AIM: To evaluate the impact of orthostatic load for sensitivity of short-term spectral analysis of heart rate variability (HRV) assessment of potential early autonomic dysfunction in diabetes mellitus. METHODS: Comparison of results of short-term time- and frequency-domain analysis of HRV during single positions and during modified orthostatic load (supine 1-standing-supine 2, each position 300 s) in diabetic subjects with good glycemic control (n=80, age 38+/-14, diabetes duration 16+/-10 years) and without autonomic neuropathy as assessed by a standard bedside reflex test battery, and in nondiabetic controls (n=150, age 40+/-13 years). RESULTS: None of the short-term frequency-domain parameters [absolute and logarithmic (LN) values of spectral powers in total- (TF), low- (LF), and high-frequency (HF) bands and its centroid frequencies] as obtained in single positions "supine" or "standing" revealed a significant difference between well-controlled patients and healthy controls (P>.3). However, during modified orthostatic load, significant differences in DeltaLN TF((supine 1-supine 2)) and in DeltaLN LF((supine 1-supine 2)) as well as in DeltaLN LF((standing-supine 2)) values between diabetic and healthy subjects were recorded [-0.2+/-0.5 vs. -0.1+/-0.4 LN (ms(2)), P=.05; -0.3+/-0.8 vs. 0.1+/-0.7 LN (ms(2)), P=.001 and 0.2+/-1.0 vs. 0.4+/-0.9 LN (ms(2)), P=.05, respectively] with insignificant intergroup differences in related centroid frequencies. This finding suggests a delayed recovery of LF spectral power in diabetic subjects after orthostatic challenge. CONCLUSIONS: When compared with single position measurements, the modified orthostatic load protocol improves the sensitivity of short-term HRV examination. In well-controlled diabetic subjects without cardiovascular autonomic neuropathy (as excluded by standard cardiovascular reflex testing), the delayed recovery of LF band spectral power after orthostatic load with standing up indicates diminished parasympathetic activation.