Smoking as a confounder of the association of suicidality with serum lipid levels

A meta-analysis by Wu and colleagues1 reported an inverse association between serum lipid levels and suicidality. The authors compared total serum cholesterol (TC), high- (HDL-C) and low-density lipoprotein cholesterol (LDL-C) and triglyceride (TG) concentrations between individuals with psychiatric disorders (depression, schizophrenia, personality disorders, drug and alcohol addiction) with or without suicidality (ideation, threatened suicide, suicide attempts, suicide completion) to healthy controls. They found that suicidal persons had significantly lower serum TC, LDL-C and TG than nonsuicidal persons. There was no overall difference between suicidal patients and healthy controls for HDL-C with the exception of individuals with schizophrenia compared with controls.

The authors mentioned that some included studies did not adjust for important confounders, such as alcohol consumption or smoking. Smoking is the most prevalent health disorder worldwide, it is an independent risk factor for all aspects of suicide behaviour (ideation, wish to die, attempt)2 and it is dose-dependently associated with suicide.3 Smoking increases the risk of suicide attempt when controlling for all other potential confounders, among which prior suicide attempt is the strongest predictor of a subsequent suicide attempt.4,5 Moreover, smoking is strongly comorbid with psychiatric disorders,6 particularly schizoaffective disorders, and it can be hypothesized that smoking is an additive risk factor for suicide- related behaviours among the psychiatrically ill.

At the level of serum lipids, the most characteristic association with smoking is the reduced HDL-C7,8 even in healthy youth;9 HDL-C becomes normal7 or increases10 when smokers quit. Thus, adjusting for smoking status could potentially cancel the association between HDL-C and suicidality observed among suicidal patients with schizophrenia, suicidal patients with other disorders and controls. Similarly, because smoking is frequently associated with elevated TG and metabolic syndrome,8 controlling for smoking would reduce the difference between suicidal and nonsuicidal patients or controls. It is therefore of major interest to control for smoking behaviour when the association of suicidality with serum lipids is assessed.