Background/Purpose:   Vitamin K is an essential
cofactor in bone and cartilage mineralization.  Low serum vitamin K has been
associated with increased prevalence of hand and knee osteoarthritis (OA), higher
risk of developing knee OA and cartilage lesions, and, in one study, worsening of
meniscal lesions; however, incidence of meniscal damage was not assessed. 
Meniscal damage is a potent risk factor for knee OA, and severe knee OA is associated
with meniscal damage.  We therefore sought to identify whether vitamin K status
was associated with the prevalence and incidence of meniscal pathology on MRI.

Methods: The Multicenter Osteoarthritis (MOST) Study
is a NIH-funded longitudinal cohort study of adults with or at high risk for
developing knee OA.  Participants had knee x-rays and 1.0T MRIs at 0 and 30
months. MRI evidence of meniscal pathology was evaluated using WORMS. Meniscal
integrity was graded on a 0-4 scale in 3 subregions of the medial and lateral
menisci, respectively, and medial and lateral meniscal extrusion were graded on
a 0-2 scale.  We defined meniscal pathology as any abnormality of meniscal
integrity (score ≥1 in any subregion) or meniscal extrusion (medial or
lateral score ≥1).  Vitamin K status was determined by plasma
phylloquinone levels at baseline, and categorized as insufficient (<1.0nM)
or sufficient (≥1.0nM); there were too few to analyze true deficiency
(<0.5nM).  We excluded baseline coumadin users. We examined the relation of
vitamin K status to knee-specific baseline presence of meniscal pathology in
the whole sample, and to development of meniscal pathology at 30 months among
subjects without meniscal pathology at baseline using binary regression with
robust variance estimation to obtain prevalence ratios (PR) and risk ratios
(RR), respectively. We used GEE to account for correlations between two knees
within a person.  All analyses were adjusted for age, sex, BMI, vitamin D,
smoking status, race, and prior knee injury.  Analyses were also stratified by radiographic
OA status.

Results: 1457 knees from 1048 subjects (mean age 62.0±7.9
years, BMI 29.9±4.9 kg/m², 62.4%
female, 34.0% with insufficient vitamin K levels) were included in the
cross-sectional analysis.   Of these, 535 knees did not have meniscal pathology
at baseline and were used in the longitudinal analysis.  The prevalence of
meniscal pathology was high.  Vitamin K status was not significantly associated
with either the prevalence or incidence of meniscal pathology (Table). 
Effects were similar when stratified by radiographic OA status.

Conclusion: Plasma vitamin K insufficiency was not associated with
prevalence or incidence of meniscal pathology over 30 months. Given the high
prevalence of meniscal pathology at baseline, we may need to examine the
effects of vitamin K on meniscal pathology at an earlier age, or in a cohort
not specifically selected for OA risk to avoid the possibility of depletion of
susceptibles.