The quality and quantity of saliva produced at rest and during exercise is thought to be important in defending the body against pathogens affecting the upper respiratory tract (Brandzaeg, 1992; Gleeson, 2000). Exercise has been reported to affect saliva flow rate and decrease the concentration and/or secretion rate of immunoglobulin A (IgA), the major secretory antibody in saliva (Chicarro et al. 1998; Gleeson, 2000). Previous studies have investigated the effect of exercise on unstimulated saliva flow rate and composition, but no information is available on the responses when saliva flow rate is stimulated. Since athletes may consume food items or drinks during exercise, it is of interest to know how exercise affects stimulated saliva flow composition.

In the present study, with local ethics committee approval, nine healthy men (age 21 ± 1 years, body mass 75.4 ± 3.0 kg,V_O2,max 49.2 ± 3.1 ml kg^-1 min^-1, means ± S.E.M.) volunteered to participate in the study. Resting unstimulated (UNSTIM) and stimulated (STIM) saliva samples were obtained by dribbling into a pre-weighed tube for a 2 min period. For STIM, subjects were given a mint to suck for 1 min prior to saliva collection (Rudney et al. 1985). Following an overnight fast, subjects cycled on an ergometer at a work rate equivalent to 85 %V_O2,max until volitional fatigue (24 ± 5 min). Timed UNSTIM and STIM samples were taken pre-exercise, after 10 min of exercise and 30 min post-exercise. UNSTIM samples were always collected first. Saliva samples were stored at -20 °C and centrifuged to remove any sediment prior to analysis. Saliva volume was estimated by weighing to the nearest mg and saliva density was assumed to be 1 g ml^-1. Saliva flow rate was determined by dividing the volume of saliva by the collection time. The s-IgA concentration was determined using a sandwich type ELISA method (Blannin et al. 1998). Saliva IgA secretion rate was calculated by multiplying the saliva flow rate by the IgA concentration. Results were analysed using a two-factor (trial X time) repeated measures ANOVA with post-hoc Tukey and paired t tests applied where appropriate. Statistical significance was accepted at P < 0.05.

STIM saliva flow rate was ~3-fold greater than UNSTIM (Table 1) both at rest and during exercise, though exercising saliva flow rate was significantly lower for both treatments than at rest. IgA concentration was significantly lower in STIM compared with UNSTIM saliva at all time points. IgA concentration increased during exercise in both STIM and UNSTIM saliva. IgA secretion rate was higher in STIM compared with UNSTIM at all time points and IgA secretion rate fell in both treatments during exercise.

Our findings indicate that stimulating saliva flow rate at rest and during high-intensity exercise results in higher rates of IgA secretion, which might be beneficial against oral pathogens. It is possible, of course, that this may only be a temporary phenomenon and may be due to a washout effect.