Have there been sustained impacts of the COVID-19 pandemic on trends in smoking prevalence, uptake, quitting, use … – BMC Medicine

There were 102,371 respondents to the Smoking Toolkit Study between June 2017 and August 2022. We excluded 411 people (0.4%) who did not report their smoking status, leaving a sample of 101,960 for analysis. Of these, 55,349 were surveyed before the start of the pandemic (June 2017February 2020) and 46,611 were surveyed during the pandemic (April 2020August 2022). There was a small proportion of missing cases on quitting outcomes (4.1% for quit attempts; 0% for cessation, number of quit attempts, and use of support). Table 1 presents weighted descriptive statistics for the sample as a whole and as a function of the timing of the pandemic (unweighted characteristics are shown in Additional File 5: Table S1).

Table 2 summarises the GAM results. Figure1 shows trends in current smoking over the study period.

Current smoking, overall and by age and social grade. Panels show trends in the prevalence of current smoking among A adults in England (unweighted n: overall=101,960, ABC1=64,088, C2DE=37,872), B 1824-year-olds (unweighted n: overall=12,455, ABC1=7766, C2DE=4689), and C 4565-year-olds (unweighted n: overall=34,332, ABC1=22,401, C2DE=11,931), June 2017 to August 2022. Lines represent modelled weighted prevalence over the study period, adjusted for covariates. Points represent unadjusted weighted prevalence by month. The vertical dashed line indicates the timing of the start of the COVID-19 pandemic in England (March 2020). ABC1, managerial/professional/intermediate; C2DE, small employers/lower supervisory/technical/semi-routine/routine/never workers/long-term unemployed

Overall, among adults in England, the onset of the COVID-19 pandemic was associated with a negligible step-level change in current smoking (Fig.1A). However, there was a notable change in trend. Before the pandemic, smoking prevalence fell by 5.2% per year (relative risk, trend [RRtrend]=0.948; note this percentage represents the relative rather than absolute percentage point reduction, i.e. a 5.2% decrease compared to the previous year [(1-RR)*100], rather than a decrease of 5.2 percentage points within a given year). After the onset of the pandemic, this rate of decline slowed to 0.3% per year (RRtrendRRtrend=0.9481.052=0.997; Fig.1A). The change in trend from pre- to post-onset of the pandemic was significant (relative risk, change in trend [RRtrend]=1.052, 95% confidence interval [CI]=1.014,1.090). In June 2017, smoking prevalence was estimated at 16.2%. At the start of the pandemic (March 2020), it was 15.1%. In August 2022, it was virtually unchanged, at 15.0%.

Stratified analyses showed a 20.1% (95% CI=10.1, 31.0%) step-level increase in smoking prevalence among adults from more advantaged social grades (ABC1) at the start of the pandemic, followed by a slowing in the pre-pandemic decline to the point where progress in reducing smoking reversed (+3.6% per year compared with9.5% per year before the pandemic, RRtrend=1.145, 95% CI=1.083,1.211; Fig.1A). By contrast, there was no increase in smoking prevalence among those from less advantaged social grades (C2DE), and it appeared the modest (~3% per year) pre-pandemic decline continued (Fig.1A).

When we looked at current smoking in different age groups, we saw divergent changes associated with the pandemic: a 34.9% (95% CI=17.7,54.7%) step-level increase among 1824-year-olds (Fig.1B) but a 13.6% (95% CI=4.4, 21.9%) step-level decrease among 4565-year-olds (Fig.1C). While the rise in smoking among young adults was similar across social grades, the fall among middle-aged adults was limited to those from less advantaged social grades (22.4%, 95% CI=10.7,32.6%). As we observed overall, progress in reducing smoking stopped among more advantaged social grades during the pandemic (from12.4% to0.3% per year among 1824-year-olds, RRtrend=1.138, 95% CI=1.004, 1.290; and from11.7% to+3.4% per year among 4565-year-olds, RRtrend=1.171, 95% CI=1.0551.300) but was similar to pre-pandemic rates within less advantaged social grades (Fig.1B and C).

The data indicated these changes were sustained over time (Fig.1), rather than short-lived pulse effects during the early months of the pandemic (Additional File 5: Table S3).

Data on cessation were available for all of the 17,964 past-year smokers in our sample. There were 741 (4.1%) with missing data on quit attempts and, among those eligible, 0 with missing data on the number of quit attempts. Table 2 summarises the GAM results. Figure2 shows trends in quitting activity over the study period.

Quitting activity, overall and by social grade. Panels show trends in the prevalence of A) cessation and B making at least one quit attempt in the past year among past-year smokers (unweighted n: overall=17,964, ABC1=8802, C2DE=9162), and C the weighted geometric mean number of past-year quit attempts among past-year smokers who made at least one quit attempt (unweighted n: overall=5754, ABC1=2908, C2DE=2846), June 2017 to August 2022. Lines represent modelled weighted prevalence (or means) over the study period, adjusted for covariates. Points represent unadjusted weighted prevalence (or means) by month. The vertical dashed line indicates the timing of the start of the COVID-19 pandemic in England (March 2020). Corresponding data without adjustment for dependence are shown in Additional File 5: Fig.1 and Additional File 5: Table 4. ABC1, managerial/professional/intermediate; C2DE, small employers/lower supervisory/technical/semi-routine/routine/never workers/long-term unemployed

Among past-year smokers, the pandemic was associated with a 120.4% (95% CI=79.4170.9%) step-level increase in cessation (Fig.2A). This increase was similar at 154.4% (95% CI=104.8216.1%) when cigarette dependence was not adjusted for (Additional File 5: Table S4; Additional File 5: Fig. S1A) despite mean cigarette dependence only decreasing very slightly during the pandemic (Additional File 5: Table S5; Additional File 5: Fig. S3). There was also a change in trend: the prevalence of cessation was reducing before the pandemic at a rate of 16.1% per year (RRtrend=0.839); this rate of decline slowed during the pandemic (RRtrend=1.219, 95% CI=1.0791.379) to 2.3% (Fig.2A). The change in trend was driven by the less advantaged social grades, among whom the rate of cessation was reversed from24.5% per year before the pandemic to+9.8% per year during the pandemic (RRtrend=1.454, 95% CI=1.2001.762; Fig.2A). By contrast, the more modest (7.4%) pre-pandemic decline in cessation among those from more advantaged social grades appeared to continue (Fig.2A). This pattern of results was largely replicated when we analysed data separately for smokers aged25years (Additional File 5: Table S6; Additional File 5: Fig. S4). However, among the much smaller group aged 1824years, while we observed a significant step-level increase in cessation, there was uncertainty in all the other results with the confidence intervals crossing zero and including the point estimate from the overall analyses for the trend in cessation before the pandemic, the change in trend, and the patterning of the socio-economic results (Additional File 5: Table S6; Additional File 5: Fig. S4).

The pandemic was also associated with a 41.7% (95% CI=29.754.7%) step-level increase in the proportion of past-year smokers who made1 quit attempt (Fig.2B). This increase occurred across ages but was larger among smokers aged 1824 (90.8% [95% CI=57.0131.9%]) than those aged25 (31.5% [95% CI=19.145.2%]) (Additional File 5: Table S6; Additional File 5: Fig. S4). The rate of decline in quit attempts slowed from 8.2 to 1.4% per year (RRtrend=1.074, 95% CI=1.0161.136; Fig.2B); again, this was driven by those from less advantaged social grades, with no significant change in trend among the more advantaged social grades (Fig.2B), and was only observed among those aged25 (Additional File 5: Table S6; Additional File 5: Fig. S4). Among those who tried to quit, there was little change in the mean number of attempts made (Fig.2C).

While analyses of pulse effects showed increases in quitting activity in the first 23months of the pandemic (Additional File 5: Table S3), it is clear from visual inspection of the data in Fig.2 and the change in trend results (Table 2) that these increases were sustained through to August 2022.

Table 2 summarises the GAM results. Figure3 shows trends in use of cessation support over the study period.

Use of support by smokers in quit attempts, overall and by social grade. Panels show trends in the prevalence of use of A prescription medication, B behavioural support, and C e-cigarettes in the most recent quit attempt among past-year smokers who made a least one quit attempt (unweighted n: overall=5754, ABC1=2908, C2DE=2846), June 2017 to August 2022. Lines represent modelled weighted prevalence over the study period, adjusted for covariates. Points represent unadjusted weighted prevalence by month. The vertical dashed line indicates the timing of the start of the COVID-19 pandemic in England (March 2020). Corresponding data without adjustment for dependence are shown in Additional File 5: Fig.2 and Additional File 5: Table 4. ABC1, managerial/professional/intermediate; C2DE, small employers/lower supervisory/technical/semi-routine/routine/never workers/long-term unemployed

Among past-year smokers who made a quit attempt, the onset of the COVID-19 pandemic was associated with little change in the use of prescription medication (Fig.3A). Point estimates for a step-level change were in opposite directions for those from more and less advantaged social grades, but neither group had a statistically significant change. This finding was robust to the exclusion of varenicline from this variable (Additional File 5: Table 7).

However, the pandemic was associated with changes in the use of behavioural support and e-cigarettes for quitting smoking. There was a 133.0% (95% CI=55.3249.6%) step-level increase in use of behavioural support, followed by a continuation of the modest pre-pandemic decline (Fig.3B). By contrast, there was a 21.2% (95% CI=6.833.4%) step-level decrease in use of e-cigarettes (Fig.3C). This change was short-lived (Additional File 5: Table 3) because there was also a change in trend, reversing this step-level decline: before the pandemic, the proportion of smokers using e-cigarettes in a quit attempt fell by 4.1% per year; during the pandemic, it increased by 18.1% per year (RRtrend=1.232, 95% CI=1.1111.365, Fig.3C). These changes were similar across social grades.

Changes in the use of cessation support were similar when cigarette dependence was not adjusted for (Additional File 5: Table 4; Additional File 5: Fig.2).

See the article here:

Have there been sustained impacts of the COVID-19 pandemic on trends in smoking prevalence, uptake, quitting, use ... - BMC Medicine