Cardiorespiratory fitness is the primary determinant of cardiac afterload during exercise in type 2 diabetes mellitus
N Kinoshita, S Yamada, E UchiyamaAbstract
Background
Cardiovascular–kidney–metabolic (CKM) syndrome reflects the interconnected burden of cardiovascular, renal, and metabolic dysfunction, including type 2 diabetes mellitus (T2DM). Exaggerated blood pressure response to exercise is a relevant haemodynamic phenotype within this spectrum; however, determinants of exercise haemodynamic load in T2DM remain incompletely defined.
Purpose
To characterise workload-indexed systolic blood pressure at peak and submaximal exercise and to identify its determinants in T2DM, considering KDIGO-defined stage of diabetic nephropathy.
Methods
We retrospectively analysed 430 cardiopulmonary exercise (CPX) tests in individuals with T2DM (60±13 years, 64% male). Exclusion criteria included known cardiovascular disease, systemic illness, and beta-blocker use. CPX was performed on a cycle ergometer using a ramp protocol. Systolic blood pressure indexed to work rate at peak (SBP/W@peak) and at ventilatory threshold (SBP/W@VT) were analysed using multivariable linear regression with forced entry. The core model included age, sex, resting systolic blood pressure, KDIGO stage, and peak oxygen uptake (VO₂peak, mL/min). Medication-adjusted models additionally included SGLT2 inhibitor, DPP-4 inhibitor, ACEi/ARB, and Ca channel blocker use. Sensitivity analysis included ln(HOMA-IR). Statistical significance was defined as p<0.05.
Results
At peak exercise, SBP/W@peak was independently associated with higher age, higher resting SBP, KDIGO stage, and lower VO₂peak, with a marked sex difference (Table 1; adjusted R²=0.694). After medication adjustment, age, sex, resting SBP, and VO₂peak remained significant; KDIGO retained a small independent association (β=0.072), whereas medications were not (adjusted R²=0.698). ln(HOMA-IR) was not independently associated. For SBP/W@VT, a similar pattern was observed: age (β=0.178), resting SBP (β=0.258), sex (β=−0.341), and VO₂peak (β=−0.307), while KDIGO stage and medications were not significant (adjusted R²=0.617). Visual inspection suggested a non-linear inverse association between VO₂peak and SBP/W@peak, with steeper increases in afterload per workload at lower fitness (Figure 1).
Conclusion
Workload-indexed systolic blood pressure, reflecting cardiac afterload per unit mechanical output, is predominantly determined by cardiorespiratory fitness, age, sex, and resting blood pressure. Diabetic nephropathy shows a modest association only at peak exercise, and insulin resistance appears minimal. Lower fitness is associated with higher afterload during both maximal and submaximal exercise, implying greater myocardial loading during routine activities. Within the CKM spectrum, renal impairment may accentuate haemodynamic stress at higher workloads, whereas fitness remains the dominant modifiable determinant. These findings support fitness optimisation to attenuate activity-related myocardial stress and potentially reduce downstream cardiac risk in T2DM.