Table 3

Summary of observational studies of associations between work factors and ill health: non-health care workers

Study	Country	Design	Participants	Response rate	Work factors	Outcomes	Results
Bacharach et al, 1991²⁹	USA	Cross sectional	430 public sector engineers	79%	Role conflict, time pressure	Emotional exhaustion (MBI)	High role conflict (B=0.24) associated with emotional exhaustion
Carayon et al, 1995³⁰	USA	Prospective	148 public sector office workers	71%	Job demands, content and control, social support, task clarity and job future ambiguity	Worker strain, including anxiety and depression (McNair)	Predictors of worker strain at one and two years were task clarity and job future ambiguity
Driscoll et al, 1995³¹	USA	Cross sectional	4900 public sector workers	70%	Job demands, control and social support	Anxiety and depression (NIOSH Generic Job Stress Questionnaire)	High demand and low support associated with anxiety (ORs 1.46 and 2.01) and depression (ORs 2.02 and 1.69)
Ferrie et al, 1998³²	UK	Prospective	10308 public sector workers	80%	Job insecurity	Psychological distress (GHQ-30)	Non-significant association
Frese, 1999³³	Germany	Longitudinal	90 male blue collar workers	35–66%	Work intensity, uncertainty, organisational problems, environmental problems, social pressure, social support	Anxiety, depression	Social pressure and support associated with anxiety/depression (r=0.21/0.20 and −0.21/−0.27)
Frone et al, 1995³⁴	USA	Longitudinal	795 employed adults	67%	Work pressure, lack of autonomy, role ambiguity	Depression (CES-D)	Work pressure, lack of autonomy, role ambiguity all associated with depression (Bs=0.10, p<0.01; 0.19, p<0.001; 0.18, p<0.001, respectively)
Fusilier et al, 1987³⁵	USA	Cross sectional	312 police officers and fire fighters	65%	Role conflict and ambiguity, social support	Depression (Caplan)	Low social support associated with depression (B=−0.24, p<0.01)
Karasek, 1979: Study 1³⁶	Sweden	Longitudinal	1896 working males	92% and 85%	Decision latitude and job demands	Depression (amended from American Health Survey), absenteeism	Decision latitude negatively associated with depression and absenteeism (OR=−1.29, p<0.05 and −1.44, p<0.01). Job demands associated with depression (OR=1.45, p<0.001)
Study 2	USA	Cross sectional	911 working males	76%	As above	As above	Decision latitude negatively associated with depression and absenteeism (OR=−1.41, p<0.01 and –2.04, p<0.001); job demands associated with depression (OR=1.20, p<0.05)
Karasek, 1990³⁷	Sweden	Cross sectional	8504 white collar workers	87%	Changes in control over work	Depression and absenteeism	Decreased control associated with depression (p<0.01) and with absenteeism in men (p<0.01) but not women
LaRocco et al, 1980³⁸	USA	Cross sectional	636 male workers	Not reported	Supervisor support, participation, future ambiguity, under-utilisation, workload, role conflict	Anxiety and depression (Cobb and Kasl)	Supervisor support buffers the adverse effect of low participation on depression (p<0.1, significant ) and of future ambiguity on anxiety (p<0.01)
Niedhammer et al, 1998³⁹	France	Prospective	11552	92%	Psychological demands, decision latitude and social support	Depression (CES-D)	High psychological demand, low decision latitude and social support associated with subsequent depression. For men, OR of 1.8, 1.4 and 1.6 respectively and for women, OR of 1.4, 1.4 and 1.3 respectively.
Payne and Fletcher, 1983⁴⁰	UK	Cross sectional	148 teachers	74%	Workload demands, discretion	Anxiety and depression (CCEI)	Association with workload demands (betas=0.117 for anxiety and 0.176 for depression) and negative association with discretion (betas=−0.222 for anxiety and –0.121 for depression)
Reifman et al, 1991⁴¹	USA	Cross sectional and longitudinal	200 married, professional women	>90%	Social support at work , control over work, role ambiguity, workload	Depression (SCL-90)	Cross sectionally, association with social support at work (r=0.37), role ambiguity (r=0.35) and workload (r=0.29). No associations one year later
Romanov et al, 1996⁴²	Finland	Prospective	15530 employees	Not reported	Conflict at work	Psychiatric morbidity (hospital discharge registry)	Positive association (RR 2.18, CI₉₅ 1.34, 3.54)
Rubenowitz et al, 1983⁴³	Sweden	Cross sectional	25 departments of 5 companies. Numbers not reported	85–90%	Perceived participation (individual, group, representative)	Absenteeism	Negative association for individual participation (r=−0.53)
Shields, 1999⁴⁴	Canada	Prospective	3830 working population	80%	Long working hours	Depression (Composite International Diagnostic Interview)	>35 hours per week associated with depression in women (OR=2.2) but not men
Smulders and Nijuis, 1999⁴⁵	The Netherlands	Cross sectional and prospective	1755 male public sector workers	70%	Job control and job demands	Absence rate and absence frequency	Cross sectionally, job control associated with low absence frequency (beta=0.10, p<0.01) and job demand associated with low absence rate (beta=−0.08, p<0.05)
Sparks and Cooper, 1999⁴⁶	UK	Cross sectional	7099 from 13 occupations	Not reported	Work control, career achievement, organisational climate, job factors, home/work interface, work relationships	Mental health (OSI)	All associated (r=0.22 to −0.28, p<0.001)
Stansfeld et al, 1995⁴⁷	UK	Cross sectional	10314 public sector employees	73%	Job variety and skill use, control, social support, work pace, conflicting demands	Psychological distress (GHQ-30)	All significantly associated (intertile trend p values < 0.001)
Stansfeld et al, 1998⁴⁸	UK	Prospective	7372 public sector workers	72%	Job demands, decision latitude, social support and effort-reward imbalance	Psychological functioning (SF-36)	Low support and effort-reward imbalance associated with poor psychological functioning (OR=1.2 for men and 1.4 for women; 1.8 for men and 2.3 for women respectively). In men, low decision latitude (OR=1.2) and in women, high job demand (OR=2.0) were associated with poor psychological functioning