SETTING AND OBJECTIVE : In low-incidence countries, most tuberculosis (TB) cases are foreign-born. We explored the temporal relationship between immigration and TB in first-generation immigrants between 1995 and 2012 to assess whether immigration can be a predictor for TB in immigrants from high-incidence countries. DESIGN: We obtained monthly data on immigrant TB cases and immigration for the three countries of origin most frequently represented among TB cases in the Netherlands: Morocco, Somalia and Turkey. The bestfit seasonal autoregressive integrated moving average (SARIMA) model to the immigration time-series was used to prewhiten the TB time series. The crosscorrelation function (CCF) was then computed on the residual time series to detect time lags between immigration and TB rates. RESULTS: We identified a 17-month lag between Somali immigration and Somali immigrant TB cases, but no time lag for immigrants from Morocco and Turkey. CONCLUSION: The absence of a lag in the Moroccan and Turkish population may be attributed to the relatively low TB prevalence in the countries of origin and an increased likelihood of reactivation TB in an ageing immigrant population. Understanding the time lag between Somali immigration and TB disease would benefit from a closer epidemiological analysis of cohorts of Somali cases diagnosed within the first years after entry. © 2017 The Union.

TB Asylum seekers Cross-correlation Prewhitening CCF

Netherlands Somali (people) human immigration statistics and numerical data Turkey (republic) time factor Time Factors ethnology Humans migrant lung tuberculosis Emigrants and Immigrants tuberculosis population research prevalence Incidence Turk (people) Article time series analysis Retrospective Studies seasonal variation major clinical study migration Somalia Turkey cohort analysis Emigration and Immigration retrospective study Moroccan Morocco