Upper extremity exercise is associated with a significant metabolic and ventilatory cost that is particularly evident in patients with severe chronic airflow obstruction. In these patients abnormal ventilatory muscle recruitment has been hypothesized to relate to impaired diaphragm function resulting from hyperinflation. Similar data have never been reported in patients with isolated diaphragm weakness but without airflow obstruction or hyperinflation, a group that would ideally define the role of diaphragm function during arm elevation (AE). We prospectively studied 15 patients with isolated diaphragm weakness of varying severity (Pdi(sniff), 31.74 +/- 3.75 cm H(2)O) as contrasted with eight normal subjects (Pdi(sniff), 111. 77 +/- 13.35 cm H(2)O) of similar age. Patients with diaphragm weakness demonstrated significant lung volume restriction with normal DL(CO)/VA. There was no difference in resting oxygen consumption (V O(2)), carbon dioxide production (V CO(2)), minute ventilation (V E), and tidal volume (VT) between the two groups; however, a borderline difference in resting breathing frequency (f(b)) (p = 0.056) was evident. Both groups demonstrated a rise in V O(2), V CO(2), and V E during 2 min of AE anteriorly. Normal subjects demonstrated a statistically significant rise in VT but a statistically insignificant rise in f(b) during AE. In contrast, patients with diaphragm weakness demonstrated a statistically significant rise in f(b) during AE but a statistically insignificant rise in VT. In patients the observed rise in VT directly correlated with baseline Pdi(sniff) (r = 0.59, p = 0.02) and Pdi(max) (r = 0.81, p = 0.002). Both groups demonstrated a rise in Pdi during AE. The rise in Pdi during AE directly correlated to Pdi(sniff) in the patients (r = 0.69, p = 0.004). Observed end-expiratory Ppl rose during arm elevation in both the patient group and in the normal control group, but no evidence of a differential response to AE was found. In those patients with greater diaphragm weakness (Pdi(sniff) < 30 cm H(2)O), abnormal respiratory muscle function (lesser rise in Pdi) and a lesser increase in VT during AE were more evident. These data highlight the importance of diaphragm function in determining the metabolic and respiratory muscle response to arm elevation.