Intervening to reduce workplace sitting: mediating role of social-cognitive constructs during a cluster randomised controlled trial

Table 4 Mediation of short- and long-term intervention effects on workplace sitting (min/8 h day) by concurrent changes in social-cognitive constructs

Mediators		Direct effect	Indirect effect	Percentage of total intervention effect^d mediated
Mediators		c' (95% CI)	a*b (95% CI)	Percentage of total intervention effect^d mediated
Perceived behavioural control	3 months^a	−106.34 (−129.35, −75.37)	−8.57 (−18.46, −3.57)	7.46%
Perceived behavioural control	12 months^b	−74.52 (−114.78, −26.61)	−0.28 (−11.03, 12.99)	0.37%
Barrier self-efficacy	3 months^a	−110.15 (−135.25, −81.62)	−4.89 (−17.12, 3.82)	4.25%
Barrier self-efficacy	12 months^c	−64.09 (−105.64, −23.41)	−10.34 (−27.26, −2.16)	13.89%
Perceived organisational norms	3 months^a	−113.74 (−133.79, −85.90)	−1.31 (−5.49, 1.94)	1.14%
Perceived organisational norms	12 months^c	−74.85 (−112.07, −30.85)	−0.04 (−5.44, 2.86)	0.05%
Knowledge	3 months^a	−112.85 (−134.65, −85.80)	−2.07 (−8.31, 0.45)	1.80%
Knowledge	12 months^b	−73.59 (−109.00, −32.54)	−0.03 (−3.85, 4.15)	0.05%

Note: For each construct, minimum score = 1 and maximum score = 5
Models adjusted for baseline values of the following potential confounders: workplace sitting (min/8 h), age (years), gender (men/women), Caucasian ethnicity (yes/no), current smoking (yes/no), body mass index (log-transformed), AQoL-8D physical superdomain score (log-transformed), AQoL-8D mental superdomain score (log-transformed), TV viewing time (log-transformed), job control category (high/low), weekly headaches (yes/no), musculoskeletal symptoms in the upper extremities (none/does not interfere with activities/interferes with activities)
^a n = 186; ^b n = 146 ^c n = 145 ^d Total intervention effect (c) comprises direct effect that occurs independently of the mediator (c’) and the indirect effect (ab) that that occurs via the mediator

ISSN: 1479-5868