Abstract
Although Japan experienced a transient increase followed by sudden decrease in the reported number of HIV infections among the non-Japanese in 1991-1992, we still do not have an explanation with persuasive power for the trends of HIV infections among the non-Japanese. Our study investigated ecological associations (associations between aggregated risk factor and outcome) between reported number of HIV infections and estimated number of HIV-infected among the non-Japanese staying in Japan, focusing specifically on those originating from six selected Southeast Asian countries. Through the use of backcalculation method, the crude time-series prevalence of HIV infections in each country of our subject was estimated. Furthermore, we estimated the crude number of HIV-infected among foreign nationals from Southeast Asian countries who were staying in Japan between 1986 and 2001, and performed univariate and multivariate linear regression analyses in order to investigate ecological associations. Both the number of foreign nationals from Southeast Asian countries (adjusted R^2=0.2800) as well as the estimated number of HIV-infected among them (adjusted R^2=0.6007) could roughly predict the reported number of HIV infections in Japan. Our model suggested that the trends in the number of foreign nationals from Southeast Asian countries with the simultaneous background of ongoing HIV epidemic was the major factor that influenced the trends of the reported number of HIV infections. Other possible factors affecting the trends of the reported number of HIV infections among non-Japanese were discussed.
