Table 3

Characteristics and results of included analytic epidemiological studies

Study*	Study design	Study population	Risk factors	Outcomes	Adjustment for confounders	Results
Only statistically significant results are reported.
*First author and reference number.
CI, confidence interval; m, months; OA, osteoarthritis; OR, odds ratio; PF, patellofemoral; TF, tibiofemoral.
Llana¹³⁷	Cross sectional study	A sample of 146 tennis players in Spain selected from a sample of 4000 who had completed a questionnaire on “discomfort” associated with tennis shoes	Perceived design errors	Discomfort	No	Significant correlation (p = 0.02) between incorrect arch support and plantar discomfort
				Pain
Spector¹³⁵	Retrospective cohort study	81 female ex-elite athletes (67 middle and long distance runners, and 14 tennis players), aged 40–65, recruited from original playing records, and 977 age matched female controls from London UK	Long term weight bearing sports activity	OA as defined by radiological changes (joint space narrowing and osteophytes) in hip joints, PF joints, and TF joints	Yes	The ex-athletes had greater rates of radiological OA at all sites. This association was strongest for the presence of osteophytes at the TF joints (OR = 3.57 (95% CI, 1.89 to 6.71)), at the PF joints (OR = 3.50 (1.80 to 6.81)), narrowing at the PF joints (OR = 2.97 (1.15 to 7.67)), femoral osteophytes (OR = 2.52 (1.01 to 6.26)), and hip joint narrowing (OR = 1.60 (0.73 to 3.48)), and was weakest for narrowing at the TF joints (OR = 1.17 (0.71 to 1.94)). The tennis players tended to have more osteophytes at the TF joints and hip

Nigg¹³⁶	Prospective cohort study (2 m follow up)	171 members of tennis clubs	Shoe, temperature, type and duration of match play, subjective assessment of shoe comfort, sole grip and lateral stability	Pain	No	Stiffness of shoe and subjective evaluation of frictional properties of the shoe were significantly associated with pain