A different approach to diabetes prevention has been attempted in small numbers of studies in adults and children. These studies have focused on community-, church- or school-based interventions aimed at improving cardiovascular (CV) and diabetes risk factors. Such studies are difficult to evaluate using the usual RCT paradigm since randomization is often not feasible or is not at the individual level (e.g. schools or classrooms may be the unit of randomization, and the number of randomization units are often small). Limited individual assessments may be available. Often the goals include increased knowledge and behavioral change, as well as sustainability of the intervention in addition to changes in physiological parameters. They are briefly reviewed here, since community-based interventions may be the most economically feasible, if they can be shown to be effective.
Ramaiya et al. reported a six-year follow-up of a community health education program among a Hindu Indian subcommunity from Dar Es Salaam, Tan-zania81. Using a pre-post design, this program resulted in a decrease in diabetes prevalence from 11.8% to 8.2%, and IGT prevalence was said to decrease from 26.5% to 10%. Small significant reductions in fasting and 2 h glucose levels, lipids, BP and weight were noted, along with an increase in physical activity. The senior author has indicated that only the abstract has been published and no follow-up has been continued (George Alberti, University of Newcastle, UK, personal communication). Because published details are limited, it is not clear whether this community intervention, which superficially appears successful, was due to selection and non-response bias, regression to the mean, or to a real effect. If the results are true, it would lend strong support to the idea that community-based diabetes interventions in developing countries can work.
Simmons et al. developed a risk factor reduction program for Independent Samoans (from the former Western Samoa) at high risk of type 2 diabetes living in Auckland, New Zealand82. A church-based intervention was developed in two churches over two years. The intervention subjects had diabetes awareness sessions, physical activity groups, nutrition education and cooking demonstrations. The control church members, disappointed at having to wait for the intervention, had lower participation in glucose tolerance testing, and started a physical activity group which did not continue.
Over two years, subjects attending the intervention church had no weight gain (compared with +3.1 kg on average in the control group), a decreased waist circumference, and increased knowledge about diabetes and nutrition. Self-reported intake of high-fat foods was lower after the intervention, and the majority of respondents felt it had been useful to them. The intervention church continued the physical activity and nutrition program on a self-sustained basis. It appears that this type of intervention had some success in preventing weight gain (though it did not achieve weight loss).
Due to the non-response at the second evaluation (20% intervention, 25% control) it is possible that response bias 'caused' the effects seen. Since no glucose tolerance testing was done at follow-up, the actual impact on diabetes and glucose intolerance is unknown. Given these limitations, the primary lesson may be that it is possible to motivate social groups to learn more about diabetes prevention, to increase their knowledge and to sustain some of the activities after investigators depart. The program was extended to nine other churches after the completion of this pilot study, largely by community members themselves.
Several studies have attempted to increase knowledge and awareness of diabetes and risk factors in children. These include a small hospital-based health education program83 and more extensive school-based interventions for fourth grade84 (aged ~9 years), and fifth grade (aged ~10 years) Mexican American students85. While each of these found some pre- to post-intervention improvements in knowledge and reported behavior, the failure to use controls or any random assignment, lack of direct observations of behavior (in contrast to self-reported behavior), as well as the short duration and small numbers of subjects limits the interpretability and utility of the findings.
Other studies have not focused specifically on diabetes, but on risk factors such as physical activity and obesity in children. A large cardiovascular primary prevention project in elementary schools had few of the deficiencies of smaller pilot studies. The Child and Adolescent Trial for Cardiovascular Health (CATCH) randomized 56 intervention and 40 control schools86. Outcomes were assessed using a pre-post design. Over 5100 initially third grade students (~8 years of age) from ethnically diverse backgrounds in state schools located in California, Louisiana, Minnesota and Texas participated in a third-grade to fifth-grade intervention including school food service modifications, enhanced physical education (PE), and classroom health curricula.
In intervention school lunches, the percentage of energy intake from fat fell more than in control school lunches. Physical activity intensity increased in PE classes in the intervention schools compared with the control schools. Self-reported daily energy intake from fat among students in the intervention schools was significantly reduced by small amounts (from 32.7% to 30.3% of energy from fat) compared with that among students in the control schools (from 32.6% to 32.2%) (p < 0.001).
Intervention school pupils reported significantly more daily vigorous activity than controls (58.6 minutes versus 46.5 minutes of activity). However, body size, BP and cholesterol measures did not differ significantly between treatment groups. Thus, some important evidence of changes in school lunch content and self-reported behavior occurred, although over the three-year period no weight loss or prevention of weight gain occurred. The CATCH intervention is now being implemented in many of the schools in Texas, as well as in other regions, and its dissemination and sustainability are being studied. Another shorter-term CVD prevention activity in schools has also been reported, with some success in decreasing cholesterol, and small decreases in body fat compared to non-intervened children87.
Two randomized trials of school-based interventions to reduce obesity suggest that schools may be an effective place to counter obesity trends in the USA88,89. The 'Planet Health' intervention randomized 10 schools from four communities to intervention or control status in Boston88. Gortmaker and colleagues used an extensive interdisciplinary curricular approach focused on improving activity and dietary behaviors of all pupils in grades 6 and 7 (aged ~11-12 years), without specifically targeting obese children. Curricular relevance to other school objectives was used to gain school acceptance of this intensive intervention.
Obesity prevalence declined in intervention schools from 23.6% to 20.3% compared to increases in control schools from 21.5% to 23.7% . The effect was significant for girls, though not for boys, where larger decreases were seen in control schools. Of interest was that television viewing time was reduced to a greater extent and there was less increase in estimated energy intake as well as an increase in fruit and vegetable intake reported in intervention schools. The decreased television viewing time significantly predicted the weight loss among girls, and was similar across ethnic groups.
The second trial, aimed in this intervention at reducing television viewing time, validated these findings89. In this study, the natural age-related increases in BMI and skin-folds were significantly lower in intervention than control children, and television watching and video game playing was reduced significantly over six months with no change in reported activity or diet.
These studies stand out as positive examples of an otherwise largely negative set of community-based studies aimed at reduction of body fatness90. The limited evidence available suggests that community-based lifestyle approaches may be more useful in changing behavior early in life. They require substantial resources to implement, and are often difficult to evaluate. Of interest are the observations that reduced television watching, at least in the USA, may be an important pathway to reduced obesity rates. Future studies should be done on larger populations over longer time periods to assure that such interventions are effective. Such studies should receive high priority given the rapid increases in diabetes and obesity risks that are occurring across the age span.
Was this article helpful?