Study design and participants
The sample dataset consisted of adults (≥18 years) from the Smoking Toolkit Study (STS) in England between July 2020 to January 2021. This timeframe was selected because July 2020 was the first month in which questions on menthol cigarettes were added to the STS following the TPD ban and January 2021 is the latest month with available data. The STS is a monthly repeated cross-sectional survey of a representative sample of ~1700 adults in England. Further details on the STS, including sampling and weighting techniques are described elsewhere7.
Measures
Current smoking
Respondents were classified as current cigarette smokers (see supplementary materials) if they reported smoking cigarettes (including hand-rolled).
Those who indicate that they do not smoke cigarettes, but do smoke tobacco of some kind were excluded from the analysis because these products were not included in the menthol ban.
Menthol cigarette smoking
Current smokers were asked the following question:
“Cigarettes can be sold in different flavours. They can also be flavoured by capsules, filter tips, cards inserted into a packet or flavoured rolling papers. How would you describe the flavour of the cigarettes you usually smoke?”
- Just tobacco
- Tobacco and menthol
- Tobacco and some other flavour
- Refused
- Don't know
Answers of 2) were categorised as menthol cigarette smoking, responses of 3) were categorised as other flavour smoking. Those responding with 4) and 5) were excluded (2.2%).
Motivation to stop smoking
Motivation to stop smoking was assessed using the Motivation To Stop Scale8, a single-item measure with seven response options representing increasing motivation to quit (see supplementary material). For ease of interpretation, responses were collapsed to reflect high (6–7) vs. low (1–5) motivation to stop smoking.
Cigarette dependence
Cigarette dependence was measured using two questions from the Fagerström Test for Cigarette Dependence: time to first cigarette in the morning after waking and the number of cigarettes smoked per day.
Socio-demographic covariates
The sociodemographic characteristics gender, and occupational social grade and ethnicity (white or non-white) were included (see supplementary materials).
Source of purchase
Data on the source of purchase of menthol cigarettes in the past six months was also collected (see supplementary materials)
Sample selection
Overall, 11,715 (unweighted) adults were surveyed of which 1,557 were smokers who answered the question on cigarette flavours.
Analysis
The overall (weighted) proportion of menthol cigarette use was calculated according to time period (July-October 2020 and November 2020-January 2021), smoking and socio-demographic characteristics. The months were collapsed into two periods to provide a broad analysis of change over time and because source of purchase data reflects self-reported purchase in the past six months, with November onwards therefore representing a period after the ban was implemented. Chi-squared statistics (with Rao & Scott second order corrections) were used to describe the strength of the relationship between menthol cigarette smoking status and the specified variables of interest, and Cramer’s V was used as an effect size.
The sources of purchase of menthol cigarettes in the past six months was summarised for July-October 2020 and November 2021-January 2021 as a prevalence for each source. Chi-squared statistics were used to describe the strength of the relationship between source of purchase and the specified time periods.
Missing data on variables included in each respective chi-squared test were excluded and reported in Table 1 as ‘Missing’. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline were used in the design and reporting of this study9.
Sensitivity analysis
To explore a potential shift towards or away from illicit sources of purchase of menthol cigarettes, a sensitivity analysis compared illicit sources of purchase (in the past six months) in July-October 2020 and November 2020-January 2021. Response variables for illicit sources were collapsed into a single dichotomous variable indicating ‘Any illicit’ or ‘No illicit’ purchase (see supplementary materials). Further sensitivity analyses explored the source of purchase of just tobacco cigarettes across the same time periods.
All planned analyses were pre-registered on the open science framework (https://osf.io/d2nv6/) and carried out in R version 4.0.3.