Table 2

Characteristics and selected results of studies examining occupational kneeling and risk of knee injury (males only)

	Study period	Population	Exposure	Outcome*		OR	(95% CI)
*K&L and Ahlbäck refer to classifications of radiographic appearance of arthrosis of the knee.
†Results given are for male and female subjects combined.
Strong evidence
Felson et al⁴⁴	1983–85	Cohort (US). Sample of population already taking part in heart study (n = 569)	Assessment of physical demand and knee bending requirement of occupations	K&L⩾2	Bending and light work	1.07	(0.53 to 2.17)
Felson et al⁴⁴	1983–85			K&L⩾2	Bending and heavy work	2.22	(1.38 to 3.58)
Kivimäki et al⁴⁶	Not specified	Case control (Finland). Cases: carpet and floor layers (n = 168). Controls: painters (n = 146)	Carpet/floor layers v painters	Osteophytes and knee pain	Floor layers	1.40	(1.10 to 1.70)
Cooper et al⁴⁸†	Not specified	Case control (UK). Cases: knee pain and radiographic OA (n = 109). Controls: no knee pain or OA (n = 218)	Time spent squatting or kneeling in longest held job before symptom onset	K&L⩾3	Squatting	3.70	(0.80 to 16.60)
Cooper et al⁴⁸†	Not specified			Knee pain	Kneeling	1.80	(0.60 to 5.70)
Sandmark et al⁵¹	1991–93	Case control (Sweden). Cases: knee replacement (n = 325). Controls: population register (n = 264)	Frequency of squatting/knee bends during working day. Time spent kneeling	Knee replacement	Squatting/knee bending	2.90	(1.70 to 4.90)
					Kneeling	2.10	(1.40 to 3.30)

Coggon et al⁵²	Not specified	Case control (UK). Cases: waiting list for knee surgery (n = 205). GP controls (n = 205)	Time spent kneeling or squatting in all jobs held for one year or more	Awaiting knee surgery	Kneeling	1.70	(1.00 to 3.00)
Coggon et al⁵²	Not specified			Awaiting knee surgery	Squatting	2.20	(1.00 to 4.90)
Moderate evidence
Sharrard & Liddell¹¹	1958–60	Case control (UK). Cases: meniscectomy patients (n = 957). Controls: appendicectomy patients (n = 1075)	Miners v non-miners	Meniscectomy	<25 years	2.00
					25–54 years	4.00
					⩾55 years	2.00
Anderson & Felson²⁷	1971–75	Cohort (US): Sample aged 35 to 64 years (n = 1853)	Knee bending demand of job at time of interview (dichotomous variable)	K&L⩾2	35–44 years	0.85	(0.20 to 3.61)
					45–54 years	0.82	(0.32 to 2.11)
					55–64 years	2.45	(1.21 to 4.97)
Sahlström & Montgomery⁴⁹†	1982–86	Case control (Sweden). Cases: knee pain and radiographic OA (n = 340). Controls: random selection from population registers (n = 680).	Degree of knee moment for working activity over three 15 year periods	Ahlbäck⩾1	Medium/heavy	1.10	(0.70 to 1.80)
Sahlström & Montgomery⁴⁹†	1982–86			Knee pain
O’Reilly et al⁵⁰	Not specifed	UK postal survey to patients of two GP practices (n = 1961)	Job title (longest held job)	Knee pain	Miners	1.90	(1.30 to 2.80)
Manninen et al⁵³	1992–93	Case control (Finland). Cases: knee arthroplasty (n = 55). Controls: random sample from population registry (n = 140)	Time spent squatting or kneeling in job at age 49 years	Knee arthroplasty	<2 hours per day	0.58	(0.21 to 1.64)
Manninen et al⁵³	1992–93		Time spent squatting or kneeling in job at age 49 years	Knee arthroplasty	⩾2 hours per day	1.68	(0.66 to 4.28)
Miranda et al¹⁷†	1994–95	Cohort (Finland). Employed by forestry company (n = 3312).	Time spent working in a kneeling or squatting position.	Knee pain	High exposure	1.30	(0.70 to 2.30)
Weak evidence
Vingård et al⁴⁵	1980	Cohort (Sweden). Hospital care for OA of the knee born 1905 and 1945 reporting same occupation in two consecutive censuses (n = 221).	Occupational exposure to forces on the lower extremities (high v low)	Hospital care	Born 1905–24	1.20	(0.90 to 1.50)
Vingård et al⁴⁵	1980			Hospital care	Born 1925–45	1.40	(1.10 to 1.90)