Abstract
BACKGROUND: Endothelial dysfunction has been identified as an etiologic factor for osteonecrosis of the femoral head (ONFH) and major adverse cardiovascular and cerebrovascular events (defined as major cardiovascular disease [CVD] and cerebrovascular accident [CVA]). However, the incidence of major adverse cardiovascular and cerebrovascular events in patients with nontraumatic ONFH and any association between the two diagnoses remain unclear. QUESTIONS/PURPOSES: We compared a large cohort of patients with nontraumatic ONFH and a matched control group without this diagnosis and (1) examined the frequency and hazard ratio (HR) of major adverse cardiovascular and cerebrovascular events in both groups adjusted for age, sex, socioeconomic status, and associated comorbidities (which we defined as the adjusted HR), (2) determined whether any association of ONFH and major adverse cardiovascular and cerebrovascular events was stable after adjusting for confounding variables, and (3) compared the occurrence of major adverse cardiovascular and cerebrovascular events with time in both groups. METHODS: A population-based cohort with a 14-year dataset period (1997-2010) from the Taiwan National Health Insurance Research Database was used for this retrospective study. The database includes a greater than 99.5% Asian population randomly selected from more than 23 million citizens and foreigners residing in Taiwan for longer than 6 months. A total of 1562 patients with nontraumatic ONFH were identified from a population of one million patients in the database after excluding initially concomitant diagnoses of major CVD and CVA. The comparison group (n = 15,620) without ONFH was analyzed in a one-to-10 ratio by matching the study cohort based on age, sex, income, and urbanization. RESULTS: The patients with ONFH had a higher frequency of major adverse cardiovascular and cerebrovascular events than their counterparts without ONFH (19% versus 14%; p < 0.001). The patients with ONFH had 1.34- and 1.27-fold adjusted HRs for occurrence of major CVD and CVA as compared with the normal population (95% CI, 1.11-1.61, p = 0.002, and 95% CI, 1.09-1.47, p = 0.002, respectively). Sensitivity analysis showed a consistent association between ONFH and major adverse cardiovascular and cerebrovascular events after controlling for potentially relevant confounding variables such as hypertension and diabetes. After adjusting for potential confounders including surgery and medications, ONFH remained independently associated with major CVD (adjusted HR, 1.51, 95% CI 1.09-2.03, p = 0.026) and CVA (adjusted HR, 2.44, 95% CI 1.69-3.51, p < 0.001), apart from age older than 65 years and traditional atherosclerotic risk factors. The cumulative incidence of major adverse cardiovascular and cerebrovascular events also was higher in the ONFH group than the non-ONFH group (30.3% vs 23.1% at the end of followup, p < 0.001). CONCLUSIONS: Patients with ONFH have a strong association with major adverse cardiovascular and cerebrovascular events as compared with the normal population, suggesting a potential common pathway involving endothelial dysfunction. In view of this association in the relatively young population with ONFH, it is important to closely monitor these patients, treat relevant comorbidities early, and investigate preventative measures for these major adverse events. LEVEL OF EVIDENCE: Level III, prognostic study.