Physical Activity Is Associated With Glucose Tolerance Independent of Microvascular Function: The Maastricht Study.

David Montero, Alfons J. H. M. Houben, Annemarie Koster, Dennis M. J. Muris, Miranda T. Schram, Ed H. Gronenschild, Simone J. S. Sep, Ronald M. A. Henry, Carla J. H. van der Kallen, Nicolaas C. Schaper, Pieter C. Dagnelie, Tineke A. C. M. van Geel, Stef P. J. Kremers, Hans H. C. M. Savelberg & Coen D. A. Stehouwer

Context and Objective:

Moderate-to-vigorous physical activity (MVPA) and physical fitness (PF) are positively associated with glucose tolerance. Such associations may be partly conditioned by microvascular function, which is a common correlate to MVPA, PF, and glucose tolerance. To test this hypothesis, the present study sought to investigate independent associations of MVPA and PF with glucose tolerance and to what extent these associations are mediated by microvascular function.

Design, Setting, Participants, and Outcome Measures:

Data from The Maastricht Study were used (n = 512 for MVPA and n = 488 for PF analyses; mean age, 59 [SD = 9] y, 52 % men). Glucose tolerance was assessed by 2-hour postload plasma glucose levels (2hPG). The total number of weekly hours of MVPA was estimated with the Community Healthy Activities Model Program for Seniors questionnaire. Walking speed during the 6-minute walk test was used to evaluate PF. Microvascular function was determined by postocclusive capillary recruitment and flowmotion with capillaroscopy and laser Doppler flowmetry in skin microcirculation.


In univariate analyses, MVPA, PF, and microvascular function variables were associated with 2hPG. MVPA (n = 512, β = −0.056, P = .019) and PF (n = 488, β = −0.368, P = .006) remained associated with 2hPG after adjustment for established cardio-metabolic risk factors and history of cardiovascular disease; addition of microvascular function variables as potential mediators did not materially change the associations of MVPA (β = −0.054, P = .024) and PF (β = −0.364, P = .006) with 2hPG. No mediation effects of microvascular function variables were detected.

MVPA and PF were independently associated with 2hPG, irrespective of established risk factors and generalized microvascular function. The possibility that specific microvascular functions, eg, insulin-mediated vasodilation, influence the association of MVPA and PF with 2hPG needs further investigation.