Marcel Slockers

64 Chapter 4 2006). All analyses were corrected for sex and age. Age was grouped in 10-year age groups. Differences in mortality between the periods 2001–05 vs. 2006–10 were assessed. In sensitivity analyses, we explored whether an alternative distinction in periods (<2003, 2003–08, >2008 and single calendar years) would change the results. Results The presented cohort includes 2130 homeless adults aged ≥18 years in 2001. This cohort consisted of 1870 men and 260 women, with a mean age of 40.3 years (table 1). A total of 17,909 person years were observed, with a mean follow-up of 8.4 years. In total, 265 homeless persons (232 male and 33 female) died during follow-up. Table 1. Characteristics of cohort of homeless men and women in Rotterdam, 2001 Characteristic N = 2130 Age Mean, year 40.3 <20 27 20–29 352 30–39 711 40–49 665 50–59 260 60–69 89 70+ 26 Sex Men 1870 Women 260 Mean duration of follow up 8.4 Type of service Only day care 640 Convalescence care 80 Other (night care, day care + night care, sleeping rough) 1410 Adjusted for age and sex, we found no significant difference in mortality between 2001–05 and 2006–10 ( P = 0.97) (table 2). These results did not change when the interactions age– sex ( P = 0.99) or type of service used (only day care, convalescence care, other) was added