e-Cigarettes: a double-edged sword?

The use of e-cigarettes has rapidly increased. A UK based randomised trial by Hajek et al (NEJM 2019;380:629) evaluated the effectiveness of e-cigarettes versus nicotine replacement therapy (NRT) as a tool for smoking cessation. Adults attending National Health Service smoking cessation services not currently using NRT or e-cigarettes were eligible. Participants had to agree to not use the non-assigned treatment for at least 4 weeks post quit date. 886 adults already attending stop smoking services were randomised to receive either refillable e-cigarettes (439/886) or a range of NRT (447/886) of their choice. Behavioural support was offered to both groups. NRT was offered for 12 months (~£120 for 3-month supply) or a starter e-cigarette pack with one liquid refill (~£30.25) and encouragement to purchase additional refills themselves. Sustained abstinence was defined as no more than five cigarettes 2 weeks following the quit date and exhaled CO <8 ppm at 52 weeks. Those in the e-cigarette group were statistically more likely to achieve abstinence at each timepoint analysed (4 weeks, 26 weeks and 26–52 weeks) including the primary outcome at 1 year (18.0% vs 9.9%, relative risk (RR) 1.83; 95% CI 1.30 to 2.58). The side effect profile was also investigated. Participants in the NRT arm were more likely to suffer from nausea (37.9% vs 31.3%) and ongoing cough with phlegm production. Those using e-cigarettes reported more frequent throat and mouth irritation (65.3% vs 51.2%). e-Cigarettes were used more frequently and for longer with 40% of patients still using at 1 year compared with 4% in the NRT group. Although e-cigarette use, delivered with behavioural support, improves smoking cessation rates the long-term health risks remain unknown.

The adverse impact of smoking combustible cigarettes on cardiovascular health is well established, however the impact of e-cigarette use among both current cigarette smokers and never smokers is yet to be determined. A cross-sectional analysis was conducted by Osei et al (Am J Med 2019. doi.org/10.1016/j.amjmed.2019.02.016) from 449 092 responders from previous US surveys. Of these 3.5% (15 863) were current e-cigarette users, 13.1% (58 789) current combustible cigarette smokers and 2.9% (12 908) dual e-cigarette and combustible cigarette smokers. 10% (44 582) had self-reported cardiovascular disease, as defined as the presence of a stroke, myocardial infarction or coronary heart disease. Current dual combustible and e-cigarette smokers had significantly higher odds of having cardiovascular disease when compared with current combustible but never e-cigarette smokers (OR 1.36, 95% CI 1.18 to 1.56) and never (combustible or e-cigarette) smokers (OR 2.44, 95% CI 2.14 to 2.78). The odds of established cardiovascular disease increased with level of e-cigarettes exposure (daily or occasional use) in the dual smoker category (OR 1.59, 95% CI 1.20 to 2.08). The authors adjusted for cardiovascular risk factors including age, race, sex, education, body mass index, physical activity, heavy alcohol consumption, income and family history of certain medical conditions. The data represent a large study population but care needs to be taken in inferring a causal relationship as the exposure and outcomes were self-reported and not verified. Furthermore, comprehensive smoking exposure information was not collected; the use of e-cigarettes may be a marker of higher combustible cigarette exposure or there may be a reverse association with e-cigarettes. Future studies are awaited to provide clarity on the long-term safety of this new form of nicotine delivery.

Maternal smoking impacts on respiratory hospitalisations in infants

There is a well-established link between maternal smoking and poor neonatal outcomes. Lawder et al (BMJ Open; 2019;9:e023213. doi:10.1136/bmjopen-2018-023213) examined a wider range of outcomes, including hospitalisation in the first 5 years after birth, than previously published by using a large linked database cohort study of 697 003 singleton births in Scotland, UK. Maternal smoking status was collected at pregnancy booking visits from 1997 to 2009 and infant outcomes followed up until March 2012. At maternal booking visits 23% of mothers self-reported smoking during pregnancy, 9% reported former smoking, 56% never smoking and 12% of mothers had missing data. Adjustments were made for a range of socioeconomic factors. In current smokers both neonatal (OR 1.32, 95% CI 1.17 to 1.49) and postneonatal (OR 2.18, 95% CI 1.87 to 2.53) mortality was elevated compared with babies of non-smokers.

Hospital admissions were increased in infants of smokers for acute upper and lower respiratory tract infections, asthma and bronchiolitis in both the first year and second to fifth years. The highest risk was of asthma related admissions (first year adj HR 2.15, 95% CI 1.80 to 2.57; second to fifth years adj HR 1.29, 95% CI 1.22 to 1.37). The rates of admissions for long bone fractures were similar across groups. This is reassuring as unmeasured confounders have not impacted on the hospitalisation estimates. These data confirm the ongoing significant proportion of mothers smoking during pregnancy as well as the conferred short-term and long-term risks of harm to infants and should be used to support further provision for harm reduction strategies in this group.

Nicotine in schizophrenia: what we do know

People with schizophrenia are four times more likely to smoke than the general population. Those that smoke, do so more heavily with a greater burden of adverse health events. The reason behind this remains unclear. Nicotine may mitigate some of the negative symptoms associated with schizophrenia or patients with schizophrenia may be biologically more susceptible to the reinforcing action of nicotine. Weeks et al (Nicotine & Tobacco Research 2019. doi:10.1093/ntr/ntz048) conducted a laboratory study to test the latter hypothesis. A previously validated tool, Methylazoxymethanol acetate (MAM) rodent model of schizophrenia was used alongside saline infused control rats (CTL). There were four key experiments. The first experiment demonstrated that MAM and CTL animals had a nicotine dose response relationship. Nicotine reinforcement behaviour however, did not differ between groups (p=0.886). The second experiment showed no significant effect in nicotine self-administration between the two groups, across a 23 hours period (p=0.34). The third experiment explored differences between male and female MAM rats of which there was no significant effect (p=0.25). The final experiment used sucrose or light stimuli as a reinforcement factor. It showed that MAM animals responded less than CTL (p<0.05). The authors infer that this indicates anhedonia consistent with a schizophrenia model with some sparing of this effect with nicotine. This study suggests that primary nicotine reinforcement does not account for increased smoking in schizophrenia but that some differential and thus the underlying mechanism remains to be determined.