Serum Leptin and Risk of Knee Pain and
Osteoarthritis: Multicenter Osteoarthritis Study

D. Misra, J. Niu, D.T. Felson, M.
Nevitt, J. Torner, C.E. Lewis, A. Hu, S. Jinno, T. Neogi

Background/Purpose: Obesity is a major risk factor for
knee osteoarthritis (OA). Leptin, a proinflammatory adipose tissue product that
is highly correlated with adiposity, has been associated with risk of knee OA. Given
that leptin is proinflammatory, it could contribute to synovial inflammation.
Since synovial inflammation leads to pain in knee OA, it is particularly
relevant to study the association of leptin with pain and symptomatic knee OA.
Thus, we evaluated the longitudinal association of serum leptin with risk of incident
knee pain, incident radiographic and symptomatic knee OA in community-dwelling older
men and women.

Methods: We included participants from
Multicenter Osteoarthritis (MOST) study, a NIH-funded, longitudinal,
observational study of individuals with or at high risk for knee OA, who had
serum leptin measured at baseline and knee radiographs at baseline and follow
up visits (30-, 60- and 84-month visits). We limited our study sample to those
without radiographic OA (KL grade < 2 and no patellofemoral OA) in either
knee at baseline. Serum leptin was not normally distributed and highly
correlated with body mass index (BMI); thus, we first created BMI residuals of
the log of serum leptin and then created tertiles. Incident radiographic knee
OA was defined by development of Kellgren-Lawrence grade≥2 or
patellofemoral OA by the 84-month visit. Incident symptomatic knee OA was
defined by presence of knee pain (both at the telephone screen and clinic
visit) in addition to incident radiographic OA changes in the knee by the 84-months. 
The association of leptin tertiles with incident knee pain, incident
radiographic and incident symptomatic OA was assessed using logistic regression
with GEE to calculate Odds Ratios (OR), adjusting for age, BMI, race, site,
knee injury, knee surgery and CES-D score. All analyses were sex-specific since
leptin levels and potentially its effects may be sex-specific.

Results: Among 690 subjects (mean age 60.6 yrs,
mean BMI 29.1 kg/m² and 62.5% women), mean serum leptin level at
baseline was 37022.5 pg/ml in women and 15099.1 pg/ml in men. At the 84 month
visit, there were 80, 226 and 64 knees with incident pain, incident radiographic
and incident symptomatic OA, respectively. We found a significant increase in
the risk of incident knee pain in women (p for trend =0.003) but not in men (Table
1). Although the results did not reach statistical significance, there was
a slight increase in the risk of symptomatic OA but not in radiographic OA, in
women and men (Table 1).

Conclusion: Serum leptin is associated with
incident knee pain in women and may be associated with symptomatic OA in men
and women but not with radiographic OA.  Larger studies are needed to confirm
the findings and evaluate leptin as a target for treatment of pain in knee OA.