Reachout Orthopedics - Issue 1

disability. However, these studies presented some limitations, most notably for limited adjustment for potential confounders. Finally, due to the different criteria adopted, it was difficult to ascertain whether only symptomatic or radiological OA predicted the onset of CVD. The purpose of this study was thus to determine whether: i) people with OA are at increased risk of incident CVD compared to people without OA; ii) exists any difference among sites usually affected by OA (hand, hip, knee, back/neck, other) in predicting CVD onset; and iii) if any difference between self-reported and clinical/radiological OA exists in predicting CVD onset. Methods Data Source and Subjects All participants in this study were recruited as part of the ongoing, publicly and privately funded, multicenter, and longitudinal Osteoarthritis Initiative (OAI study) (http:// www. oai.ucsf.edu/) . Specific datasets used are those recorded during baseline and screening evaluations (November 2008) (V00) and those evaluating the participants until the last evaluationavailable(96months;V10).Patients with a high risk of knee osteoarthritis were recruited from 4 clinical sites in the United States (Baltimore, Maryland; Pittsburgh, Pennsylvania; Pawtucket, Rhode Island; and Columbus, Ohio) between February 2004 and May 2006 and were eligible if they: 1) had knee osteoarthritis and reported knee pain in a 30 day period in the past 12 mo, or 2) were at high risk of developing knee osteoarthritis (e.g., overweight or obese, knee injury or operation, parents or siblings with total knee replacement, frequent knee- bending activities that increase risk, and hand or hip osteoarthritis). All of the participants provided written informed consent. The OAI study protocol was approved by the institutional review board of the OAI Coordinating Center, University of California at San Francisco. Exposure The diagnosis for OA in our analysis was self- reported for the most common sites usually affected by OA (knee, hip, hand, back/neck, and other joints) asking to participant if one doctor said that he/she suffers from OA during his/her life. A summary variable ascertained as the presence of at least one site affected by self-reported OA was then calculated. Since for knee OA, radiological diagnosis was also ascertained, an additional analysis was undertaken assessing knee OA defined as a combination of the clinical reporting and assessment of pain and stiffness (i.e. pain, aching or stiffness in or around the knee on most days during the last year), and radiographical OA on the baseline fixed flexion radiograph based on the presence of tibiofemoral osteophytes (correspondent to Osteoarthritis Research Society International atlas grades 1-3, clinical center reading). Outcomes The main outcome of interest was the onset of CVD during the follow-up period. As for OA, CVD was recorded through self-reported information. We defined the development of CVD as the presence of heart attack, heart failure, unclog or bypass arteries in legs, and stroke, cerebrovascular accident, blood clot in brain, or transient ischemic attack. The presence of CVD in the OAI was recorded, other than baseline, during the V3 (24 months), V6 (48 months) and V10 (96 months). Covariates A number of variables was identified from the OAI dataset to explore the relationship between OA and incident CVD. These included: (1) physical activity evaluated through the Physical Activity Scale for the Elderly, a validated scale for assessing physical activity level in the elderly [15]. The scale covers 12 different activities, such as walking, sports, and housework, and is scored from 0 to 400 andmore (nomaximum score has been defined); (2) race was defined as “whites” vs. others; (3) smoking habits as “previous/ current” vs. never; (4) educational level was categorized as “degree” vs. others; (5) yearly income as < and missing data vs. > $50,000; (6) co-morbidities assessed through the modified Charlson comorbidity score, with higher scores indicating an increased severity of conditions [16]; and (7) body mass index (BMI) recorded by a trained nurse. Among the several medical conditions assessed through the Charlson comorbidity score, we reported descriptively the prevalence of some common diseases that could influence the association between OA and CVD, namely diabetes and cancer. Hypertension was diagnosed through self-reported information or in case of systolic blood pressure >140 and/or diastolic blood pressure > 90 mmHg. Blood pressure in the OAI study. Statistical Analyses For continuous variables, normal distribu- tions were tested using the Kolmogorov- Smirnov test. The data are shown as means and standard deviations (SD) for quantitative measures, and frequency and percentages for all discrete variables by OA presence at baseline. P-values were calculated for continuous variables using the independent Student T-test and for categorical parameters the chi-square test by OA presence at baseline. Multivariate Cox’s regression models were conducted using as exposure the presence of OA and as outcome incident CVD at follow-up visits. People dead during follow- up period were censored. Time to event was calculated as time to first CVD event. Factors which reached a statistical significance between participants with OA vs. those without or significantly associated with CVD at follow-up (taking a p-value<0.05 as statistically significant) were included. Multi- collinearity among covariates was assessed through variance inflaction factor, taking a cut-off of 2 as reason of exclusion, but no variablewas excluded for this reason.The basic model was not adjusted for any confounders, while the fully adjusted model included baseline values of: age (as continuous); gender; race (whites vs. others); BMI (as continuous); education (degree vs. others); smoking habits (current and previous vs. others); yearly income (categorized as > or < $50,000 and missing data); PASE score (as continuous); Charlson comorbidity index (as continuous); Individuals with OA were almost three times as likely to have heart failure or coronary heart disease compared with matched non-osteoarthritis cohorts. 24 reachOut Orthopedics