Abstract
Aims:
We assessed costs and cost-effectiveness of implementing Fit Body and Soul (FBAS), a churchbased 18-session lifestyle education intervention for African Americans.
Methods:
We calculated incremental cost-effectiveness ratios (ICER) using data from a cluster randomized controlled trial comparing FBAS with health education (HE) among 604 overweight participants in 20 churches. The ICER was the adjusted difference in costs to deliver FBAS versus HE over the difference in weight change at one-year follow-up. Costs included those incurred for participant identification and program implementation. We fitted linear mixed-effects regression models, accounting for clustering of participants within churches and for age, sex, and educational attainment. We repeated these analyses for secondary outcomes (waist circumference (cm), physical activity (MET), glucose, blood pressure, and quality of life).
Results:
Per-person intervention cost of FBAS was $50.39 more than HE ($442.22 vs. $391.83 perperson), and adjusted differences in weight change (1.9 kg [95% CI 1.0, 2.8]) and waist circumference (2.4 cm [95% CI, 1.3 to 3.4]) were both significant. FBAS did not result in statistically significant differences in METs, glucose, blood pressures, or quality of life. We estimated that compared to HE, FBAS costs an additional $26.52 per kg weight lost and $21.00 per cm reduction in waist circumference.
Conclusions:
For a modest increase in cost, FBAS led to greater weight and waist reductions among African Americans in a church setting.
Keywords: diabetes prevention, obesity, lifestyle intervention, cost, African American, church
1. INTRODUCTION
Over 30 million adults in the United States (US) are living with diabetes, a debilitating and costly condition that disproportionately affects African Americans [1]. Diabetes is associated with a multitude of complications, including cardiovascular disease and chronic kidney disease, and cost an estimated $327 billion in direct medical expenditures and productivity loss in 2017 [1–3]. Randomized controlled trials such as the US Diabetes Prevention Program (DPP) trial and translation studies have shown that lifestyle interventions that promote weight loss through diet and physical activity can reduce type 2 diabetes mellitus (henceforth referred to as diabetes) incidence, improve cardiometabolic health indicators such as blood pressure, and improve health-related quality of life in high-risk adults [4–10]. Several recent efforts to reach high risk populations, such as African Americans, have focused on delivering lifestyle interventions through faith-based institutions.
Despite their efficacy, lifestyle interventions can be resource intensive and costly, which pose a major challenge to the large-scale implementation and economic sustainability of such programs [7]. To date, no studies have estimated the value that could be gained from investing in faith-based lifestyle interventions targeting African Americans. In a systematic review and meta-analysis of 28 US-based translation studies of DPP-adapted lifestyle interventions in real-world settings, four evaluated lifestyle interventions implemented in AA churches, but these studies were single-group intervention or nonrandomized controlled studies [7]. The meta-analysis, as a whole, concluded that programs can lower implementation costs and still achieve clinically significant weight loss, the single most important factor in reducing diabetes incidence [7]. However, there are few translation studies that have evaluated the cost-effectiveness of lifestyle interventions [7, 11]. There is an even greater paucity of research on the cost-effectiveness of lifestyle interventions targeting African Americans, a high-risk group.
To address this gap in the evidence and inform decision-making for diabetes prevention among African Americans, this study assessed the costs and cost-effectiveness of implementing Fit Body and Soul [FBAS], a faith-based lifestyle intervention adapted from the Group Lifestyle Balance program, which was based on the DPP. In a cluster randomized controlled trial, African Americans participating in FBAS achieved significantly greater weight loss compared to a Health Education (HE) control group [12]. We build on these prior findings to examine the potential costs and value associated with investing in this effective intervention.
2. SUBJECTS, MATERIALS AND METHODS
2.1. Design
Full study methodology, including detailed inclusion criteria and descriptions of the curricula for both interventions, and intervention effectiveness are described elsewhere [12, 13]. Briefly, the Fit Body and Soul (FBAS) study was a cluster-randomized, controlled, single-blinded, faith-based lifestyle intervention trial for diabetes prevention conducted in African American churches from October 2009 to March 2012 (ClinicalTrials.gov Identifier NCT01730196). Twenty churches were randomized to receive FBAS (intervention), or a HE program (control). FBAS was a faith-based adaptation of the DPP Group Lifestyle Balance intended to promote weight loss and improve diet and physical activity. The HE program sessions covered health topics selected from the Centers for Disease Control and Prevention Guide to Community Prevention Services and represents a standard (non-faith based) lifestyle intervention program.
Overweight, non-diabetic adult participants in both groups (n=604) attended 12-weekly group core sessions, followed by six-monthly group maintenance sessions. All sessions lasted one hour and were taught by trained Church Health Advisors who were health care professionals and members of the respective churches’ health ministry.
The main outcome of FBAS was weight change at three months and 12 months after baseline. FBAS participants lost more weight compared with those in the HE control group at three months (2.6 vs 0.4 kg, p<0.0001) and 12 months (2.0 vs 0.1 kg, p<0.0001) and were more likely than the HE control group to achieve a 7% weight loss at both time points (13% vs. 3% at three months, p<0.001; 19% vs. 8% at 12 months, p<0.001) [12].
2.2. Cost Data
To obtain cost data, we accessed FBAS study records. From these records, we extracted all costs associated with implementation of FBAS and HE. Costs included participant identification costs (blood test screening at baseline) and program implementation costs, including payment for Church Health Advisors at three time points, materials, equipment, non-monetary incentives to participants (e.g., gift cards, pedometers, t-shirts). All costs, which were originally recorded in 2012 US dollars (USD), were adjusted for inflation in a compound fashion using the CPI Inflation Calculator from the Labor of Bureau Statistics to reflect the costs of the interventions in 2017 USD [14].
2.3. Health data
Our primary outcome was the difference in weight reduction in kilograms (kg) between FBAS and HE groups. We examined secondary outcomes such as: physical activity (MET), fasting plasma glucose (mg/dl), waist circumference (in centimeters (cm)), systolic and diastolic blood pressure, and health-related quality of life (assessed using Short Form 12 (SF-12v®)). Measurements were obtained at baseline, three months, and 12 months post-intervention. A detailed description of these health measurements and data collection procedures has been published elsewhere [13].
2.4. Analysis
Our analysis was conducted adhering to the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement guidelines and is from a healthcare provider perspective [15]. We added all implementation costs of FBAS to determine the total cost of FBAS. The total cost was then divided by the number of participants in the FBAS group to produce the per-participant cost of delivering FBAS. We used the same calculation to determine the per-participant cost of HE. To show the cost of replicating similar faith-based lifestyle interventions in ‘real-world’ settings, we computed costs after excluding all research expenses. Research expenses included research and support staff salaries, travel-related expenses, and costs of equipment for management and analysis of trial.
To evaluate the cost-effectiveness of FBAS compared to HE, we estimated the incremental cost effectiveness ratio (ICERs). An ICER compares the differential costs and outcomes between the two groups using the formula: (cost per participant FBAS – cost per participant HE) / (mean change in outcome FBAS – mean change in outcome HE) [16]. The mean change in the primary outcome (weight) and all six secondary outcomes (waist circumference, fasting plasma glucose, physical activity, systolic blood pressure, diastolic blood pressure, and quality of life) was measured at 12 months. Since all implementation costs and measured health effects occurred within the 12-month period of the study, we did not apply a discount rate in our analyses.
We estimated between-group mean differences in weight at 12 months using an intention-totreat analysis. We fitted a linear mixed-effects regression model with weight over time as the outcome. We included indicators for time of assessment (baseline, three months, and 12 months), intervention (FBAS versus HE), and time x intervention interaction as fixed effects and participants as random effects. We also adjusted for age, sex, and educational attainment. Because participants were nested within churches, we also accounted for clustering of participants within churches in the model and included church as a random effect [17]. We tested the statistical significance of the difference in distributions of weight with a likelihood ratio test (null hypothesis Ho = β = 0, with the coefficient β as the intervention by 12-month assessment interaction). If the interaction was found to be statistically significant (p < .05), we used the least squares means (LS-means) of fixed effects to estimate the between-group mean differences in mean weight change from baseline to three months and from baseline to 12 months. These modeling steps were repeated for all six secondary outcomes. All available data from the 604 randomized participants (FBAS, n=317; HE, n=287) were included in the analysis of weight, waist circumference, fasting plasma glucose, physical activity, systolic blood pressure, and diastolic blood pressure. Seven participants were missing data on quality of life at all assessment points, and thus, were excluded from that analysis. All analyses assumed that data were missing at random. All statistical analyses were conducted using SAS/STAT 14.1 (SAS Institute, Inc. Cary, NC). Finally, for outcomes that were statistically significantly different between the intervention and control groups (i.e., models in which the intervention x time interaction coefficient had a p < .05), we calculated an ICER.
To estimate how variations in the cost of delivery and impacts of FBAS might affect our findings, we performed one-way sensitivity analyses. We allowed the cost of Church Health Advisors in FBAS to vary ± 10%, ± 20%, and ± 30%. We also examined scenarios where the adjusted difference in weight change between the FBAS and the HE control group and the adjusted difference in waist circumference was higher or lower by ± 10%, ± 20%, and ± 30%. Finally, we calculated an ICER that compared the differential costs and outcomes between FBAS and no intervention.
3. RESULTS
At baseline, demographic participant characteristics between the FBAS and HE control groups were similar. FBAS and HE control group participants’ mean age was 46.6 and 46.4 years, respectively. Among the FBAS and HE control groups, 84.2% and 82.6% were female, and 48.9% and 53.0% completed college or more. The groups were also similar in terms of the following health indicators: mean weight (FBAS, 98.4 kg; HE, 99 kg); body mass index (FBAS, 35.8 kg/m2; HE, 35.6 kg/m2), waist circumference (FBAS, 107.8 cm; HE, 106.7 cm), systolic blood pressure (FBAS, 130.4; HE, 130.6), diastolic blood pressure (FBAS, 82.9; HE, 82.2), glycated hemoglobin (FBAS, 5.8; HE, 5.8), fasting plasma glucose (FBAS, 90.1 mg/dl; HE, 89.9 mg/dl), total physical activity (FBAS, 1736.9 MET; HE, 1998.2 MET), SF-12 physical component summary (FBAS, 48.7; HE, 49.1), and SF-12 mental component summary (FBAS, 52.4; HE, 51.8).
The cost of delivering FBAS was $442.22 per person, $50.39 more than the cost of delivery of HE ($391.83 per person). Materials accounted for the largest component of the cost difference, with the materials for the FBAS group costing $36.78 more than the materials for the HE group per participant. The remainder of the cost difference was due to payments to Church Health Advisors for leading training sessions, which were provided at three time points during implementation. FBAS Church Health Advisors were paid a total of $129.24 per participant compared to $115.63 per participant for the HE Church Health Advisors, resulting in per participant payments for the FBAS Church Health Advisors that were $13.61 more than the per participant payments for HE Church Health Advisors. Participant incentives were the same for both groups at $166.50 per participant (Table 1). The total cost of delivering FBAS (for 317 participants) and HE (for 287 participants) was approximately $139,500 and $113,500, respectively (Appendix Table 1).
Table 1:
Total Implementation Cost Comparison of the FBAS and HE Control Groups per Participant
| Fit Body and Soul | Health Education | Difference | |
|---|---|---|---|
| Part a. Participant Identification | |||
| Screenings costs* | $91.41 | $91.41 | $0.00 |
| Part b. Program Implementation | |||
| Church Health Advisors | |||
| Training Payment | $27.21 | $13.60 | $13.61 |
| Payment at session 12 | $68.02 | $68.02 | $0.00 |
| Payment at booster 6 | $34.01 | $34.01 | $0.00 |
| Materials and Equipment | |||
| Binder and handouts | $40.92 | $9.58 | $31.34 |
| Resistance bands | $5.44 | $0.00 | $5.44 |
| Pedometer | $8.71 | $8.71 | $0.00 |
| Participant Incentives | $166.50 | $166.50 | $0.00 |
| TOTAL | $442.22 | $391.83 | $50.39 |
Screening cost included blood work at baseline
All costs are adjusted for inflation and are presented in USD 2017
Participants in both arms lost weight over time (p < .0001 for time), but this weight loss was greater in the FBAS group (p < .0001 for intervention x time interaction) (Table 2). The FBAS group had an estimated mean weight change of −2.6 kg (95% CI: −4.0 to −1.1) at three months and −2.0 kg (95% CI: 3.5 to −0.6) at 12 months. In comparison, the estimated mean weight change for the HE control group was −0.4 kg (95% CI: −1.9 to +1.1) and −0.1 kg (95% CI: −1.6 to +1.4) at three months and 12 months, respectively. Thus, weight loss was 2.1 kg (95% CI: 1.3 to 3.0) and 1.9 kg (95% CI: 1.0 to 2.8) greater in the FBAS group compared with the HE control group at three months and 12 months, respectively. The mean weight at 12 months was 95.6 ± 20.5 kg and 98.5 ± 22.6 kg for the FBAS and HE groups, respectively (Appendix Table 2).
Table 2:
Change in Weight, Fasting Plasma Glucose, Physical Activity, Waist Circumference, Blood Pressure, and Quality of Life by Treatment Group
| Adjusted Least-Squares Mean (95% CI) * | ICER ($) at 12 months | ||||||
|---|---|---|---|---|---|---|---|
| Change from baseline | P value | ||||||
| Baseline | 3 months | 12 months | Group | Time | Group × Time | ||
| PRIMARY OUTCOME | |||||||
| Weight, kg | 0.35 | <.0001 | <.0001 | 26.52 | |||
| Fit Body and Soul | 101.2 (94.8, 107.6) | −2.6 (−4.0, −1.1) | −2.0 (−3.5, −0.6) | ||||
| Health Education | 101.4 (94.9, 108.0) | −0.4 (−1.9, 1.1) | −0.1 (−1.6, 1.4) | ||||
| Difference | −2.1 (−3.0, −1.3) | −1.9 (−2.8, −1.0) | |||||
| P value | <.0001 | <.0001 | |||||
| SECONDARY OUTCOMES | |||||||
| Fasting plasma glucose, mg/dl | 0.6 | <.0001 | 0.4 | N/A** | |||
| Fit Body and Soul | 90.3 (87.3, 93.2) | −0.02 (−2.9, 2.8) | −1.0 (−3.9, 1.9) | ||||
| Health Education | 90.0 (87.0, 92.9) | 0.7 (−2.2, 3.6) | −1.5 (−4.5, 1.4) | ||||
| Difference | −0.7 (−2.4, 1.0) | 0.5 (−1.2, 2.3) | |||||
| P value | 0.4 | 0.6 | |||||
| Total physical activity, MET | 0.5 | <.0001 | 0.2 | N/A** | |||
| Fit Body and Soul | 1714.1 (824.7, 2603.6) | 566.4 (−674.1, 1807.0) | −509.5 (−1750.1, 731.0) | ||||
| Health Education | 1951.3 (1042.4, 2860.2) | 795.9 (−468.9, 2060.7) | −991.4 (−2256.2, 273.4) | ||||
| Difference | −229.4 (−992.1, 533.3) | 481.9 (−280.8, 1244.5) | |||||
| P value | 0.6 | 0.2 | |||||
| Waist circumference, cm | 0.73 | <.0001 | <.0001 | 21.00 | |||
| Fit Body and Soul | 108.0 (103.3, 112.6) | −2.0 (−3.8, −0.1) | −1.2 (−3.1, 0.6) | ||||
| Health Education | 106.8 (102.0, 111.6) | 0.4(−1.4, 2.3) | 1.1 (−.7, 3.0) | ||||
| Difference | −2.4 (−3.5, −1.4) | −2.4 (−3.4, −1.3) | |||||
| P value | <.0001 | <.0001 | |||||
| Systolic blood pressure | 0.7 | <.0001 | 0.8 | N/A** | |||
| Fit Body and Soul | 132.8 (128.3, 137.3) | −3.5 (−7.5, 0.6) | −2.1 (−6.2, 1.9) | ||||
| Health Education | 132.8 (128.3, 137.4) | −2.9 (−7.0, 1.2) | −2.1 (−6.2, 2.0) | ||||
| Difference | −0.5 (−3.0, 1.9) | −0.1 (−2.5, 2.4) | |||||
| P value | 0.7 | 0.96 | |||||
| Diastolic blood pressure | 0.3 | <.0001 | 0.5 | N/A** | |||
| Fit Body and Soul | 85.3 (82.7, 88.0) | −1.8 (−4.2, 0.6) | −0.9 (−3.3, 1.4) | ||||
| Health Education | 84.6 (81.9, 87.3) | −1.1 (−3.5, 1.3) | −1.1 (−3.5, 1.4) | ||||
| Difference | −0.7 (−2.1, 0.7) | 0.1 (−1.3, 1.5) | |||||
| P value | 0.3 | 0.9 | |||||
| SF-12 physical component summary | 0.7 | 0.0 | 0.5 | N/A** | |||
| Fit Body and Soul | 47.4 (45.3, 49.6) | 0.4 (−1.9, 2.7) | 0.1 (−2.2, 2.3) | ||||
| Health Education | 47.7 (45.5, 49.9) | −0.2 (−2.5, 2.1) | −0.6 (−2.9, 1.8) | ||||
| Difference | 0.6 (−0.7, 2.0) | 0.6 (−0.8, 2.0) | |||||
| P value | 0.4 | 0.4 | |||||
| SF-12 mental component summary | 0.3 | 0.0 | 0.98 | N/A** | |||
| Fit Body and Soul | 51.7 (49.6, 53.8) | 2.0 (−0.6, 4.7) | 1.2 (−1.5, 3.8) | ||||
| Health Education | 51.1 (49.0, 53.3) | 1.7 (−1.0, 4.4) | 0.5 (−2.2, 3.2) | ||||
| Difference | 0.3 (−1.3, 1.9) | 0.7 (−1.0, 2.3) | |||||
| P value | 0.7 | 0.4 | |||||
All models were adjusted for sex, age, and educational attainment.
N/A, not applicable. There was not a statistically significant difference between the FBAS and HE control groups, and thus, an ICER was not calculated.
We also found a statistically significant difference in waist circumference change between groups over time (p < .0001 for time and p < .0001 for intervention x time interaction). The reduction in waist circumference was 2.4 cm (95% CI: 1.3 to 3.4) greater in the FBAS group compared to the HE control group at 12 months. The mean waist circumference was 105.4 ± 15.0 cm and 107.1 ± 16.6 cm for the FBAS and HE groups, respectively (Appendix Table 2). There were no significant differences between the FBAS and HE control groups for fasting plasma glucose, physical activity, systolic blood pressure, diastolic blood pressure, and quality of life.
Evaluating the cost-effectiveness at 12 months, it would cost an additional $26.52 to achieve each one kg of weight lost. For waist circumference, over 12 months, it would cost $21.00 per one cm reduction. ICERs for fasting plasma glucose, physical activity, systolic blood pressure, diastolic blood pressure, and quality of life were not calculated, since we did not find significant differences between the FBAS and HE control groups in these outcomes (Table 2).
The results of one-way sensitivity analyses are presented in Table 3. The ICERs were highly sensitive to the cost of Church Health Advisors. For example, a, 30% cost increase in the cost of Church Health Advisors nearly doubled the cost of an additional kg weight lost and cm reduction in waist circumference in the FBAS group compared to the HE control group. The ICERs were less sensitive to variations in outcomes. Varying the adjusted difference in weight change between the groups resulted in ICERs similar to the base case. Similarly, variations in the adjusted difference in waist circumference reduction between the groups produced ICERs close to the base-case ICERs. When we compared FBAS to no intervention, we found that it would cost an additional $221.11 to achieve one kg of weight lost and $368.52 per one cm reduction.
Table 3:
One-way sensitivity analyses
| Parameter | Base-case value | Basis for alternative | Alternative value | ICER ($) | |
|---|---|---|---|---|---|
| Weight* | Waist circumference** | ||||
| Church Health Advisors | $129.24 | 30% less | $90.47 | 6.11 | 4.84 |
| 20% less | $103.39 | 12.92 | 10.23 | ||
| 10% less | $116.32 | 19.72 | 15.61 | ||
| 10% more | $142.16 | 33.32 | 26.38 | ||
| 20% more | $155.09 | 40.13 | 31.77 | ||
| 30% more | $168.01 | 46.93 | 37.15 | ||
| Weight (kg) | 1.9 | 30% less | 1.33 | 37.89 | − |
| 20% less | 1.52 | 33.15 | − | ||
| 10% less | 1.71 | 29.47 | − | ||
| 10% more | 2.09 | 24.11 | − | ||
| 20% more | 2.28 | 22.10 | − | ||
| 30% more | 2.47 | 20.40 | − | ||
| Waist circumference (cm) | 2.4 | 30% less | 1.68 | − | 29.99 |
| 20% less | 1.92 | − | 26.24 | ||
| 10% less | 2.16 | − | 23.33 | ||
| 10% more | 2.64 | − | 19.09 | ||
| 20% more | 2.88 | − | 17.50 | ||
| 30% more | 3.12 | − | 16.15 | ||
Cost per kg decrease in weight.
Cost per cm decrease in waist circumference.
4. DISCUSSION
Although evidence suggests that lifestyle interventions delivered at faith-based institutions have indeed been successful [12, 18–20], data are scarce on the costs of delivery and cost-effectiveness of a faith-based lifestyle intervention for diabetes prevention among African Americans. The incremental implementation cost of FBAS over HE, $50.39 per participant, was minimal. As such, the intervention provides good value, wherein each additional kg reduction in weight and each cm reduction in waist circumference cost $26.52 and $21, relative to the HE intervention, respectively.
These study findings can be useful to payers who invest in the health of communities. Notably, there has been increased interest in wellness and prevention programs among payers in the past decade. Most recently, the Centers for Medicare and Medicaid Services (CMS) expanded the Medicare Diabetes Prevention Program (MDPP) model to increase Medicare beneficiaries’ access to DPP, with the aim of reducing the rate of progression to diabetes and lowering spending [21]. Beginning in 2018, CDCrecognized organizations delivering DPP can enroll in Medicare as MDPP suppliers [21]. Furthermore, evidence of the importance of investing in lifestyle interventions as a strategy for addressing diabetes continues to accumulate. For example, a recent meta-analysis showed that lifestyle interventions have sustained benefits, whereas glucose lowering medications only suppress glucose while they are in use [22].
Our findings regarding the cost-effectiveness of implementing faith-based lifestyle interventions are helpful towards evidence-informed decisions about investing in and scaling such programs. Moreover, our findings add to a growing body of literature suggesting that diabetes prevention programs culturally tailored to high-risk groups have improved efficacy with similar cost to a conventional lifestyle program [23]. Although there is currently no widely accepted cost effectiveness threshold for reduction in weight or waist circumference, the results of this analysis are comparable to ICERs calculated for the same health effects in other studies. In a recent systematic review, Sun et al. reported that the lowest average cost per kg weight loss was $34.06 over 6 months, while the highest was $1,005.36 over 12 months [11]. As expected, because of the high cost of HE and resulting small cost difference between FBAS and HE, the ICERs reported in this study were on the lower end of the range of ICERs from similar studies. The high cost of HE is particularly worth noting given that the HE group did not achieve health benefits. Results from our sensitivity analyses showed that compared to no intervention, FBAS would cost an additional $221.11 to achieve one kg of weight lost and $368.52 per one cm reduction in waist circumference. Overall, our results are consistent with previous studies showing that lifestyle interventions for adults at increased diabetes risk can be delivered in community settings at lower costs than those implemented in clinical trials and are cost-effective [24–26].
Many cost-effectiveness analyses of lifestyle interventions for diabetes prevention do not include outcomes beyond weight. A strength of our study is that it had an expanded scope by including additional outcomes that are critically important for cardiometabolic health and risk for diabetes, such as waist circumference, fasting plasma glucose, physical activity level, and blood pressure. Randomized controlled trials have demonstrated that intensive lifestyle modification for adults with impaired glucose tolerance is associated with many beneficial effects (slow progression to diabetes; regression to normal glucose regulation; reduced need for medications that lower blood pressure and lipid levels; reduced disability, retinopathy, and urinary incontinence; increased quality of life) [4, 8–10, 27–30]. Although we calculated ICERs for weight and waist circumference only, most modeled cost-effectiveness analyses of lifestyle interventions would incorporate other health benefits and the estimated value of such interventions would be even greater.
A notable weakness is that we were unable to assess the long-term cost-effectiveness of FBAS. Since our study duration was one year, we captured the high upfront expenditure of delivering FBAS, but were only able to partially capture the benefits of the intervention, which accrue over years. Evidence from the DPP clinical trial and other similar trials demonstrate that lifestyle interventions provide lasting health benefits, with effects persisting up to 15 years after the intervention [4–6, 27, 28]. Thus, the incremental health effects of the FBAS compared to HE may be underestimated in a 12-month trial, leading to high ICERs in the short-term. The highly sensitive nature of cost effectiveness analyses to the time horizon of the trial is a well-known challenge and should be taken into consideration in the interpretation of the findings [31]. Moreover, it is important that payers recognize that they may not receive high returns on investment early, but long-term studies of lifestyle interventions provide robust evidence that these benefits will accrue over time [4–6, 27, 28]. Also, while we adhered to the recommendations from the Second Panel on Cost-effectiveness in Health and Medicine where possible, it was not possible to conduct analyses from the health care sector perspective or societal perspective due to data limitations, such as limited health care utilization data [32]. However, our findings provide an important, internally valid perspective of a within-trial analysis.
In conclusion, our study helps to address the dearth of literature on cost-effectiveness of faithbased lifestyle interventions targeting African Americans, a high-risk minority population. We found that a faith-based lifestyle intervention is effective in modestly reducing weight and waist circumference among African Americans for a relatively small cost compared to a health education intervention. Since this is just one modality of delivering lifestyle modification programs, future studies should assess the long-term sustainability of such programs and the cost-effectiveness of varying approaches to delivering these interventions.
ACKNOWLEDGEMENTS
This work was supported by the National Institutes of Health (NIH) National Institute of Diabetes and Digestive and Kidney Diseases [grant number R18DK082401], National Heart, Lung, And Blood Institute [grant number R01HL125442], and Fogarty International Center [grant number R25Tw09337]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. MKA and KMVN are partially supported by the Georgia Center for Diabetes Translation Research funded by the National Institute of Diabetes and Digestive and Kidney Diseases [grant number P30DK111024].
APPENDIX
Table 1:
Total Implementation Cost Comparison of the FBAS and HE Control Groups (n=604)
| Fit Body and Soul | Health Education | Difference | |
|---|---|---|---|
| n=317 | n=287 | ||
| Part a. Participant Identification | |||
| Screenings costs* | $28,978.02 | $26,235.62 | $2,742.40 |
| Part b. Program Implementation | |||
| Church Health Advisors | |||
| Training Payment | $8,488.38 | $4,026.54 | $4,461.84 |
| Payment at session 12 | $21,220.95 | $20,132.69 | $1,088.26 |
| Payment at Booster 6 | $10,610.47 | $10,066.35 | $544.12 |
| Materials and Equipment | |||
| Binder and handouts | $12,971.11 | $2,748.49 | $10,222.62 |
| Resistance bands | $1,724.88 | $0.00 | $1,724.88 |
| Pedometer | $2,759.81 | $2,498.63 | $261.18 |
| Participant Incentives | $52,781.39 | $47,786.30 | $4,995.09 |
| TOTAL | $139,535.01 | $113,494.62 | $26,040.39 |
Screening cost included blood work at baseline
All costs are adjusted for inflation and are presented in USD 2017.
Table 2:
Outcome Parameters at Baseline, 3 months, and 12 months by Treatment Group
| Fit Body and Soul (n= 317) | Health Education (n=287) | |||||
|---|---|---|---|---|---|---|
| Baseline | 3 months | 12 months | Baseline | 3 months | 12 months | |
| Weight, kg | 98.4 ± 21.0 | 95.1 ± 19.9 | 95.6 ± 20.5 | 99.0 ± 22.1 | 98.4 ± 22.4 | 98.5 ± 22.6 |
| Waist circumference, cm | 107.8 ± 15.0 | 104.7 ± 14.3 | 105.4 ± 15.0 | 106.7 ± 15.9 | 106.3 ± 16.0 | 107.1 ± 16.6 |
| Systolic blood pressure** | 130.4 ± 16.8 | 125.6 ± 16.6 | 127.1 ± 17.6 | 130.6 ± 16.3 | 126.5 ± 16.4 | 127.7 ± 16.3 |
| Diastolic blood pressure** | 82.9 ± 9.5 | 80.9 ± 9.5 | 81.9 ± 9.9 | 82.2 ± 9.9 | 80.9 ± 10.0 | 80.8 ± 9.4 |
| Fasting plasma glucose, mg/dl | 90.1 ± 10.0 | 89.7 ± 10.4 | 88.7 ± 12.6 | 89.9 ± 9.4 | 89.5 ± 10.9 | 87.4 ± 12.3 |
| Total physical activity, MET | 1736.9 ± 3334.1 | 2381.3 ± 4034.1 | 1305.3 ± 2660.0 | 1998.2 ± 3976.4 | 2844.5 ± 6295.9 | 1057.2 ± 2207.7 |
| SF-12 physical component summary1 | 48.7 ± 8.9 | 50.1 ± 7.8 | 49.7 ± 8.5 | 49.1 ± 7.7 | 49.8 ± 8.1 | 49.2 ± 9.3 |
| SF-12 mental component summary2 | 52.4 ± 8.4 | 53.5 ± 8.2 | 52.7 ± 8.3 | 51.8 ± 9.2 | 53.0 ± 8.2 | 52.1 ± 9.3 |
Plus-minus values are mean ± SD.
The body-mass index is the weight in kilograms divided by the square of the height in meters
Systolic and diastolic blood pressure were calculated as an average of two measurements (mmHg) taken 1 min apart
Seven participants were missing data at all assessments points, and thus, were excluded from the analysis. n=597
Seven participants were missing data at all assessments points, and thus, were excluded from the analysis. n=597
Footnotes
ClinicalTrials.gov Identifier NCT01730196.
DECLARATIONS OF INTEREST
None
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
REFERENCES
- 1.Federation ID. IDF Diabetes Atlas Eighth Edition 2017 2017. [Google Scholar]
- 2.Huo X, Gao L, Guo L, Xu W, Wang W, Zhi X, et al. Risk of non-fatal cardiovascular diseases in early-onset versus late-onset type 2 diabetes in China: a cross-sectional study. The lancet Diabetes & endocrinology. 2016;4(2):115–24. [DOI] [PubMed] [Google Scholar]
- 3.Economic Costs of Diabetes in the U.S. in 2017. Diabetes Care. 2018. [DOI] [PMC free article] [PubMed]
- 4.Orchard TJ, Temprosa M, Barrett-Connor E, Fowler SE, Goldberg RB, Mather KJ, et al. Long-term effects of the Diabetes Prevention Program interventions on cardiovascular risk factors: a report from the DPP Outcomes Study. Diabet Med. 2013;30(1):46–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lindstrom J, Peltonen M, Eriksson JG, Ilanne-Parikka P, Aunola S, Keinanen-Kiukaanniemi S, et al. Improved lifestyle and decreased diabetes risk over 13 years: long-term follow-up of the randomised Finnish Diabetes Prevention Study (DPS). Diabetologia. 2013;56(2):284–93. [DOI] [PubMed] [Google Scholar]
- 6.Li G, Zhang P, Wang J, Gregg EW, Yang W, Gong Q, et al. The long-term effect of lifestyle interventions to prevent diabetes in the China Da Qing Diabetes Prevention Study: a 20-year follow-up study. Lancet. 2008;371(9626):1783–9. [DOI] [PubMed] [Google Scholar]
- 7.Ali MK, Echouffo-Tcheugui J, Williamson DF. How effective were lifestyle interventions in realworld settings that were modeled on the Diabetes Prevention Program? Health Aff (Millwood). 2012;31(1):67–75. [DOI] [PubMed] [Google Scholar]
- 8.Florez H, Pan Q, Ackermann RT, Marrero DG, Barrett-Connor E, Delahanty L, et al. Impact of lifestyle intervention and metformin on health-related quality of life: the diabetes prevention program randomized trial. J Gen Intern Med. 2012;27(12):1594–601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346(6):393–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mudaliar U, Zabetian A, Goodman M, Echouffo-Tcheugui JB, Albright AL, Gregg EW, et al. Cardiometabolic Risk Factor Changes Observed in Diabetes Prevention Programs in US Settings: A Systematic Review and Meta-analysis. PLoS Med. 2016;13(7):e1002095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sun Y, You W, Almeida F, Estabrooks P, Davy B. The Effectiveness and Cost of Lifestyle Interventions Including Nutrition Education for Diabetes Prevention: A Systematic Review and MetaAnalysis. J Acad Nutr Diet. 2017;117(3):404–21.e36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sattin RW, Williams LB, Dias J, Garvin JT, Marion L, Joshua TV, et al. Community Trial of a FaithBased Lifestyle Intervention to Prevent Diabetes Among African-Americans. J Community Health. 2016;41(1):87–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Williams LB, Sattin RW, Dias J, Garvin JT, Marion L, Joshua T, et al. Design of a clusterrandomized controlled trial of a diabetes prevention program within African-American churches: The Fit Body and Soul study. Contemp Clin Trials. 2013;34(2):336–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.United States Department of Labor Bureau of Labor Statistics. CPI Inflation Calculator 2017. [Available from: https://data.bls.gov/cgibin/cpicalc.pl?cost1=25000&year1=198601&year2=201703.
- 15.Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, et al. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. BMJ. 2013;346:f1049. [DOI] [PubMed] [Google Scholar]
- 16.Owens DK. Interpretation of cost-effectiveness analyses. J Gen Intern Med. 1998;13(10):716–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Gomes M, Ng ES, Grieve R, Nixon R, Carpenter J, Thompson SG. Developing appropriate methods for cost-effectiveness analysis of cluster randomized trials. Med Decis Making. 2012;32(2):35061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Boltri JM, Davis-Smith M, Okosun IS, Seale JP, Foster B. Translation of the National Institutes of Health Diabetes Prevention Program in African American churches. J Natl Med Assoc. 2011;103(3):194202. [DOI] [PubMed] [Google Scholar]
- 19.Boltri JM, Davis-Smith YM, Seale JP, Shellenberger S, Okosun IS, Cornelius ME. Diabetes prevention in a faith-based setting: results of translational research. J Public Health Manag Pract. 2008;14(1):29–32. [DOI] [PubMed] [Google Scholar]
- 20.Davis-Smith YM, Boltri JM, Seale JP, Shellenberger S, Blalock T, Tobin B. Implementing a diabetes prevention program in a rural African-American church. J Natl Med Assoc. 2007;99(4):440–6. [PMC free article] [PubMed] [Google Scholar]
- 21.Services CfMM. Fact Sheet: Final Policies for the Medicare Diabetes Prevention Program Expanded Model in the Calendar Year 2018 Physician Fee Schedule Final Rule.
- 22.Haw JS, Galaviz KI, Straus AN, Kowalski AJ, Magee MJ, Weber MB, et al. Long-term Sustainability of Diabetes Prevention Approaches: A Systematic Review and Meta-analysis of Randomized Clinical Trials. JAMA internal medicine. 2017;177(12):1808–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Patel RM, Misra R, Raj S, Balasubramanyam A. Effectiveness of a Group-Based Culturally Tailored Lifestyle Intervention Program on Changes in Risk Factors for Type 2 Diabetes among Asian Indians in the United States. Journal of diabetes research. 2017;2017:2751980. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Li R, Qu S, Zhang P, Chattopadhyay S, Gregg EW, Albright A, et al. Economic Evaluation of Combined Diet and Physical Activity Promotion Programs to Prevent Type 2 Diabetes Among Persons at Increased Risk: A Systematic Review for the Community Preventive Services Task Force. Ann Intern Med. 2015;163(6):452–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lawlor MS, Blackwell CS, Isom SP, Katula JA, Vitolins MZ, Morgan TM, et al. Cost of a group translation of the Diabetes Prevention Program: Healthy Living Partnerships to Prevent Diabetes. Am J Prev Med. 2013;44(4 Suppl 4):S381–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Roberts S, Barry E, Craig D, Airoldi M, Bevan G, Greenhalgh T. Preventing type 2 diabetes: systematic review of studies of cost-effectiveness of lifestyle programmes and metformin, with and without screening, for pre-diabetes. BMJ open. 2017;7(11):e017184. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Long-term effects of lifestyle intervention or metformin on diabetes development and microvascular complications over 15-year follow-up: the Diabetes Prevention Program Outcomes Study. The lancet Diabetes & endocrinology. 2015;3(11):866–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Li G, Zhang P, Wang J, An Y, Gong Q, Gregg EW, et al. Cardiovascular mortality, all-cause mortality, and diabetes incidence after lifestyle intervention for people with impaired glucose tolerance in the Da Qing Diabetes Prevention Study: a 23-year follow-up study. The lancet Diabetes & endocrinology. 2014;2(6):474–80. [DOI] [PubMed] [Google Scholar]
- 29.Brown JS, Wing R, Barrett-Connor E, Nyberg LM, Kusek JW, Orchard TJ, et al. Lifestyle intervention is associated with lower prevalence of urinary incontinence: the Diabetes Prevention Program. Diabetes Care. 2006;29(2):385–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Gong Q, Gregg EW, Wang J, An Y, Zhang P, Yang W, et al. Long-term effects of a randomised trial of a 6-year lifestyle intervention in impaired glucose tolerance on diabetes-related microvascular complications: the China Da Qing Diabetes Prevention Outcome Study. Diabetologia. 2011;54(2):300–7. [DOI] [PubMed] [Google Scholar]
- 31.Cohen DJ, Reynolds MR. Interpreting the results of cost-effectiveness studies. J Am Coll Cardiol. 2008;52(25):2119–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Sanders GD, Neumann PJ, Basu A, Brock DW, Feeny D, Krahn M, et al. Recommendations for Conduct, Methodological Practices, and Reporting of Cost-effectiveness Analyses: Second Panel on CostEffectiveness in Health and Medicine. JAMA. 2016;316(10):1093–103. [DOI] [PubMed] [Google Scholar]