INTRODUCTION AND AIMS:
The relationship between cannabis use and diabetes is puzzling. Although cannabis users versus non-users should theoretically have a higher likelihood of diabetes, epidemiological studies suggest otherwise. However, previous epidemiological studies have not considered the potential confounding effects of mental health disorders. As such, the relationship between cannabis use and diabetes was examined while accounting for a range of potential confounders, including mental health disorders.
DESIGN AND METHODS:
Data were obtained from the National Epidemiologic Survey on Alcohol and Related Conditions III. Chi-square tests were used to compare socio-demographics, lifestyle behaviours, physical health disorders and mental health disorders between diabetics and non-diabetics. Measures that exhibited statistical significance in these tests were subsequently included in multiple logistic regression analyses to quantify the relationships between lifetime and 12-month cannabis use and diabetes.
Although there was a considerable attenuation in the magnitude of the odds ratios after adjustment for confounders, there was still a decreased likelihood of diabetes for cannabis users versus non-users. The corresponding odds ratios of diabetes were 0.81 (95% confidence interval 0.70, 0.94) and 0.51 (95% confidence interval 0.41, 0.63) for lifetime and 12-month cannabis use, respectively.
DISCUSSION AND CONCLUSIONS:
A decreased likelihood of diabetes for cannabis users versus non-users was indicated after accounting for a range of potential confounders, including mental health disorders. Before the protective effects of cannabis use for diabetes can be suggested, further epidemiological studies are needed that incorporate prospective designs, as well as feature innovative exposure measurements and statistical analyses.