Table 4.

The association between quadriceps weakness and patellofemoral joint (PFJ) cartilage damage and bone marrow lesions, stratified by those with and without patella alta.

	Among those without patella alta				Among those with patella alta
	Tertile 1 (Strong)	Tertile 2	Tertile 3 (Weak)	p trend	Tertile 1 (Strong)	Tertile 2	Tertile 3 (Weak)	p trend	p interaction
# of knees	192	204	174		76	64	97
Lateral PFJ
Cartilage Damage
WORMS ≥2 n/N (%)	101/384 (26)	152/408 (37)	138/348 (40)		58/152 (38)	67/128 (52)	99/194 (51)
Adjusted PD* (%) (95% CI)	0 (reference)	6.6 (−1–14)	7.8 (−1–16)	0.005	0 (reference)	14.9 (1–29)	12.7 (−1–26)	0.08	0.80
Bone Marrow Lesions
WORMS ≥1 n/N (%)	52/384 (14)	84/408 (21)	69/348 (20)		36/152 (24)	44/128 (34)	67/194 (35)
Adjusted PD* (%) (95% CI)	0 (reference)	3.7 (−2–10)	2.7 (−3–9)	0.16	0 (reference)	11.3 (−1–24)	11.4 (0–23)	0.02	0.73
Medial PFJ
Cartilage Damage
WORMS ≥2 n/N (%)	142/384 (37)	205/408 (50)	191/348 (55)		75/152 (49)	75/128 (59)	108/194 (56)
Adjusted PD* (%) (95% CI)	0 (reference)	8.7 (1–16)	12.2 (4–21)	0.0003	0 (reference)	7.8 (−6–21)	0.5 (−13–14)	0.78	0.13
Bone Marrow Lesions
WORMS ≥1 n/N (%)	62/384 (16)	79/408 (19)	76/348 (22)		39/152 (27)	33/128 (26)	28/194 (14)
Adjusted PD* (%) (95% CI)	0 (reference)	0.3 (−5–6)	2.6 (−4–9)	0.63	0 (reference)	−1.5 (−13–10)	−13.9 (−24 – −4)	0.01	0.002**

^*PD=prevalence difference, adjusted for age, sex, and BMI

^**Interaction for medial PFJ BMLs implies that while quadriceps weakness decreased the likelihood of BMLs in those with patella alta it did NOT do so in those without patella alta. This is the opposite direction of the hypothesized interaction (knees with patella alta and weakness would have the greatest damage).