Can comparison of MZ- and DZ-twin concordance rates be used invariably to estimate heritability?

Proponents of the validity of the classical MZ-DZ twin comparison model for calculating heritability claim that the environments influencing MZ and DZ twin individuals are essentially identical. This 'equal environments assumption' may or may not be universally true when applied to the analysis of subjective traits. We examined the validity of this assumption as applied to the propensity for smoking cigarettes, reasoning that equality of environments should lead to equal smoking prevalences in MZ and DZ twin individuals. We identified 8 twin populations with data on smoking. We compiled odds ratios (ORs) for ever smoking in MZ and DZ twin individuals in these 8 studies and overall, using a fixed-effects meta-analytic method based on the Mantel-Haenszel procedure. The prevalence of smoking was less in MZ twin individuals than in DZ twin individuals in 7 of 8 studies. The overall OR was 0.86 (95% confidence interval 0.84, 0.89). ORs were virtually unchanged when the analyses were stratified for gender and age, and no differences were found in relation to the location of the study, the date of the study or the birth years of the cohorts. For cigarette smoking, the environments of MZ and DZ twins may not be co-equal. For subjective traits, heritability estimates may be influenced by these unequal environmental factors that differentially affect their development and characteristics in MZ and DZ twins.