Cardiorespiratory Fitness Is Impaired in Type 1 and Type 2 Diabetes: A Systematic Review, Meta-Analysis, and Meta-Regression

Abstract

Introduction: Low cardiorespiratory fitness (CRF) increases the risk of cardiovascular disease by up to eightfold and is one of the strongest predictors of mortality. Some studies demonstrate impaired CRF in people living with type 1 and type 2 diabetes compared with those without diabetes, whereas others demonstrate no diabetes-associated impairment in CRF.

Purpose: We aimed to determine whether diabetes can influence CRF and, if so, identify clinical associations underlying diabetes-associated exercise impairments.

Methods: Sixty-eight studies were included in the quantitative analysis. Standardized mean difference (SMD) was calculated and meta-analyses and meta-regressions were performed by using a random-effects model.

Results: Diabetes is associated with a large negative effect on CRF (SMD = -0.80, P < 0.001)-an effect that is partially mitigated, but still significant, in those with high physical activity levels (SMD = -0.50, P = 0.007). A sedentary lifestyle (SMD = -0.83, P = 0.007) and the presence of clinical complications related to diabetes (SMD = -1.66, P < 0.001) predict a greater magnitude of CRF reduction in people with diabetes compared with controls without diabetes. Both type 1 diabetes and type 2 diabetes are independently associated with impaired CRF compared with controls without diabetes; however, the effect is significantly greater in those type 2 diabetes (SMD = -0.97, P < 0.001). Meta-regression analysis demonstrates the effects of diabetes on CRF are primarily associated with HbA1c levels for type 1 diabetes ( B = -0.07, P < 0.001) and body mass index for type 2 diabetes ( B = -0.17, P = 0.005).

Conclusions: These data demonstrate a negative influence of diabetes on the key risk factor of low CRF and provide critical insight into specific clinical markers of low CRF associated with diabetes.

FIGURE 1

Flow diagram of study selection process.

FIGURE 2

Summary of the overall and subgroup meta-analysis assessing the impact of the PAS on CRF. T1D, type 1 diabetes; T2D, type 2 diabetes. The forest plot for the overall and subgroup meta-analysis are available in the Supplemental Digital Content (Fig. S4—Supplemental Digital Content 7; Fig. S5—Supplemental Digital Content 8; Fig. S6—Supplemental Digital Content 9; Fig. S7—Supplemental Digital Content 10; and Fig. S8—Supplemental Digital Content 11).

FIGURE 3

Meta-regression analysis of the relationship between (A) age, (B) changes in BMI (∆BMI), (C) glycosylated hemoglobin (HbA1c), and (D) DBP and CRF in subjects with type 1 diabetes and type 2 diabetes combined. Each effect is represented by a circle, and the size of a circle reflects the degree of weighting for that data point. The line reflects the best fit for the linear meta-regression. ΔBMI (difference between BMI from diabetes and control group) was presented instead of absolute BMI because some experimental groups from the included studies were not BMI matched. Therefore, using ΔBMI allows us to make a more precise judgment regarding the contribution of BMI on CRF when analyzing a pool of studies in which some of the groups were and others were not BMI matched.

FIGURE 4

Meta-regression analysis of the relationship between (A) changes in BMI difference (∆BMI) and (B) glycosylated hemoglobin (HbA1c) and CRF in subjects with type 1 diabetes and type 2 diabetes, respectively. Each effect is represented by a circle, and the size of a circle reflects the degree of weighting for that data point. The line reflects the best fit for the linear meta-regression.