Table 1.
Description of studies examining Rhesus D incompatibility as a schizophrenia risk factor
| Study | Birth years (cases) |
Measure of Rhesus D incompatibility |
Rhesus D incompatibility effect |
Sex-specific Rhesus D incompatibility effect |
|---|---|---|---|---|
| Hollister et al. (1996) | 1959–1961 | Rhesus D serotypes from medical records on mother and child |
RR=1.99 (p=0.12) | All males: RR=2.78 (p=0.03) 2nd or later born males: RR=3.32 (p=0.05) All Females: RR=0.53 (p=1.0) 2nd or later born females: not examined due to small sample size |
| Insel et al. (2005) | 1959–1967 | Rhesus D serotypes from blood samples collected from mother and child at delivery |
aRRadj=1.80 (95% CI 0.71–4.58) | 2nd or later born males: RR=2.02 (95% CI 0.71–5.79) RRadj=2.37 (95% CI 0.82–6.86) 2nd or later born females: RR=0.90 (95% CI 0.12–6.74) RRadj=0.93 (95% CI 0.12–7.01) |
| Palmer et al. (2002) | 1940–1969 | Rhesus D genotypes from blood samples collected from mother and child |
RR=2.26 (p=0.027) | Not examined |
| Kraft et al. (2004) b | 1940–1969 | Rhesus D genotypes from blood samples collected from mother and child |
2nd or later born offspring RR=1.7 (p=0.014) |
Not examined |
| Sacker et al. (1995) | 1958 | Mother noted as Rhesus negative in medical record |
OR=1.79 (p=0.05) | Not examined |
| Kendell et al. (2000) | 1971–1974 | Presence of Rhesus D antibodies noted in medical record |
OR=6.0 (p=0.10) | Reported as non-significant |
| Byrne et al. (2000) | Before 1972c | Not explicit | Not reported | Males: 1:255 cases to 0:256 controls (sample too small to compute stable estimates) Females: 2:173 cases to 2:173 controls |
| Hollister and Kohler (2001) | 1955–1967 | Evidence of hemolytic disease of the newborn in medical record |
OR=2.0 (p=0.05) | Not examined |
| Geddes and Lawrie (1995) (meta-analysis) | Before 1976d | Evidence of hemolytic disease of the newborn in medical record or by maternal recall |
OR=1.4 (p=0.49) | Reported as non-significant |
| Cannon et al. (2002) (meta-analysis of Sacker et al., 1995; Kendell et al., 2000; Byrne et al., 2000) | ~1958–1974 | Mother noted as Rhesus negative in medical record; Presence of Rhesus D antibodies noted in medical record; or not explicit |
OR=2.0 (p=0.05) | Not examined |
a
RRadj are adjusted for maternal age and ethnicity.
b
Contains same nuclear families analyzed in Palmer et al., 2002, but expanded to include all genotyped affected and unaffected offspring.
c
Cases most likely born before 1972 because they were diagnosed with schizophrenia between 1972 and 1992.
d
Meta-analysis of studies published between 1990 and 1996, so cases likely born before 1976.