Improved Metabolic Risk Markers Following Two 6-Month Physical Activity Programs Among Socioeconomic Marginalized Women of Native American Ancestry in Lima, Peru

The study population consisted of 142 Amerindian women. They were all examined in 1999 (6) and had a normal fasting plasma glucose concentration (range 2.8–5.6 mmol/l). Five years later, a total of 83 women participated in a follow-up examination (7). Of these, 76 consented to take part in the present training study. Mean age was 41 years (range 25–64 years). The women were randomly assigned into group A (one training session per week) or group B (three training sessions per week). No economic compensation was given besides free athlete suits and shoes.

All exercise sessions were verified by the direct supervision by a physiotherapist and took place outdoors on a square with a concrete surface. A warm-up with stretching, light jogging, and flexibility movements was followed by a mixture of traditional folk and modern aerobic dances for a total of 60 min. This was undertaken once or three times per week for 6 months.

Body weight and height, BMI, waist circumference, fasting plasma glucose, and cardiorespiratory capacity (Vo_2max) were measured at baseline and after the intervention period. All subjects refrained from any severe physical activity 48 h before the measurements. Glucose was determined with the glucose oxidase technique. Vo_2max was estimated indirectly after subjects ran around two cones, spaced 20 m apart, for 12 min, as previously decribed (8).

Comparisons within groups were performed by paired Student's t test and between groups by Mann-Whitney U test. Correlations were performed using Spearman rank correlations and multiple regression analysis for training sessions, changes of plasma glucose, body weight, BMI, waist circumference, and Vo_2max. A P value of <0.05 was considered statistically significant.

Fifty-nine patients completed the study, with 33 in group A and 26 in group B. The mean total attendance was 21 in group A (maximal possible 27) and 64 in group B (maximal possible 77). During the 6-month study period, 20% in group A (one session per week) and 16% in group B (three sessions per week) discontinued the training sessions.

No significant differences were observed between groups A and B in 1999 before the start of the exercise program (P > 0.05) (Table 1). In both groups, plasma glucose decreased (P < 0.01), waist circumference decreased (P < 0.05), and Vo_2max increased significantly (P < 0.05) after the training sessions. These effects were more pronounced among members of group B (P = 0.019).

The number of training sessions was associated with changes in plasma glucose (Rho = 0.37, P = 0.013), waist circumference (Rho = 0.29, P = 0.034), and Vo_2max (Rho = 0.38, P = 0,009). In contrast, these variables did not correlate significantly with BMI. Multiple regression analysis showed that the number of training sessions and the increase of Vo_2max independently contributed to changes in plasma glucose, whereas waist circumference and BMI did not correlate significantly.

The dramatic increase of type 2 diabetes in Latin America among subjects with Amerindian heritage (9–11) is seen mainly among populations with low socioeconomic status. For prevention and management, these at-risk populations require a tailored approach. This study shows that a supervised physical group activity is feasible among women living in poor urban areas. The attendance rate was high and the drop out rate acceptable. Furthermore, the study shows that after following such an activity for 6 months, circulating glucose and waist circumference are reduced and Vo_2max increased. Furthermore, the number of exercise sessions independently contributed to the reduction of glucose concentrations. In fact, in group B, the mean plasma glucose concentration decreased to the same level as in 1999. Also, in group A, the glucose concentration was significantly reduced, although not to the same degree as in group B.

The exercise regime resulted in a reduction of waist circumference but with no significant change in body weight. Exercise in women is associated with a marked increase in lipolysis in the abdominal subcutaneous adipose tissue in comparison with the femoral adipose tissue (12). This suggests that exercise-induced weight loss would be associated with a preferential reduction in abdominal obesity.

Although encouraging, it may be surprising that only one exercise session per week significantly reduced fasting glucose. This supports results showing that marked changes of the metabolic potential of muscle occur with small variation in the level of physical activity (13). It is also known that men with impaired glucose tolerance normalize their glucose tolerance by increasing their weekly physical activity pattern, without any change in body weight (14).

The findings in this study are extremely encouraging in that supervised exercise may be a low-cost safe therapy with favorable benefits, exercise need not be strenuous or prolonged, and exercise does not have to be done every day.

The study was supported by grants from Stiftelsen för forskning inom diabetes och kärlsjukdom, the Swedish Research Council (Grant 6834), the Albert Påhlsson Foundation, the Swedish Diabetes Foundation, the Region Skåne, and the Faculty of Medicine, Lund University, Lund, Sweden.

M. Benavente Ercilla, M. Tamashiro, and L. Retamozo at the Alternativa Center for Social Research and Popular Education in Lima, Peru, participated in the design and data collection and provided information to participants.