10 August 2007. A new Finnish study has identified a possible negative association between schizophrenia and type 1 diabetes. The results suggest that individuals with this type of diabetes are less than half as likely as those without to develop schizophrenia.
A positive link between schizophrenia and type 2 diabetes is well established (see, e.g., Bushe and Holt, 2004), but it remains unclear whether genes involved in schizophrenia, side effects of antipsychotic medications, or some other factors underlie the insulin resistance that develops in many patients with schizophrenia. Type 1 diabetes, on the other hand, is an autoimmune disorder that develops in childhood or early adulthood. Interestingly, several lines of inquiry (see, e.g., SRF related news story) have raised the possibility of a close relationship between autoimmunity and schizophrenia.
Using a register that comprehensively records governmental reimbursements for drugs used to treat chronic diseases, Hannu Juvonen and colleagues from the National Public Health Institute in Helsinki assembled a nationwide cohort of 5,009 people with type 1 diabetes born in Finland between 1950 and 1959; previous studies of this register concluded that it documents more than 95 percent of cases of type 1 diabetes in the country.
The researchers then used this same register, plus two similar registers that record hospital discharges and diagnoses related to disability pensions, to identify Finnish citizens with schizophrenia during a follow-up period of 1961 through 1991. These registers relied on ICD-8 classifications until 1987, when DSM-III-R classifications were instituted. Applying a broad definition of schizophrenia, “including schizoaffective and schizophreniform disorders and also simple schizophrenia and latent schizophrenia in ICD-8, which correspond to schizotypal personality disorder in DSM-III-R,” the Helsinki team identified 10,931 patients born between 1950 and 1959 who were diagnosed with schizophrenia during the follow-up period.
To assess the overall reliability of these recorded schizophrenia diagnoses, the group employed a “best-estimate case-note consensus procedure” with a register-based sample of 902 individuals, in which two psychiatrists independently reviewed all the available case notes from hospitals and mental health centers and arrived at a DSM-IV diagnosis. When their opinions differed, the psychiatrists met and attempted to arrive at a consensus diagnosis. If they continued to disagree, a third psychiatrist was consulted and a consensus diagnosis was agreed upon by all three. Using this technique, the team identified false-positive schizophrenia diagnoses in 123 patients, or 13.6 percent of the sample.
The registers identified 49 patients as comorbid with schizophrenia and type 1 diabetes, but when the researchers applied the consensus procedure for both diabetes and schizophrenia to these cases, they found a much higher false-positive rate for schizophrenia than in their earlier sample: only 24 of the 49 patients had a consensus diagnosis of schizophrenia. (There were no false-positive diagnoses of type 1 diabetes.)
Halving the risk
The authors calculate the incidence of schizophrenia in members of the cohort born between 1950 and 1959 without type 1 diabetes to be 0.56 per 10,000 person-years, but 0.21 per 10,000 person-years for those with type 1 diabetes (p <.001). Even when postulating a 20 percent false-positive rate in schizophrenia diagnoses in the cohort—6.4 percent greater than that the team had established using the consensus diagnosis procedure—the incidence of schizophrenia was more than twice as high in the subpopulation without type 1 diabetes (0.45 per 10,000 person-years vs. 0.21 per 10,000 person-years; p <.001).
Type 1 diabetes and schizophrenia share many features—both are highly heritable, both have been tentatively associated with gestational and childhood infections, and both have been linked to the HLA antigen alleles A24 and DQB1*0602—and the Helsinki team concedes that the negative association between the two illnesses revealed in their study is puzzling. They suggest that “early insults (e.g., prenatal and childhood infections and obstetric complications) evoke different responses among individuals with a genetic predisposition to type 1 diabetes or schizophrenia, leading to the development of type 1 diabetes in the former and schizophrenia in the latter.” The authors also propose that genes, hormone profiles, or insulin treatments may modify the phenotype of schizophrenia in patients with type 1 diabetes; as evidence, they note that in the consensus diagnoses done in their own study, schizoaffective disorder, schizophreniform disorder, and schizotypal personality disorder were disproportionately represented among patients with diabetes.—Peter Farley.
Juvonen H, Reunanen A, Haukka J, Muhonen M, Suvisaari J, Arajarvi R, Partonen
T, Lonnqvist J. Incidence of schizophrenia in a nationwide cohort of patients with type 1
diabetes mellitus. Arch. Gen. Psychiatry. 2007 Aug;64(8):894-9. Abstract