M.E.; Klapp, B.F. Psychological and sociodemographic predictors of premature discontinuation of a 1-year multimodal outpatient weight-reduction program: An attrition analysis. Patient Prefer. Adherence; 2012; 6, pp. 165-177. [DOI: https://dx.doi.org/10.2147/PPA.S28022]
29. Patsalos, O.; Keeler, J.; Schmidt, U.; Penninx, B.W.; Young, A.H.; Himmerich, H. Diet, obesity, and depression: A systematic review. J. Pers. Med.; 2021; 11, 176. [DOI: https://dx.doi.org/10.3390/jpm11030176]
30. Van Dammen, L.; Wekker, V.; De Rooij, S.; Groen, H.; Hoek, A.; Roseboom, T. A systematic review and meta-analysis of lifestyle interventions in women of reproductive age with overweight or obesity: The effects on symptoms of depression and anxiety. Obes. Rev.; 2018; 19, pp. 1679-1687. [DOI: https://dx.doi.org/10.1111/obr.12752]
31. Alyahya, R.A.; Alnujaidi, M.A. Prevalence and outcomes of depression after bariatric surgery: A systematic review and meta-analysis. Cureus; 2022; 14, e25651. [DOI: https://dx.doi.org/10.7759/cureus.25651] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35784972]
32. Firth, J.; Marx, W.; Dash, S.; Carney, R.; Teasdale, S.B.; Solmi, M.; Stubbs, B.; Schuch, F.B.; Carvalho, A.F.; Jacka, F. et al. The effects of dietary improvement on symptoms of depression and anxiety: A meta-analysis of randomized controlled trials. Psychosom. Med.; 2019; 81, pp. 265-280. [DOI: https://dx.doi.org/10.1097/PSY.0000000000000673] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30720698]
33. Heissel, A.; Heinen, D.; Brokmeier, L.L.; Skarabis, N.; Kangas, M.; Vancampfort, D.; Stubbs, B.; Firth, J.; Ward, P.B.; Rosenbaum, S. et al. Exercise as medicine for depressive symptoms? A systematic review and meta-analysis with meta-regression. Br. J. Sports Med.; 2023; 57, pp. 1049-1057. [DOI: https://dx.doi.org/10.1136/bjsports-2022-106282]
34. Birch, J.M.; Jones, R.A.; Mueller, J.; McDonald, M.D.; Richards, R.; Kelly, M.P.; Griffin, S.J.; Ahern, A.L. A systematic review of inequalities in the uptake of, adherence to, and effectiveness of behavioral weight management interventions in adults. Obes. Rev.; 2022; 23, e13438. [DOI: https://dx.doi.org/10.1111/obr.13438]
35. Seguin, R.A.; Eldridge, G.; Graham, M.L.; Folta, S.C.; Nelson, M.E.; Strogatz, D. Strong Hearts, Healthy Communities: A rural community-based cardiovascular disease prevention program. BMC Public Health; 2016; 16, 86. [DOI: https://dx.doi.org/10.1186/s12889-016-2751-4]
36. Sriram, U.; Sandreuter, K.; Graham, M.; Folta, S.; Pullyblank, K.; Paul, L.; Seguin, R. Process evaluation of Strong Hearts, Healthy Communities: A rural community-based cardiovascular disease prevention program. J. Nutr. Educ. Behav.; 2019; 51, pp. 138-149. [DOI: https://dx.doi.org/10.1016/j.jneb.2018.10.011]
37. Szeszulski, J.; Rolke, L.J.; Ayine, P.; Bailey, R.; Demment, M.; Eldridge, G.D.; Folta, S.C.; Graham, M.L.; MacMillan Uribe, A.L.; McNeely, A. et al. Process evaluation findings from Strong Hearts, Healthy Communities 2.0: A cardiovascular disease prevention intervention for rural women. Int. J. Behav. Nutr. Phys. Act.; 2024; 21, 122. [DOI: https://dx.doi.org/10.1186/s12966-024-01670-y]
38. Wang, H.; Kenkel, D.; Graham, M.L.; Paul, L.C.; Folta, S.C.; Nelson, M.E.; Strogatz, D.; Seguin, R.A. Cost-effectiveness of a community-based cardiovascular disease prevention intervention in medically underserved rural areas. BMC Health Serv. Res.; 2019; 19, 315. [DOI: https://dx.doi.org/10.1186/s12913-019-4117-y]
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Demment, Margaret M 1 ; Ha, Seungyeon 2
; Folta, Sara C 3
; Graham, Meredith L 1
; Eldridge, Galen D 1 ; Seguin-Fowler, Rebecca A 4
1 Institute for Advancing Health Through Agriculture, Texas A&M University, Dallas, TX 75252, USA;
2 Statistical Consultation Center, Texas A&M University, College Station, TX 77843, USA;
3 Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA 02111, USA;
4 Institute for Advancing Health Through Agriculture, Texas A&M University, College Station, TX 77840, USA