The role of perseverative cognition in the job strain-health outcome relationship

Figures & data

Table 1. Interrelationships between study variables (SD).

	M	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age	28.90	10.90	—
2. Gender	1.43	0.50	.24***	—
3. BMI	25.30	7.88	.08	−.02	—
4. Hours work week	31.50	15.5	.27***	.01	.03	—
5. Job Strain	0.96	0.07	−.02	.06^†	.09*	.05	—
6. General worry	46.70	9.41	−.19***	.23***	.01	−.07**	.21***	—
7. Work-rel worry	11.90	4.00	−.01**	.06	.04	−.08*	.35***	.55***	—
8. General rum	24.80	7.06	−.22***	.07	.06	−.10**	.25***	.62***	.52***	—
9. Work-rel rum	42.70	7.16	−.09**	.04	.04^†	−.05	.19***	.34***	.62***	.42***	—
10. Neuroticism	31.60	8.00	−.12***	.18***	.03	−.09*	.24***	.44***	.44***	.50***	.18***	—
11. Burnout	21.60	4.85	−.08*	.10**	.12**	.04	.43***	.48***	.64***	.47***	.57***	.43***	—
12. HBs	49.80	12.10	−.03	−.18***	.04	.01	.03	.05	.02	−.01	−.02	.21***	.11**	—
13. Somatization	13.70	6.39	−.17***	.07	.13**	−.08*	.29***	.42***	.46***	.49***	.38***	.40***	.46***	.02	—
14. Sleep quality	10.00	4.34	−.01	.14***	.05	−.07*	.24***	.33***	31***	.30***	.17***	.44***	.38***	.19***	.38***	—

^*** p < .001; ^**p < .01; ^*p < .05, †p = .05 - 0.7; for gender 1 = male, 2 = female; Hours work week = number of hours worked per week; Work-rel worry = work-related worry; General rum = general rumination; Work-rel rum = work-related rumination; HBs = health behaviours.

Table 2. Participant demographics.

Characteristics
Age (years), M (SD)	28.9 (10.9)
Gender, % female (N)	43.23 (281)
Hours worked per week, M (SD)	31.5 (15.5)
Body Mass Index, M (SD)	25.3 (7.88)
Ethnicity, N (%)
White	542 (83.38)
Mixed	27 (4.15)
Asian	16 (2.46)
Black	11 (1.69)
Other	54 (8.31)
Nationality, N (%)
UK & Ireland	176 (27.08)
North or West Europe	22 (3.38)
USA or Canada	24 (3.69)
Central or Eastern Europe	186 (28.62)
South Europe	176 (27.08)
Latin America, South Pacific, Middle East, African	66 (10.15)
Occupation, N (%)
Managers (e.g. chief executives, legislators)	33 (5.08)
Trained professionals (e.g. scientists, lawyer)	143 (22.00)
Technicians and associate professionals (e.g. health assistants, business service agents)	105 (16.15)
Clerical support workers (e.g. secretarial, clerks)	13 (2.00)
Service and sales workers (e.g. waiters, hairdressers, traders)	103 (16.85)
Skilled agriculture (e.g. crop growers, animal producers).	7 (1.08)
Craft workers (e.g. building trade, garment trade)	17 (2.62)
Plant and machine operators (e.g. mining, truck drivers).	18 (2.77)
Elementary occupations (e.g. cleaners, labourer)	203 (31.23)
Armed forces (e.g. army, navy)	8 (1.23)
Education, N (%)
Some secondary school	118 (18.15)
GCSE (or equivalent)	53 (8.15)
A-level (or equivalent)	93 (14.31)
Foundation degree (or equivalent)	38 (5.85)
Degree	201 (30.92)
Masters	118 (18.15)
PhD	29 (4.46)
Mental health condition, N (%)
Yes	102 (15.69)
Physical health condition, N (%)
Yes	101 (15.54)

Table 3. Mediation analysis for job strain, PC, and health outcomes.

Predictor	Outcome	Effect	b (95% CI)	S. E	R^2
Job strain	Burnout	Total	28.97***	2.32
		Direct	15.94***	1.88
		Indirect	12.99*** (9.62 − 16.62)	1.70	.54***
Job strain	Somatization	Total	25.01***	3.26
		Direct	10.91***	3.00
		Indirect	14.16*** (10.15 − 18.35)	2.04	.33***
Job strain	Health behaviours	Total	5.66	6.43
		Direct	4.59	6.89
		Indirect	1.03 (-.4.22 − 6.67)	2.79	.01
Job strain	Sleep quality	Total	13.42***	2.25
		Direct	7.06**	2.28
		Indirect	6.36*** (4.31 − 8.58)	1.09	.16***

Note: General worry/rumination as well as work-related worry/rumination are the mediator terms in these mediation models. Beta-coefficients are unstandardized; p-values and 95% CI’s for indirect effects are from 5,000 bootstrapped samples; ***p < .001 **p < .01 * p < .05.

Figure 1. Mediation path analysis for Job Strain, PC, and Health Outcomes. Note: The figure shows the individual paths from job strain to each outcome, via each type of PC. With the exception of the paths from all types of PC to health behaviours, as well as the path from work-related rumination to sleep, and general rumination to burnout, all paths from job strain to PC and from PC to health outcomes (and the indirect paths) were significant. Numbers reflect the unstandardized regression coefficients; C = the total effect of x on y; C’ = the direct effect of x on y; ***p < .001; **p < .01; *p < .05.