BackgroundSalivary alpha amylase (sAA) and chromogranin A (sCgA) have both been suggested as non-invasive markers for sympathetic nervous system (SNS) activity. A complete cervical spinal cord injury leading to tetraplegia is accompanied with sympathetic dysfunction; the aim of this study was to establish the exercise response of these markers in this in vivo model.MethodsTwenty-six elite male wheelchair athletes (C6–C7 tetraplegia: N = 8, T6–L1 paraplegia: N = 10 and non-spinal cord injured controls: N = 8) performed treadmill exercise to exhaustion. Saliva and blood samples were taken pre, post and 30 min post exercise and analysed for sAA, sCgA and plasma adrenaline concentration, respectively.ResultsIn all three subgroups, sAA and sCgA were elevated post exercise (P < 0.05). Whilst sCgA was not different between subgroups, a group × time interaction for sAA explained the reduced post-exercise sAA activity in tetraplegia (162 ± 127 vs 313 ± 99 (paraplegia) and 328 ± 131 U mL−1 (controls), P = 0.005). The post-exercise increase in adrenaline was not apparent in tetraplegia (P = 0.74). A significant correlation was found between adrenaline and sAA (r = 0.60, P = 0.01), but not between adrenaline and sCgA (r = 0.06, P = 0.79).ConclusionsThe blunted post-exercise rise in sAA and adrenaline in tetraplegia implies that both reflect SNS activity to some degree. It is questionable whether sCgA should be used as a marker for SNS activity, both due to the exercise response which is not different between the subgroups and its non-significant relationship with adrenaline.