Table 3

Results from multivariate analysis among women for adjusted RRs* (95% CI) for seeking care because of neck or shoulder pain associated with different factors in subjects with long term exposure† compared with long term unexposed subjects (different analyses, one including factors with lower 95% confidence limit ≥0.9 in table 2 and three separate analyses including interactive effects)

	Women: 210 cases, 698 controls
Multivariate analyses	RR*†	95% CI
*Adjusted for age (5 year age groups), earlier neck or shoulder disorders, irregular working hours, and socioeconomic group; †defined as current exposure and exposure 5 years ago to physical factors, and exposure during a 5 year period to psychosocial factors; ‡(Borg, 1970); §reports of exposure more than 1 day/week, alternatives, hardly ever/never, 1–3 days/month, 1 day/week, 2–4 days/week, every working day.
Work related factors:
High perceived workload ≥14 on the RPE scale‡	1.3	0.7 to 2.3
Hindrance at work	1.5	0.9 to 2.5
Frequent hand or finger movements at work§	1.9	1.2 to 3.0
Interactive effects between work related factors:
High perceived workload ≥14 on the RPE scale‡+hindrance at work	2.0	1.0 to 4.0
Hindrance at work but no high perceived workload	1.0	0.6 to 1.7
High perceived workload but no hindrance at work	0.9	0.4 to 1.6
Interactive effects between work related factors:
Work requiring frequent bending and twisting§+hindrance at work	2.2	1.3 to 6.7
Work requiring frequent bending and twisting but no hindrance at work	1.0	0.6 to 1.5
Hindrance at work but no work requiring frequent bending and twisting	0.8	0.4 to 1.4
Interactive effects between work and non-work related factors:
Frequent hand or finger movements at work§+75% of the spare time used for housework	2.3	1.2 to 4.3
Frequent hand or finger movements at work but not a lot of housework	1.3	0.9 to 1.9
A lot of house work but not frequent hand or finger movements at work	0.4	0.1 to 0.9