New insights on the association between family history of type 2 diabetes and comorbid diabetes in patients with schizophrenia and other nonaffective psychoses.
Longstanding evidence suggests that schizophrenia is associated with type 2 diabetes mellitus (DM2). The research includes maternal and birth studies that identify common risk factors for both conditions, including nulliparity, gestational diabetes, preeclampsia, birth weight, and season of birth.1 Studies before the advent of antipsychotics showed an increased prevalence of abnormal glucose metabolism in patients with schizophrenia, albeit with methodological limitations.2 There is also evidence for impaired glucose tolerance and insulin resistance in patients with first-episode psychosis who are either antipsychotic-naÃ¯ve or have minimal antipsychotic exposure.3,4
These findings suggest that some of the increased risk of DM2 in schizophrenia may be independent of antipsychotic medications and involve host-agent-environment interactions. However, this hypothesis has largely been overshadowed by known metabolic adverse effects of second-generation antipsychotic medications, which increase the risk of DM2.
Schizophrenia is associated with an increased risk of the metabolic syndrome, a constellation of metabolic risk factors-including abnormal glucose tolerance-associated with cardiovascular disease morbidity and mortality.5 In an earlier study, patients with schizophrenia and other nonaffective psychoses (NAP) and a parental history of DM2 were found to have a 3.7-fold increased odds of comorbid DM2 compared with patients without this parental history.6
A meta-analysis was undertaken to understand the association between a family history of DM2 and DM2 comorbidity in patients with NAP.7 Studies were identified by systematically searching Medline, PsycInfo, Web of Science, and Science Direct. Nonaffective psychosis was defined to include schizophrenia, schizophreniform disorder, brief psychotic disorder, delusional disorder, schizoaffective disorder, and psychotic disorder not otherwise specified. One hundred thirty-six potential studies were identified; however, most were excluded because of missing family history data about DM2, or because the diagnoses were primarily affective psychosis. Case reports were also excluded. After a detailed review, 10 studies met the inclusion criteria.
Data used included number of subjects with and without a family history of DM2 based on comorbid DM2 status, as well as other clinical and demographic variables. Effect size estimates (odds ratios [ORs] and 95% confidence intervals [95% CIs]) were calculated using the random effects method. Given significant between-study heterogeneity, a sensitivity analysis and a series of meta-regressions to explore possible moderating variables we used to account for heterogeneity (eg, age, sex, geographic region, body mass index, year of publication, study quality).
The total sample consisted of 3780 patients with nonaffective psychoses, including 804 subjects with a family history of DM2 and 2976 without a family history of DM2. A family history of DM2 was associated with more than a four-fold increased odds of comorbid DM2 (OR = 4.3, 95% CI 2.9-6.4, P < .001), with significant between-study heterogeneity. In sensitivity analyses, after removing two outlying studies with the highest OR, the association remained significant (OR = 3.3, 95% CI 2.6-4.3, P < .001), but between-study heterogeneity was no longer significant. A funnel plot and Egger’s test were not significant for publication bias, although findings should be interpreted with some caution given the small number of studies. In meta-regression, there was a significant, positive association with age, but all other potential moderating factors were unrelated to the association between comorbid DM2 and family history of DM2.
The findings showed a family history of DM2 is associated with more than a four-fold increased odds (OR=4.3) of comorbid DM2 in patients with nonaffective psychoses. Sex, geography, BMI, study quality, or year of publication did not moderate this association. However, the association was moderated by older age, which is a known risk factor for DM2. The results are broadly consistent with those in general population studies, which have reported a 2- to 4-fold increase risk of DM2 with a parental history of DM2.8 Therefore, the results do not suggest that a family history of DM2 is associated with differentially increased odds of DM2 in patients with schizophrenia compared with individuals in the general population.
Strengths and weaknesses
An important strength of the study was the relatively large cumulative sample size of patients with and without comorbid DM2. A limitation was the heterogeneity in the definition (eg, parents versus any first-degree relative) of what constitutes family history of DM2. For most studies, a family history of DM2 was based on subject self-report. Thus, there is potential for selective recall bias. Data on family psychiatric history were not available for the vast majority of studies, limiting inferences on what proportion of cases of a family history of DM2 might be attributed to psychotropic medication-induced diabetes. However, it is reassuring that the odds of a family history of DM2 in patients with nonaffective psychosis was broadly consistent with findings in the general population. The study did not permit inferences regarding the mechanism(s) of this association, which may be due to shared environmental or genetic risk factors, or gene-environment interactions.
The bottom line
Given the high risk for metabolic syndrome, the association in the study suggests that screening for a family history of DM2 is germane to the clinical care of patients with schizophrenia and other nonaffective psychoses, as it can inform on risk of incident diabetes with antipsychotic treatment. However, it is important to emphasize that patients without a family history of DM2 are also vulnerable to the diabetogenicity of some antipsychotics. Nevertheless, stratifying patients based on DM2 family history status may increase the signal-to-noise ratio of future trials of adjunctive agents for antipsychotic-induced metabolic dysfunction.
This article was originally posted on December 5, 2019, and has since been updated. -Ed.
Dr Miller is Associate Professor of Psychiatry, Department of Psychiatry and Health Behavior, Augusta University, Augusta, Georgia. He is the Schizophrenia SectionChief for Psychiatric Times.
The author reports that he receives research support from Augusta University, the National Institute of Mental Health, the Brain and Behavior Research Foundation, and the Stanley Medical Research Institute.
1. Kandhal P, Miller BJ. Shared early life risk factors for schizophrenia and diabetes. Minerva Psichiatrica. 2013;54:197-210.
2. Kohen D. Diabetes mellitus and schizophrenia: historical perspective. Br J Psychiatry. 2004;184:64-66.
3. Greenhalgh AM, Gonzalez-Blanco, L, Garcia-Rizo C. et al. Meta-analysis of glucose tolerance, insulin, and insulin resistance in antipsychotic-naÃ¯ve patients with nonaffective psychosis. Schizophr Res. 2017;179:57-63.
4. Perry BI, McIntosh G, Weich S, et al. The association between first-episode psychosis and abnormal glycaemic control: systematic review and meta-analysis. Lancet Psychiatry. 2011;11:1049-1058.
5. McEvoy JP, Meyer JM, Goff DC, et al. Prevalence of the metabolic syndrome in patients with schizophrenia: baseline results from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial and comparison with national estimates from NHANES III. Schizophr Res. 2005;80:19-32.
6. Miller BJ, Goldsmith DR, Paletta N, et al. Parental type 2 diabetes in patients with non-affective psychosis. Schizophr Res. 2016;175:223-225.
7. Chung J, Miller BJ. Meta-analysis of comorbid diabetes and family history of diabetes in nonaffective psychosis. Schizophr Res. 2019 (In press).
8. Papazafiropoulou AK, Papanas N, Melidonis A, Maltezos E. Family history of type 2 diabetes: does having a diabetic parent Increase the risk? Curr Diabetes Rev. 2017;13:19-25.