Published: Sep 30, 2014
Story Source
Download Complimentary Source PDF
By John Gever, Managing Editor, MedPage Today

Action Points

• Physical activity appears to have an overall protective effect on type 2 diabetes.
• Note, however, that this relationship is attenuated in patients at high genetic risk for type 2 diabetes.

Individuals with a relatively large number of genetic factors linked to type 2 diabetes risk showed a smaller protective effect of exercise, a large cohort study indicated.

The finding emerged from an analysis of data from participants in the ARIC (Atherosclerosis Risk in Communities), Yann Klimentidis, PhD, of the University of Arizona in Tucson, and colleagues reported.

When participants were stratified by genetic risk (defined as the presence of single nucleotide polymorphisms [SNPs]) there was little difference in incidence rates for new-onset type 2 diabetes in those with high versus low levels of physical activity.

In contrast, those with low genetic risk scores were half as likely to develop type 2 diabetes when their physical activity was rated as high versus low, the researchers reported in Diabetologia.

“Although physical activity generally protects against type 2 diabetes, our findings suggest that the level of protection conferred by physical activity is lower in individuals (especially women) with a higher genetic risk for type 2 diabetes or insulin resistance,” Klimentidis and colleagues wrote.

The association appeared to be driven mainly by genetic risk for insulin resistance, they added. Separate analysis of genetic risk factors for other aspects of type 2 diabetes showed a weaker or absent correlation between the genetic risk burden and the protective effect of exercise.

Klimentidis and colleagues drew on data from 8,101 ARIC participants, all white, who did not have type 2 diabetes on their first study visit (defined as a fasting glucose level of less than 124 mg/dL and no report of taking anti-diabetic drugs). Follow-up averaged 7.8 years (maximum 12 years), during which 821 participants were diagnosed with type 2 diabetes.

Genetic risk scores for incident type 2 diabetes were developed using 65 SNPs identified in a previous study. Separate risk scores were developed for abnormal fasting glucose and fasting insulin levels on the basis of 36 and 17 SNPs, respectively, identified in a different analysis. The scores were calculated as weighted sums of risk alleles, using the effect sizes found in the previous studies for weighting purposes.

Results were adjusted for baseline factors including sex, age, body mass index, smoking status, dietary fiber, and total carbohydrate intake as reported in food questionnaires, and also for whether patients had normal or impaired fasting glucose (101 to 124 mg/dL) at study entry.

Physical activity was reported at baseline with a questionnaire asking about the types, frequencies, and durations of activities that participants regularly engaged in. Responses were then converted into a “sports index” with values of 1 to 5 in 0.25-point increments.

For several of the current analyses, Klimentidis and colleagues stratified physical activity into three categories (high, intermediate, and low) and did the same for genetic risk scores.

In the primary analysis, they found that participants with low genetic risk scores for incident type 2 diabetes had actual incidences of approximately 50, 80, and 115 per 10,000 person-years for high, intermediate, and low physical activity levels, respectively.

Participants with intermediate genetic risk scores and low physical activity showed an incidence rate of ab0ut 180 per 10,000 person-years, whereas those with intermediate and high physical activity levels both had incidence rates of about 125 per 10,000 person-years.

But there was almost no difference in type 2 diabetes incidence between physical activity strata in the group of participants with high genetic risk scores. Incidence rates ranged only from about 160 to 175 per 10,000 person-years between the activity strata.

Klimentidis and colleagues also noted that the interaction between genetic risk and physical activity was seen primarily in women. The hazard ratio for interaction was 1.59 in female participants (95% CI 1.17-2.15) but only 1.10 in men (95% CI 0.85-1.53). The researchers offered no specific explanation for the sex difference in study results.

“In future studies, it will be important to use less subjective measures of physical activity, to replicate this finding in another large prospective cohort study, to better understand the sex difference observed, and to generalize this finding to other age and ethnic/racial groups,” they concluded.