Schizophrenia as a Complex Trait. Evidence From a Meta-analysis of Twin Studies.
Sullivan PF, Kendler KS and Neale MC;
Commented by , 28 Jan 2004
Aim of the study
It has been well established that schizophrenia is in part a genetic disorder. Recently a number of genes have been discovered that are associated with the disorder. Much evidence of the genetic association has been derived from a number of twin studies. These studies have been reviewed several times, but to date no formal meta-analysis of their findings was available.
The advantage of a meta-analysis compared to a conventional review is that the former can provide quantitative measures of effects. In addition, the pooling of different trials increases the number of cases and thus the statistical power to detect small, but important effects.
Therefore, the authors undertook the first meta-analysis on twin studies to obtain quantitative estimates of heritability in liability but also of shared and individual-specific environmental effects from the published twin data.
Methods
The authors used MEDLINE, dissertation, and books-in-print searches to identify relevant published twin studies of schizophrenia. The fourteen studies identified were published during a period of more than 50 years (1941 – 1999) so that methods and quality varied.
Therefore, only twelve studies that met minimal inclusion criteria of systematic ascertainment were included.A multi-group twin model including parameters for additive genetic, common environmental, and individual-specific environmental components of variance was used in the analysis.
Results
As expected the authors found substantial evidence for additive genetic effects. The point estimate of heritability in liability to schizophrenia was 81% (95% confidence interval, 73%-90%). However, there was also a consistent evidence across studies for common or shared environmental influences on liability to schizophrenia. The joint estimate was lower than that for heritability, but it was statistically significant (11%, 95% confidence interval, 3%-19%).
Discussion
There are two main findings of this meta-analysis. Firstly, the authors confirmed the well-known fact that schizophrenia is to an important extent a heritable disorder. What is new is the quantative estimate that can be derived from this summary of twin studies – 81%.
Although 81% is a high number, it also shows that not the whole effect is due to heritability. What was more surprising is that the authors could secondly show, that environmental effects also have a smaller, but scientifically relevant influence on the liability for schizophrenia.
These results are consistent with the theory that schizophrenia is a complex trait that is caused by genetic and environmental etiological factors. A methodological limitation of the report is that the quality of the individual studies varied. This is understandable given the long period of time during which the studies were carried out.
In addition, the authors found a statistically significant heterogeneity of the results of the individual studies. Reasons for this finding are not clear, but could be related to sample-specific characteristics such as male-female ratios in the individual trials. Although the meta-analysis could not provide information about the specific identiy of the etiological influences, it does provide evidence supporting the search for both underlying genetic and common environmental etiological factors.