Effectiveness of a canteen and a behavioural worksite intervention to lower cardiometabolic risk in South Africa

Abstract

Background: Chronic lifestyle diseases like type-2 diabetes, hypertension and dyslipidaemia are modifiable; however, these non-communicable diseases (NCDs) are set to outpace communicable diseases in South Africa. The South African population has a high prevalence of NCDs, including genetic lipid disorders, diabetes and hypertension, meaning that the risk for undiagnosed conditions, like prediabetes and prehypertension, to develop to diabetes and hypertension, respectively, is high. As employees spend much of their waking hours at work, an opportunity exists to engage with this subset of the population for targeted NCD reduction goals. This study used the worksite setting as a backdrop for the implementation of targeted interventions aimed at reducing employee cardiometabolic risk. The resulting gains have economic significance for the employee, the employer, and the country. It also has direct potential to improve employees' quality of life. Aim: To measure the effectiveness of a canteen and behavioural intervention versus a canteen-only intervention among participants by evaluating the change in number of individuals reaching two or more cardiometabolic risk goals, namely reductions in blood pressure, triglycerides, and glycated haemoglobin (HbA1c) - the primary outcome, and through changes in secondary outcomes including rates of type-2 diabetes prevalence and regression to normoglycemia and changes in anthropometry, lipids, and glucose. Methodology: This two-arm randomised controlled trial (RCT) featuring a canteen and behavioural arm (CB) and a canteen-only arm (CO), was structured to provide a six-week intervention to employees at two multinational companies spread across eight worksites. Prior to randomisation, all employees were informed about the pending study through information packs distributed at employee staff meetings and during promotional visits to worksites. The information packs contained details on the study, the purpose, duration, partners, expectations, and privacy clause. A two-step screening process was followed to recruit eligible employees for the study. In step one, consenting employees from the eight worksites were screened for inclusion in the study using an eight-question survey. In step two, eligible participants from step one underwent clinical tests (HbA1c, lipids and blood pressure measurements). Consenting employees who were either prediabetic or prehypertensive completed the baseline assessments, which included anthropometry, a demographic and lifestyle survey, a dietary questionnaire, the Global Physical Activity Questionnaire (GPAQ) and the 24-hour food recall. Participants were randomised to the CB and CO treatment groups. The CO group received six weeks of canteen intervention (changes to enable a healthy food environment). In comparison, the CB group received six weeks of canteen intervention along with a behavioural intervention. The CB intervention included an intense six-week lifestyle programme based on the Diabetes Prevention Programme (DPP). The lifestyle classes were held two days per week at each worksite, with three time slots per day to facilitate employee attendance. After the intervention period, a post-test was used to repeat the clinical tests (HbA1c, lipids and blood pressure measurements), measurements for anthropometry and the 24- hour food recall. Data were analysed to assess the effectiveness of the CB and the CO intervention on cardiometabolic risk factors among prediabetic and prehypertensive employees. Diet quality was assessed through the dietary quality questionnaire (DQQ) indicators, and the Framingham Risk Score was used to calculate participants' 10-year risk for developing cardiovascular disease (CVD). Results: Out of a potential pool of 3000 employees, 797 employees participated in the screening process. After applying exclusion criteria and obtaining consent, 147 employees agreed to participate in the RCT. Of these, 72 were assigned to the CB arm, and 75 to the CO arm. In this study, success was defined by a systolic blood pressure decrease ≥ 5mmHg, a decrease in plasma triglycerides ≥0.1 mmol/L and a decrease of ≥0.5% in HbA1c. In the CO intervention arm, twenty-two participants met no improvement of cardiometabolic risk factors while 29 (19.7%) participants met one, 23 (15.6%) participants improved two; and one (0.68%) participant improved all three cardiometabolic risk factors. In the CB intervention group, 21 (14.2%) participants met no improvement of cardiometabolic risk factors, 38 (25.8%) participants improved one risk factor, 13 (8.8%) participants improved two, and none improved three cardiometabolic risk factors. To evaluate the effectiveness of the CO arm on diabetes risk, 6 (4.0%) participants met the intended intervention effect; however, 69 (46.9%) did not. From baseline (BL) to endline (EL), 2 (1.3%) participants were diagnosed as diabetic, 20 (13.6%) participants presenting with prediabetes at BL increased to 23 (15.6%) presenting with prediabetes at EL and finally, 53 (36%) participants without diabetes decreased to 50 (34%) at EL. To measure the effectiveness of a CB intervention on diabetes risk, 6 (4%) participants met the intended intervention effect however, 66 (44.8%) did not. From BL to EL, 1 (0.68%) participant was diagnosed as diabetic, 22 (14.9%) participants at BL increased to 26 (1.6%) presenting with prediabetes and finally, 49 (33.3%) participants without diabetes decreased to 45 (30.6%) at EL. To evaluate the combined effect of a CB intervention versus a CO intervention on diabetes risk, 12 (8.1%) participants met the intended intervention effect; however, 135 (95%) did not. Overall, post intervention improvements were noted for the whole group when no targeted cutoffs were applied, meaning that these participants improved their risk factors but not within projected reduction cutoffs. Conclusion: Cardiometabolic risk factors were improved for some participants using CB or CO interventions at worksites. Given that the CB and CO interventions produced similar results, the CO intervention has the potential to have a broader reach across the entire worksite, regardless of employees' health conditions, as most employees engage with the worksite food environment (canteens, board room meals, tea stations). Promoting an enabling worksite food environment is likely to encourage healthy eating habits. Unlike the CB intervention, which is more resource-intensive, the CO intervention is more feasible to implement. These results were achieved under a challenging COVID-19 lock-down period through the implementation of a 6-week intervention. There is potential to improve these outcomes outside the influence of COVID-19 and by using a longer duration intervention. Collectively, improved employee health has gains for the organisation, their employees and the country. Scope also exists to improve the effectiveness of the outcome outside of a research study and through more integrated communication and support from the different role players, including human resources, shift supervisors, management, occupational health staff and through the election of employee health champions.