Seasonal variation in critical limb ischemia requiring endovascular therapy: an analysis of a multicenter database of Japanese patients with critical limb ischemia undergoing endovascular therapy.

AIM We investigated whether any seasonal variation is observed in the incidence, severity and prognosis of critical limb ischemia (CLI) requiring endovascular therapy. METHODS We analyzed a multicenter database of 1,568 consecutive CLI cases undergoing primary endovascular therapy for infrainguinal lesions between July 2004 and June 2011. The monthly incidence was assessed according to the cumulative number of cases in each month, using a simple moving average. The data were fitted to a nonlinear regression model with a cosine function. The monthly proportion of cases in each Rutherford classification among the overall CLI population was assessed using a multinomial logistic regression model. The monthly risk of major amputation was evaluated using a Cox proportional hazard regression model. RESULTS Significant seasonal variation was observed in the incidence of CLI (p＜0.01). It was higher in the period from winter to spring, with a peak in March, and lower in the period from summer to autumn, with a trough in September; the fold difference between the peak and trough was 2.2. The seasonal variation was more markedly observed in the cases with a more severe Rutherford class. The proportion of cases in each Rutherford class among the overall CLI population also exhibited significant seasonal variation (p＜0.01). In addition, the risk of major amputation demonstrated significant seasonal variation (p=0.03); however, the statistical significance was lost following adjustment for the Rutherford classification (p=0.10). CONCLUSIONS Seasonal variation is observed in the incidence and severity of CLI. The seasonality of the limb prognosis is likely explained by that of the CLI severity.