less severe abdominal pain than the control group at week 8 (p = .038). Diarrhoea syndrome was rated less severe in the intervention group than the in control group at week 8(p = .044). Gastrointestinal symptoms assessed in total score at week 8 and diarrhoea syndrome at week 4 wasrated less sever in the intervention group than the in control group but did not reach the level of significance(p = .052). No other significant associations were found.In the multiple linear regression, the intervention group rated significantly less severe gastrointestinal symptoms than the control group in GSRS total score at week 8 both inrelation to baseline (p = .003) and to week 4 (p = .003)(Table 3). Constipation syndrome was significantly less severe in the intervention group than the control group at week 8 both in relation to baseline (p = .014) and week 4(p = .013). Pain syndrome was rated significantly lesssevere in the intervention group than the control group at week 8 in relation to week 4 (p = .010) but not significantly less severe at week 8 in relation to baseline (p = .074).There were no difference in ratings between intervention group and control group in severity of indigestion or diarrhoea syndrome at week 8 (Table 3). There were no significant relations at week 4. The intervention group also reported significantly more bowel movements than the control group at week 8 in relation to baseline (p = .016)than the control group (Table 3).An interaction was found between baseline value of GSRS total score and massage (p = .001); the more severe the symptoms were at baseline the greater the impact of abdominal massage was in week 8. No other interactions were found (Fig. 2).Rating of GSRS total score, constipation syndrome, and pain syndrome were stable over time in the control group while the trend in the intervention group was falling,which means a decrease in severity of symptoms.No significant differences were found between the intervention and control group according to consistency of faeces or stool size analyzed with Wilcoxon signed-rank test or laxative intake analyzed with Mann–Whitney Utest.