Table 3.
Hazard ratios (95% confidence intervals) for risk of hospital admission for a psychiatric disorder between ages 38 and 78 (1971-2011) according to evacuee status as a child during second world war
| Mental disorder | Evacuee status* | Full sample (n=45 463) | Women (n=22 021) | Men (n=23 442) | |||||
|---|---|---|---|---|---|---|---|---|---|
| Cohort | Within sibling | Cohort | Within sibling | Cohort | Within sibling | ||||
| Any disorder | Evacuee | 0.94 (0.79 to 1.12) | 0.89 (0.64 to 1.26) | 1.16 (0.90 to 1.49) | 1.21 (0.80 to 1.83) | 0.80 (0.63 to 1.02) | 0.67 (0.44 to 1.03) | ||
| Substance misuse | Evacuee | 0.85 (0.65 to 1.11) | 0.58 (0.33 to 1.03) | 0.93 (0.58 to 1.47) | 0.70 (0.33 to 1.49) | 0.82 (0.59 to 1.13) | 0.52 (0.25 to 1.08) | ||
| Psychosis | Evacuee | 0.81 (0.55 to 1.19) | 0.67 (0.35 to 1.29) | 1.00 (0.60 to 1.64) | 1.13 (0.46 to 2.81) | 0.64 (0.35 to 1.17) | 0.44 (0.18 to 1.09) | ||
| Mood | Evacuee | 1.09 (0.83 to 1.43) | 1.39 (0.82 to 2.37) | 1.38 (0.97 to 1.97) | 2.19 (1.10 to 4.33) | 0.83 (0.54 to 1.27) | 0.90 (0.44 to 1.83) | ||
| Anxiety | Evacuee | 1.20 (0.86 to 1.67) | 1.36 (0.73 to 2.54) | 1.37 (0.88 to 2.12) | 1.55 (0.72 to 3.36) | 1.03 (0.63 to 1.70) | 1.14 (0.52 to 2.52) | ||
Effects by sex were derived from one model by including an interaction with evacuee status. Sex composition of analytic sample (n=45 463) reported in table; 1321 sibling pairs were discordant for exposure. Cohort analysis adjusted for sex and its interaction with evacuee status, parental education, native language, number of children in 1940, five categorical variables for socioeconomic status in 1939 and county of residence in 1939, interaction terms between sex and each of five categorical socioeconomic variables, age (birth cohort), birth order, and region of residence (1939) (full sample results in first two columns omit interaction terms with sex). Within sibling analyses used a sibling group specific baseline hazard. All family background covariates—parental education, native language, number of children in 1940, five categorical variables for socioeconomic status in 1939, and county of residence in 1939—cancel out in within sibling analysis.
Cluster robust standard errors are adjusted for familial clustering.
*Reference category is non-evacuated participants.