21

dimensional gait analysis. The authors concluded that each pound of weight loss would result in a four-fold reduction in the load exerted across the knee joint per step and, in total, of more than 4800 lbs per mile walked (1200 strides/mile).13 In addition, although its significance is not entirely understood, obese people have gait patterns that differ from those of normally weighted people, such as walking with increased hindfoot motion and forefoot abduction.14 Also, minor trauma in overweight individuals can cause injuries typically seen in high-energy accidents, the most notable being dislocation of the knee.15 Humoral factors. Explanations that solely focus on the mechanical effect of weight fail to explain all aspects of the epidemiology of OA of the knee in the obese. While biomechanical effects certainly play a substantial role, humoral factors may also play a part. Various cytokines and growth factors contribute to the development and repair of degenerative intra-articular changes. OA develops when cartilage breaks down faster than it is produced. Adipose tissue may produce abnormal levels of certain hormones and growth factors that affect cartilage or underlying bone, possibly leading to degeneration. Lloyd et al16 suggested a modest association between elevated levels of insulin-like growth factor-1 (IGF-1) and OA of the distal interphalangeal joints (DIPJ) in the hand and more severe or bilateral knee OA in women, but interestingly failed to find a similar association at the spine and hip. They went on to find that while the subgroup with both DIPJ and knee OA had significantly higher IGF-1 levels than controls without OA, this subgroup also had significantly higher BMI, reducing the significance of the association between IGF-1 levels and OA at these two locations.16 In contrast, Denko, Boja and Moskowitz17 found reduced IGF-1 levels in patients with symptomatic OA. Meanwhile, cross-sectional studies have shown an inverse relationship between IGF-1 levels and adiposity.18 While it has been suggested that higher circulating oestrogen levels predispose obese women to OA, according to Felson et al19 there is little evidence to support this theory. Thus, while numerous papers have sought a discrete alteration in the biochemistry of the obese which predisposes them to OA, the resulting evidence is conflicting and unable to indicate a specific imbalance. Genetic factors. While there is a growing understanding of the genetic basis of OA and a strong association between genetic factors and obesity,20 conflicting views have been expressed about the strength of the genetic association between obesity and OA of the knee. Spector et al21 studied identical and non-identical female twins and estimated that up to 65% of the variance in hand and knee OA in the female population was attributable to genetic factors when controlled for environmental and demographic confounders such as age and weight. Conversely, an earlier twin study in women found a correlation between obesity and OA of the hands and knees, claiming that twins with OA of the knee were 3 kg to 5 kg heavier than their co-twin without disease, appearing to distance the influence of genetics on OA.22 Manek et al23 studied healthy monozygotic and dizygotic female twins to examine the genetic and environmental influences on the association between BMI and OA of the knee, concluding that the association was unlikely to be mediated by shared genetic factors. Their results highlight the uncertain mutual link between genetics, obesity and OA of the knee and should focus attention on the environmental modification of BMI and its effect on OA of the knee, which will be discussed in the remainder of this review. Natural history Most orthopaedic surgeons consider OA of the knee to be a gradual, degenerative process. However, clinical studies of the progression of disease suggest that many patients remain stable for long periods while others progress rapidly.24 Spector et al25 investigated the radiological and symptomatic changes in 169 patients with OA of the knee. A total of 63 patients underwent radiographs of the knee at an interval of 11 years: 42 (67%) did not deteriorate radiologically or symptomatically. An increased risk of progression was associated with pain and OA in the contralateral knee at presentation. Felson et al26 found that varus malalignment in particular was a potent risk factor for progression. Sharma et al27 found that an increased BMI was related to the severity of OA in those with a varus knee but not in those with valgus knees. Using a quantitative assessment of the loss of articular cartilage by MRI in 251 knees, they found that the local factors that independently predicted the loss of tibial and femoral cartilage included damage to the medial meniscus, varus malalignment in the medial compartment, and damage to the lateral meniscus in the lateral compartment.28 These studies did not look at obesity, but may afford some insight into the natural history of degenerative knee pain. OA of the knee in obese patients is typically bilateral and unrelated to other factors correlated with obesity, including OBESITY AND SYMPTOMATIC OSTEOARTHRITIS OF THE KNEE 435 VOL. 94-B, No. 4, APRIL 2012 hyperlipidaemia, hyperuricaemia and diabetes mellitus.29 Dawson et al30 used univariate analysis to study several risk

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25