Table 3.

Merged data from Study-1 and Study-2: Relationships of Abdominal Pain with Fatigue in Women with Irritable Bowel Syndrome (IBS; N = 356).

	Main models				Models controlling for depression and anxiety (day = t)				Bootstrap results for significances of mediation effectc
	βa	SE	t	pb	βa	SE	t	pb	βa,d	SE	95% CI	z	pb
Acrosse-Women Same Day Model-A: Relationships of Abdominal Pain (day = t) with Fatigue (day = t)
Across	488	.017	10.1	<.001	.333	.017	7.1	<.001	.155	.012	.110 to .192	3.94	<.001
Within-Woman
Same Day Model-A: Relationships of Abdominal Pain (day = t) with Fatigue (day = t)
Within	.174	.010	16.5	<.001	.138	.010	13.3	<.001	.036	.006	.001 to .071	6.04	<.001
Day-to-Day Model-B: Abdominal Pain (day = t) Predicts Next-day Fatigue (day = t+1)
Within	.059	.011	5.5	<.001	.046	.010	4.4	.001	.013	.003	.009 to .021	4.90	<.001
Reverse Model-C: Fatigue (day = t) Predicts Next-day Abdominal Pain (day = t+1)
Within	.020	.011	1.8	.060	.005	.011	0.5	.618	Not applicablef

Note. CI = confidence interval. SE = standard error.

^astandardized β-coefficient.

^btesting the significance of β-coefficient.

^cresamples n = 1,000.

^dmediation effect testing for ‘β in main model’-’β controlling for psychological distress’

^econtrolling for age and study design (Study-1 and Study-2) in across-women analyses.

^fdue to non-significant relationship of reverse model-C.