Association Between Physical Activity and Fitness in Patients with Heart Failure and Type 2 Diabetes Mellitus: Influence of a Telemedicine Program

Abstract

Highlights: Physical fitness (PF) and physical activity (PA) are closely related in individuals with heart failure and type 2 diabetes mellitus, but our telemedicine program only moderately impacted PF. Specific baseline factors influenced outcomes after six months of intervention. Younger patients with higher baseline PF, HDL cholesterol levels, and PA showed better PA outcomes at the end of the program. Meanwhile, those with lower BMI, younger age, and better baseline PF demonstrated greater improvements in PF. What are the main findings? After our telemedicine program, the 6-Minute Walking Test (6MWT) improved, while PA levels remained unchanged. The association between PA and PF increased from moderate at baseline to strong at the end of the program. PF, age, HDL cholesterol, and baseline PA were key predictors of PA, while BMI, age, and baseline 6MWT influenced PF outcomes. What are the implications of the main findings? PA and PF are distinct parameters that should be measured and monitored independently. Telemedicine programs should include targeted interventions to improve both PF and PA. Stratifying patients based on specific baseline characteristics may support selecting the most appropriate program, enhancing personalization and effectiveness. Background. Few studies have evaluated physical fitness (PF) and physical activity (PA) in individuals with heart failure and type 2 diabetes mellitus and the possibility that some telemedicine programs (TMPs) may impact them. This post hoc subgroup analysis of an RCT study aimed to describe PF and PA in this population before and after a TMP. Methods. We evaluated (a) PF as distance, assessed via the 6-Minute Walking Test (6MWT), and PA as daily step count in this population before and after a TMP; (b) their relationship; and (c) the patient characteristics that influence PF and PA. Results. Fifty-eight patients (aged 71.31 ± 7.92 years old, 84% male, BMI 28.01 ± 4.70 Kg/m2, ejection fraction 48.64 ± 10.64%) were enrolled between August 2022 and September 2024. All patients received a six-month TMP (nurse teleassistance, telemonitoring, a dedicated app, and PA evaluation using a tracker bracelet and step count goals). The 6MWT improved (from 418 ± 113 to 439 ± 120 m, p < 0.001), while PA remained unchanged after the TMP (from 7181 ± 4149 to 7229 ± 4947 steps/day, p = 0.92). The PA and PF association ranged from moderate at baseline (rho = 0.4958, p < 0.001) to strong at the end of the study (rho = 0.6807, p < 0.001). The regression model shows that following the TMP, PA(y) was associated with baseline 6MWT [β= 8.5, 95%IC −0.31; 17.24], age (β = −144.0, 95%IC −262.14; −25.90)), baseline daily steps (β = 0.5967, 95%IC 0.37; 0.82), and HDL cholesterol (β = 119.7, 95%IC 39.07; 200.31) (R2= 0.6580, F(4.53) = 25.49, p < 0.001), while PF(y) was associated with BMI (β= −2.04, 95%IC −4.30; 0.22)), age (β= 0.90, 95%IC −4.4; 1.32), and baseline 6MWT (β = 0.90, 95%IC 0.79; 1.00) (R2 = 0.9007, F(3.54) = 163.29, p < 0.001). Conclusions. Our TMP led to a statistically significant but clinically modest improvement in PF but did not impact PA despite there being variability among patients. PA and PF appear to be interdependent. PF, age, HDL cholesterol, and baseline PA were key predictors of PA, while BMI, age, and baseline 6MWT impacted PF at the end of the TMP.