Tuesday, June 15, 2010

New Study Concludes that Secondhand Smoke Causes Mental Illness

A new study published online ahead of print in the Archives of General Psychiatry concludes that secondhand smoke exposure is a cause of mental illness, including depression, psychoactive substance use, schizophrenia, delirium, and mental and behavioral disorder (see: Hamer M, Stamatakis E, Batty GD. Objectively assessed secondhand smoke exposure and mental health in adults. Arch Gen Psychiatry 2010).

The study consisted of a cross-sectional and a longitudinal component. In the cross-sectional component, a relationship was found between secondhand smoke exposure and psychological distress. In the prospective component, secondhand smoke exposure at baseline was found to be associated with a higher risk of psychiatric hospital admission at follow-up five years later, while controlling for baseline psychological distress. The sample consisted of 5,560 adult nonsmokers from the Scottish Health Survey.

The study concludes that "secondhand smoke influences mental health" and the authors "suggest a causal role of nicotine exposure in mental health."

Newspaper headlines widely disseminated to the public the conclusion that secondhand smoke causes mental illness. For example, the headline on Medpage Today read: "Secondhand smoke ups risk of mental illness."

Anti-smoking groups have also used the study results to instruct the public that secondhand smoke causes mental illness. For example, ASH Scotland was quoted as stating: "We have known about links between active smoking and mental health problems, but this new research suggests that second-hand smoke is even more harmful than we thought."

The Rest of the Story

The rest of the story is that there is a very plausible alternative explanation for the study findings: instead of secondhand smoke causing mental illness, it is quite likely that people who have psychological distress are more likely to expose themselves to secondhand smoke.

Think about it: where were the places where people were most likely to be exposed to secondhand smoke in 1998 in Scotland? Secondhand smoke in the conventional workplace was already minimal by then, so the most likely sources of exposure were pubs, restaurants, casinos, and in the home.

Are people who frequent pubs and casinos comparable to people who avoid pubs and casinos in terms of their risk of psychological distress? I think not. For example, people often go to pubs or casinos specifically to alleviate psychological distress. In addition, there is no question that people who go often to pubs are more likely to drink alcohol, and probably to use other substances as well.

In my view, the cross-sectional study establishes a correlation between secondhand smoke exposure and psychological distress, but it does not imply a causal connection. It is more likely that the direction of causation is the opposite: people with psychological distress are more likely to put themselves in situations where secondhand smoke exposure is higher.

The longitudinal study is also severely flawed. The study considers as a potential "effect" of secondhand smoke the diagnosis of psychoactive substance abuse. Yet it does not control for baseline levels of substance abuse. How can you possibly conclude that secondhand smoke exposure causes psychoactive substance abuse among a cohort of nonsmokers if you do not even measure whether they used psychoactive substances at baseline?

Other potential "effects" of secondhand smoke included schizophrenia, delirium, and mental and behavioral disorder. There is simply no biologically plausible mechanism to account for a causal relationship between secondhand smoke exposure and any of these disorders among adults.

This brings us to another major flaw of the study: the lack of biologic plausibility for the purported causal effect. A conclusion of causality should never be made based on association alone. One also needs to show biologic plausibility. In this case, there is no biologic plausibility. It has not yet even been shown that smoking causes mental illness. In the studies that have associated smoking with depression, for example, the relationship appears to be a reverse one: people who are depressed are more likely to use cigarettes. In addition, people who are depressed are less likely to quit smoking. As the study itself points out, in a large cohort study which found an association between smoking and suicide, the relationship disappeared after controlling for baseline levels of alcohol consumption and mental well-being.

The study attempts to put forward a potential mechanism for an effect of secondhand smoke on mental illness, but it is extremely weak. What is the best the study can offer?

"Low-grade inflammation."

That's another way of saying: "We have no idea how secondhand smoke could cause mental illness, but we don't want to say that, so instead we're going to invoke the classic 'chronic low-grade inflammation' excuse."

While the study is a valuable one because it demonstrates an association between secondhand smoke and mental illness, it crosses the line from solid science to hysteria when it - and groups like ASH Scotland - conclude that the observed relationship is a causal one and when newspaper headlines scare the public into believing that tobacco smoke exposure is a cause of psychiatric disease without sufficient evidence to draw such a conclusion.

For some reason, when it comes to the harmful effects of secondhand smoke, all scientific bets are off. You don't need to use the same scientific rigor that we use to study other exposures. Especially since the tobacco companies are no longer monitoring non-supported claims like this one, anti-smoking groups and researchers can get away with a lot more than they used to. This deterioration of the rigor of the science does not benefit the tobacco control movement, however. It harms the movement because it risks losing our credibility and scientific reputation with the public.

No comments: