BMI and depressed mood were found to be positively and independently associated with knee pain and activity limitations. BMI and depressed mood explained small parts of variance in knee pain. BMI explained a substantial part of variance in activity limitations, while depressed mood made a small contribution.
This is the first study that is primarily aimed at examining the interrelations between BMI, depressed mood, knee pain and activity limitations in knee OA. Five earlier studies examined these interrelations in the context of a larger study on determinants of pain and activity limitations [9, 12, 17–19]. The results of two of these five studies were in agreement with our results [12, 17]. Three studies did not find independent associations between BMI, depressed mood, and pain and activity limitations [9, 18, 19]. However, the latter studies found that BMI and helplessness  or anxiety  (psychological concepts closely related to depressed mood) were independently associated with pain or activity limitations. Possibly, although it is important to distinguish depressed mood, helplessness and anxiety, these concepts are not completely separable in empirical models. Therefore, although the results of these two studies [9, 18] differ from our results, they seem not to be in conflict with our findings.
In the present study, BMI and depressed mood were independently associated with knee pain and activity limitations in patients with knee OA. This indicates that in clinical practice BMI and depressed mood should both be monitored and targeted. However, BMI and depressed mood explained only a small part of variance in knee pain, which suggests that treatment of these conditions may result in only modest improvement in pain. On the other hand, BMI explained a substantial part of variance in activity limitations, while depressed mood made a small contribution. This suggests that bodyweight is a relevant treatment target, resulting in improvement in activity limitations. Treatment of depressed mood is expected to result in only modest improvement in activity limitations. These predictions regarding the effects of interventions seem to be supported by the literature. Studies in patients with knee OA and overweight or obesity have shown that weight reduction interventions lead to small to large effect sizes with greater improvements in activity limitations than in pain [33–35]. Little is known about the effect of depression interventions in patients with knee OA. One large study compared the effect of collaborative depression care with that of usual depression care in patients with OA and a major depression or dysthymia, and reported small to moderate effect sizes for pain and activity limitations . The results of secondary analyses in the latter study suggested that depression interventions could be improved by targeting not only depression, but also pain using a combined medication and behavioural approach [2, 36].
Several mechanisms have been proposed to explain the associations between BMI, depressed mood, knee pain and activity limitations in patients with knee OA. Increased mechanical stress [5, 6] may explain the association between BMI and knee pain and activity limitations. Overweight increases the load on the knees during weight bearing activities, which may lead to pain and activity limitations. Fatigue [7, 37] may explain the association between depressed mood and knee pain and activity limitations. Fatigue is associated with depressed mood and pain , and may lead to decreased motor activity resulting in activity limitations . The interrelations between BMI, depressed mood and knee pain may be explained by low-grade inflammation and dysregulation of the hypothalamic-pituitary-adrenal axis (HPA axis). Overweight and depressed mood have been associated with activation of inflammatory pathways, including increases in C-reactive protein [15, 37]. The HPA-axis may be involved in such pathways , and chronic low-level inflammation may lead to pain. The interrelations between BMI, depressed mood and activity limitations may be explained by low self-efficacy (i.e. low confidence in the ability to complete a task or activity) [38, 39]. In both overweight and depressed people low self-efficacy may lead to avoidance of activities [40, 41] and thereby activity limitations . The validity of these proposed mechanisms has not been examined in patients with knee OA and could be a target for further research.
Some methodological issues need to be addressed. First, in our study population the prevalence of depressed mood was rather low: only 11.2% had a HADS depression score > 7 indicating probable depressed mood . Despite this low prevalence we found significant associations between depressed mood, knee pain and activity limitations. However, the low prevalence of depressed mood could have influenced the strength of the associations found. Therefore, more studies are needed to externally validate our findings. Second, because the present study had a cross-sectional design no causal inferences can be made. It is hypothesized that high BMI and depressed mood lead to knee pain and activity limitations, however reverse causation (i.e. knee pain and activity limitations lead to a high BMI and depressed mood) is also possible and cannot be ruled out. The results of the present study may guide further research aimed at unravelling the causal pathways and mechanisms underlying the interrelations between BMI, depressed mood, knee pain and activity limitations. For this purpose longitudinal studies are needed. Third, our suggestions about the possible effects of interventions targeting depressed mood and especially BMI are hypothetical and cannot be confirmed based on the results of the present study.