Tobacco smoking, cognition and first episode psychosis: time for a rethink?


Understanding why smoking is more prevalent among people with schizophrenia is particularly important given that much of the premature mortality associated with schizophrenia can be attributed to smoking. Over the last few years I have blogged about how evidence is emerging that smoking may be a risk factor for schizophrenia, but also that nicotine (or nicotinic agonists) could normalise prefrontal cortex function and reduce some of the cognitive deficits found in people with schizophrenia. However, not all studies have found an improvement in cognition with nicotine use.

In this study, the authors argue that these inconsistencies may be explained by differences in study design and phase of illness (e.g., recruiting those with chronic illness, or those diagnosed with affective psychosis). They sought to examine the relationship between chronic exposure to nicotine through tobacco smoking on cognitive performance in a large population of first episode psychosis patients and health volunteers. They hypothesised that if nicotine improves cognition then smokers should obtain better scores than non-smokers.

The authors of this new study hypothesised that if nicotine improves cognition then smokers should obtain better scores than non-smokers.

The authors of this new study hypothesised that if nicotine improves cognition then smokers should obtain better scores than non-smokers.


This is an observational, cross-sectional study. Participants were part of a cohort of first episode non-affective psychosis patients included in the first episode psychosis programme of Cantabria, Northern Spain. Healthy volunteers in the comparison group were recruited from the community through advertisements, and matched for age, sex and years of education.

Cigarette consumption and the age at which regular smoking commenced was obtained via retrospective self-report. Non-daily smokers were classified as smokers, and former smokers as non-smokers. Smokers were minimally nicotine deprived (≤2h since last cigarette) or non-deprived when tested.

Participants completed a battery of cognitive tests, including verbal memory, visual memory, executive functioning, working memory, processing speed, motor dexterity, attention, and a measure of premorbid IQ. The battery took approximately 2 hours to administer, and smokers were allowed to take smoking breaks if requested, to avoid nicotine deprivation effects.


Of the 397 first episode psychosis patients approached, 304 (77%) agreed to take part. There were 156 healthy volunteer participants in the comparison group. The prevalence of smoking in the patient group was 57%, whereas among the healthy volunteers it was 49%. Unsurprisingly, given that the two groups were matched, there was no clear evidence of differences in age, sex, or years of education.

Each cognitive score was standardised to a z-score (with a mean of 0 and a standard deviation of 1) using the score of the healthy volunteer comparison group as a reference. One-way ANCOVA was used to compare the performance of smokers and non-smokers on each of the cognitive domains assessed. Age, sex, years of education and premorbid IQ were included as covariates.

There was no clear evidence of a difference between first episode psychosis patients who were smokers or non-smokers on:

  • Verbal memory (p = 0.41)
  • Visual memory (p = 0.22)
  • Executive functioning (p = 0.20)
  • Working memory (p = 0.78)
  • Motor dexterity (p = 0.34)
  • Attention (p = 0.70).

There was only very weak evidence of a difference on processing speed (p = 0.09).

Similarly, there was little evidence for a difference between healthy volunteer smokers and non-smokers, with only weak evidence for a difference on visual memory (p = 0.03).

This study suggests that chronic exposure to nicotine through cigarette smoking is not associated with cognitive functioning in first-episode psychosis.

This study suggests that chronic exposure to nicotine through cigarette smoking is not associated with cognitive functioning in first-episode psychosis.


The authors conclude that chronic exposure to tobacco is not associated with cognitive performance, either in first episode psychosis patients or healthy volunteers, and that these results therefore do not support the hypothesis that the use of nicotine via tobacco smoking enhances cognition in individuals diagnosed with non-affective psychotic disorder.

One possible explanation for this is that nicotine may only enhance cognition in deprived smokers; in other words, that abstinence is associated with cognitive deficits, and these are reversed after the administration of nicotine.

Disentangling any genuine cognitive enhancing effects of nicotine from withdrawal-reversal effects is notoriously challenging, particularly since the administration of nicotine to nicotine-naïve participants is difficult because first exposure to nicotine is usually highly aversive. The authors are therefore correct to highlight this as a potential source of discrepancies between studies.


The main strength of this study is the very comprehensive assessment of cognition, including a number of different cognitive domains. The sample size is also relatively large, given the nature of the population studies and the length of cognitive test battery administered.

There are also a number of general limitations to a study of this kind – for example, observational, cross-sectional studies of this kind cannot tell us about cause and effect, and are prone to residual confounding (where potential confounders have either been measured imprecisely or not measured at all). Given that the results generally do not indicate any differences, we might be less concerned about these general issues. The question is then whether we can conclude that nicotine does not improve cognition. Certainly the authors are right to highlight withdrawal reversal effects as a complicating factor in this kind of research.

It is also worth noting that the analytical method employed (ANCOVA) is arguably better suited to experimental designs, and in this case linear regression would be more appropriate. More importantly, the authors interpret their results solely on the basis of statistical significance (i.e., p < 0.05), which does not provide any information about the strength of any association. This presentation of the data also makes it more difficult to include them in future meta-analyses.

There are also a number of other issues that complicate interpretation. Principally, the effects of nicotine in nicotine-naïve individuals may be different to the effects in those who have developed tolerance. So nicotine may initially be used for self-medication but after a long period of use its efficacy in this respect may diminish. Most importantly, the study is predicated on the assumption that chronic use of nicotine will lead to sustained improvement in cognitive function. This may not be the case; effects may diminish as tolerance to nicotine develops. It is also possible that those with more pronounced cognitive deficits are more likely to be smokers.

Overall, these results do challenge the widely held belief that nicotine improves cognitive function in patients with schizophrenia. However, against a backdrop of very mixed evidence they cannot be considered definitive.

Do we need to rethink the nicotine self-medication hypothesis?

Do we need to rethink the nicotine self-medication hypothesis?


Primary paper

Hickling LM, Perez-Iglesias R, Ortiz-Garcia de la Foz V, Balanza-Martinez V, McGuire P, Crespo-Facorro B, Ayesa-Arriola R. (2017) Tobacco smoking and its association with cognition in first episode psychosis patients. Schizophrenia Research doi: 10.1016/j.schres.2017.04.018

Other references

Smoking and risk of schizophrenia: new study finds a dose-response relationship

Nicotine reverses hypofrontality in animal models of addiction and schizophrenia

Photo credits

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+