Table 3.

Studies measuring mediating or moderating effects and osteoarthritis outcomes.

Study (year)	Type of OA	Independent variables	Data source/population	Health outcomes	Covariates measured	Outcome	Country	Ref.
Sandmark et al. (2000)	Knee OA	Job titles and exposure to physical load in occupational work, housework and leisure-time activities	Men and women who had prosthetic surgery due to primary tibiofemoral OA	Development of knee OA leading to prosthetic surgery	Age, BMI, sports, for women: hormone substitution	Among men there was an association between lifting at work, squatting or knee bending, kneeling and jumping with knee OA	Sweden	[146]
Allen et al. (2010)	Knee and hip OA	Ten occupational tasks and lifetime exposure to jobs that require spending >50% of their time doing five specific tasks	Johnston County Osteoarthritis Project (NC, USA), African–American and Caucasian men and women aged ≥45 years in rural NC (USA)	Presence of knee and hip OA	Age, race, gender, BMI, prior knee or hip injury, smoking	Radiographic hip and knee OA not significantly associated with any occupational tasks; specific activities associated with symptomatic knee and hip OA; exposure to a greater number of physically demanding occupational tasks at the longest job was associated with greater odds of symptomatic knee and hip OA	USA	[102]
Andersen et al. (2012)	Knee and hip OA	Occupations with heavy workload (i.e., farming, construction, healthcare work)	National patient register data on the Danish working population from 1981 to 2006 followed-up for hip and knee OA during 1996–2006	Increased risk of OA, particularly for women	Age, personal income, personal unemployment, previous knee injury	Male floor layers and bricklayers and male and female healthcare assistants had the highest risk of knee OA; farmers had the highest risk of hip OA; generally, risk increased with cumulative years in the occupation	Denmark	[150]
Keysor et al. (2010)	Incident and symptomatic knee OA	Community mobility barriers, transportation facilitators	Multicenter Osteoarthritis Study Knee Pain and Disability subcohort	Disability: DAL and DAF, prevalence of barriers	Age, gender, race, educational attainment, comorbidity, BMI, knee pain, functional limitation and study site	People who reported more transportation facilitators reported less disability, whereas a greater presence of transportation facilitators was not associated with DAF	USA	[89]
Martin et al. (2011)	Hip and knee OA	Participant identified community factors	Johnston County Osteoarthritis Project (NC, USA), African–American and Caucasian men and women aged ≥45 years in rural NC (USA)	Facilitators/barriers to OA self-management	Not applicable	Individuals use local recreational facilities, senior centers, shopping centers, religious organizations, medical providers, pharmacies, and their social network for OA management; built environmental characteristics both facilitate and hinder use of community resources for OA management	USA	[39]
Ethgen et al. (2004)	Knee and hip OA	Social support	Patients with consecutive hip and knee OA attending an outpatient physical rehabilitation and rheumatology clinic of University Hospital of Liege in Belgium	Health-related quality of life	Age, sex, BMI, number of comorbid conditions, SES, site of survey completion, severity of OA	Greater social companionship transactions associated with higher physical functioning, general health, mental health, social functioning and vitality	Belgium	[166]
Kee (2003)	OA	Sociodemographic characteristics (age, race, education, income), psychosocial factors, health promotion practices and social support, health status and pain	Older adults	Psychological equilibrium, physical function	Not available	Hardiness, pain and social support were significant contributors to psychological status and physical function	USA	[168]
Rapp et al. (2000)	Knee OA	Pain coping skills	Older adults	Self-reported disability, walking disability	Demographic variables, BMI, medical comorbidities, history of fractures, vision and hearing impairment	Pain coping was significantly associated with disability and distance walked. Less catastrophic thinking and prayer was associated with less disability and better physical function	USA	[171]
Keefe et al. (2000)	Knee OA	Gender, catastrophizing	Patients with OA	Pain, pain behavior, physical disability	Age, ethnic background, income, education level, number of years since OA diagnosis, generalized or localized OA, disability and financial compensation or not	Significant difference in pain, pain behavior, and physical disability in men and women, mediated by catastrophizing	USA	[172]
Hausmann et al. (2011)	OA	Past perceived discrimination, classism	African–American and white patients treated for OA by orthopedic surgeons in two Veteran Affairs facilities	Affective tone of patient–provider communication	Race, age, annual income, highest educational attainment, whether patient had been seen in the orthopedic clinic previously, quality of life	Perceived racism and classism were reported by more African–American patients than by white patients. High levels of perceived racism among African–American patients was associated with less positive nonverbal affect	USA	[174]

DAF: Daily activity frequency; DAL: Daily activity limitation; OA: Osteoarthritis; SES: Socioeconomic status.