Socioeconomic Variation in Tobacco Smoking Among the Adult Population in Ireland

Abstract

Introduction

Differences in smoking prevalence across socioeconomic groups are a major driver of health inequalities. Although smoking prevalence continues to decline across most developed countries, socioeconomic inequalities in smoking still persist. While Ireland is among a small number of countries with a tobacco-endgame goal set to achieve a smoking prevalence of 5% by 2025, the challenge this presents by socioeconomic status is uncharted.

Aims and Methods

We analyzed how differences in smoking status across various socioeconomic groups have changed over time in the adult population in Ireland. We used cross-sectional smoking data from the national population-based Healthy Ireland Survey for 2015–2022 (n = 52 494). Educational attainment and area-based deprivation were used as socioeconomic indicators. Socioeconomic differences and changes in inequality over time were identified using the relative index of inequality (RII). Multinomial logistic regression was used to analyze the association between socioeconomic status and daily smoking, occasional smoking, former smoking, and never smoking with adjustment for sex, age, and survey year.

Results

We observed the highest daily smoking rates among the least educated (OR = 11.62; 95% CI = 9.91, 13.63) and individuals living in the most deprived areas (OR = 4.23; 95% CI = 3.55, 5.04). Additionally, we identified significant relative smoking inequalities over the observation period continued to increase, among the least educated (RII = 2.86, 95%CI = 2.63, 3.09) and individuals living in the most deprived areas (RII = 2.64, 95% CI = 2.36, 2.93)

Conclusions

Despite generally reducing smoking prevalence, socioeconomic inequalities continue to widen among the smoking population in Ireland.

Implications

As the tobacco endgame deadline of 2025 is fast approaching, this study highlights the urgent need to consider potential effects across the lowest socioeconomic status groups when implementing equity-oriented tobacco control policies.

Introduction

Tobacco smoking is the leading cause of mortality and morbidity worldwide.¹ Although smoking prevalence continues to decline across most high-income countries, differences in smoking across socioeconomic status (SES), indexed with educational attainment, employment, and other measures of deprivation, are persistent and a major source of health inequalities.^2–6 Despite increasing implementation of effective tobacco control measures, populations in lower SES groups are more likely to smoke than those in higher SES.^4,6,7 Similarly, smoking is more prevalent among the more deprived^8,9 and among the least educated populations.^2,6,10–14 Significant differences in smoking prevalence between males and females in lower SES groups also continue to be reported, with prevalence significantly higher among males.^13,15–17

More recently, socioeconomic inequalities in smoking prevalence have been measured using indices such as the relative index of inequality (RII).^16,18–24 It allows policy makers to monitor inequality changes over time, by considering the distributional changes across socioeconomic group hierarchies.^18,21 For instance, such index measures have been used to identify increasing smoking inequalities over time among the less educated (compared to the highly educated),^17,20 among the unemployed (relative to the employed),²⁵ and among the more deprived (relative to the better-off) populations.¹⁷ To date it is clear that the declining smoking prevalence among the higher socioeconomic groups continues to increase the inequality in smoking prevalence among the lower socioeconomic groups.²⁶

While Ireland is recognized as a leader in tobacco control,²⁷ smoking-related harm remains the leading preventable cause of ill health and premature mortality, accounting for more than 5000 deaths annually, over €1 billion lost in productivity, and a cost to the hospital system of €460 million each year.^28,29 Ireland is one of the first countries internationally to set a national tobacco endgame goal of achieving a “Tobacco Free Ireland” with the smoking prevalence of less than 5% by 2025.³⁰ Tobacco endgame emerged as a concept in tobacco control over a decade ago. Rather than simply limiting the harm caused by smoking, the tobacco endgame is characterized by the declaration of a date-bound endpoint which is achieved through policies that are designed to “change permanently the structural, political and social dynamics that sustain the tobacco epidemic, in order to end it within a specific time.”³¹ Proposals to achieve a tobacco endgame have identified a range of bold and innovative policies focused across the product, the user, the market, and the industry. Despite exceptionally high levels of public support for this approach in Ireland,³² the goal is unlikely to be achieved by 2025 since the prevalence of smoking among adults in 2023 was reported to be at 18%, with no change since 2021.³³ Tackling this challenge requires a better understanding of smoking prevalence across population groups, along with targeted intervention. However, evidence on smoking prevalence differences across various SES groups in Ireland is relatively uncharted despite this potentially being a key consideration to tobacco endgame achievement. Additionally, to the best of our knowledge, smoking prevalence inequalities in Ireland have not been previously quantified using index measures such as RII. The aim of this study is therefore to examine to what extent smoking status differed across various SES groups, defined by education and area-based deprivation, among the adult population in Ireland. Additionally, we explore the presence of potential inequalities across these SES groups over time.

Methods

Data

Population-based cross-sectional data from the Healthy Ireland surveys over the period 2015–2019, and 2021–2022 for all adults aged 18 years and older was analyzed. The Healthy Ireland survey is conducted each year for a nationally representative sample of approximately 7500 individuals aged 15 and older, and living in Ireland.³⁴ Each year, the sample is identified using probability sampling of all electoral divisions around the country, with a random sample of addresses identified from each region using the list of all addresses in the state.^35,36 The survey collects individual data on health and various health behaviors including smoking status along with various socioeconomic indicators.^34,36 The Healthy Ireland survey is used to capture the health of the Irish population and to monitor progress against key national policy targets.³⁴ From 2015 to 2019, the survey was conducted by personal interview. However, due to the COVID-19 pandemic, the survey was not carried out in 2020. From 2021 the survey has been conducted by telephone using random digit dialing.³⁴

This study is reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.

Smoking Status

We defined the smoking status of individuals into four categories: daily smoking, occasional smoking, former smoking, and never smoking.^15,17 Daily smoking was defined as current daily tobacco product users. Occasional smoking was defined as current tobacco users, who reported using tobacco less than daily. Former smoking was defined as individuals who reported to have used tobacco products in the past (including those who stopped using tobacco products in the 12 months prior to the survey) but who do not smoke anymore. Never smoking was defined as individuals who reported no current or previous use of tobacco products. Smoking status was limited to the use of smoked tobacco products (manufactured cigarettes, hand-rolled tobacco, tobacco pipe, and tobacco cigars) and did not include the use of e-cigarettes and smokeless tobacco products. Individuals who reported to be both tobacco and e-cigarette users simultaneously were categorized as “daily smoking” or “occasional smoking.”

Socioeconomic Status

We defined socioeconomic status using two measures: educational attainment and area-based deprivation. We categorized education into seven categories capturing the highest educational attainment from primary to postgraduate level: primary (early childhood combined with primary), lower secondary, upper secondary, postsecondary, short tertiary, bachelors, and postgraduate (Masters and Doctoral combined).

To capture area-based deprivation, we used the Healthy Ireland survey categorization of ten levels, ranging from 1 (most deprived) to 10 (least deprived).³⁵ This deprivation measure is known as the Haase Pratschke (HP) index,³⁷ which is Ireland’s primary social gradient tool, used by various state agencies and governmental departments, to identify geographic disadvantage and to target communities in greatest need of services and resources.³⁸ The index is constructed based on the demographic profile, social class and labor market affiliation of the local population in which the respondent was resident in the year of survey.^35,37,38 The index is constructed by using Irish population census data to measure relative deprivation across 18 919 small areas each containing approximately 100 households, from the period 2016 to 2019.^37,38 The indices are divided into deciles, each containing 10% of areas based on their deprivation level (each decile is therefore not necessarily representative of 10% of the population).^34–36 Each decile allows us to compare and identify patterns in the geographic location of disadvantage. We used this categorization in our analysis. During 2015–2022 the classification system capturing deprivation differed in 2015 and was only partly captured in 2021 and not captured in 2022.³⁴ To maintain consistency in measurement, we included in our analysis the index measured over the survey period 2016–2019 only.

Analysis

Multinomial regressions were performed to estimate the adjusted odds ratios (ORs) of smoking status across the period 2015–2022 combined. The dependent variable was smoking status (daily smoking, occasional smoking, former smoking, and never smoking) and the independent variables were each of the socioeconomic measures (education and area-based deprivation) in separate models. Smoking status variation was estimated independently for each SES indicator, using never smoking as the reference category. In our estimation, we used the postgraduate level of education and the least deprived decile as reference groups. Additionally, in each estimation model, we included an interaction between sex and age (5-year age bands), to adjust for potential differences in smoking behavior of individuals. Similarly, to capture potential differences across each survey year and to improve the accuracy of estimates, we adjusted for each survey year. All analyses applied the population sampling weights provided by the Healthy Ireland Survey. We illustrated the variation in smoking status for each SES indicator using marginal probability plots, adjusted for sex, age, and survey year. Confidence intervals were calculated at the 95% level. All estimations were performed using Stata v.18.0 software.³⁹

Relative Index of Inequality

In addition to measuring the variation in smoking status across each SES group using adjusted Odds Ratios for each SES indicator, we measured socioeconomic inequalities among daily smoking using the relative index of inequality (RII) over the period 2015–2022. The RII measures the proportionate increase or decrease in inequality between individuals in the lowest socioeconomic group relative to the highest socioeconomic group (ie, relative inequality), by fitting a model to the complete data.^21,40 This measure can be interpreted as the prevalence of being a daily smoker at the bottom of the educational hierarchy compared with the prevalence at the top of the educational hierarchy.⁴¹ A similar interpretation for area-based deprivation applies. An RII value of one indicates no inequality, and higher values (above one) indicate increasing inequality (for the lowest SES group).^40,41

For each SES indicator, we estimated the RII using separate binary logistic regression models (using the “siilogit” Stata command⁴²), for each individual survey year, and for the entire period (2015–2019 & 2021–2022). First, we ranked the weighted sample of all individuals by SES category from the most disadvantaged group, for example, most deprived to the most advantaged group, for example, least deprived.⁴³ Using this approach, we then estimated the RII among daily smoking individuals only, relative to never-smoking individuals, with other smoking categories omitted. Daily smoking was chosen since the proportion of daily smoking individuals in our sample was greater than occasional smoking. We adjusted for differences in the age profile of daily smokers across each estimation period.

Finally, we tested for trends in smoking status over the survey period, for each SES indicator, by conducting a temporal trend test.⁴⁴ We estimated a multinomial logistic regression, with smoking status as the dependent variable, each SES indicator as a categorical variable, and included adjustments for the interaction of sex by age and included survey year as a continuous variable.⁴⁵ We additionally applied the population survey weights as reported in the Healthy Ireland survey. Statistically significant trends were identified based on the significance of the survey year coefficient. This test allowed us to identify whether the changes in smoking status and smoking inequalities across each SES indicator were statistically significant over time.

Results

Descriptive Statistics

Smoking status for 52 494 individuals over the period 2015–2022 is presented in Table 1 by sociodemographic characteristics. Overall, 15.7% and 2.6% of respondents used tobacco products daily and occasionally, respectively. 28.7% were former tobacco users and 53% reported to be never smoking. Daily smoking prevalence among males was 17.5% and 14.1% among females and was greatest (23.1%) for individuals aged 25–29 years. A greater proportion of males (26.4%) and those aged 65 years or older (38.3%) were reported to be in the former smoking category. Never smoking was greatest among females (57.2%), those aged 18–24 (63.6%), individuals with postgraduate education (64.4%), and in the least deprived deprivation decile (56.7%). Daily smoking was greatest among individuals with lower secondary level education, 21.3%, and occasional smoking was greatest among those with a Bachelor’s level of education, 3.3%. The most deprived decile had the greatest proportion of daily smoking individuals (28.6%), compared to 10.2% in the least deprived decile.

Table 1.

Summary of Smoking Status by Sociodemographic Characteristics (2015–2022)

Smoking status	Daily smoking	Occasional smoking	Former smoking	Never smoking	Total
n (%)	8235 (15.7%)	1390 (2.6%)	15 055 (28.7%)	27 814 (53.0%)	52 494 (100%)
Sex, n (%)
Male	4336 (17.5%)	740 (3.0%)	7739 (31.2%)	11 955 (48.3%)	24 770
Female	3895 (14.1%)	650 (2.3%)	7316 (26.4%)	15 856 (57.2%)	27 717
Age group, n (%)
18–24	651 (19.1%)	182 (5.3%)	411 (12.0%)	2171 (63.6%)	3415
25–29	709 (23.1%)	165 (5.4%)	559 (18.2%)	1632 (53.2%)	3065
30–34	887 (21.3%)	210 (5.0%)	883 (21.2%)	2182 (52.4%)	4162
35–39	957 (18.5%)	172 (3.3%)	1335 (25.8%)	2703 (52.3%)	5167
40–44	814 (16.3%)	157 (3.1%)	1445 (28.9%)	2580 (51.6%)	4996
45–49	825 (17.7%)	118 (2.5%)	1316 (28.2%)	2405 (51.6%)	4664
50–54	787 (17.8%)	87 (2.0%)	1282 (29.0%)	2260 (51.2%)	4416
55–59	681 (16.1%)	75 (1.8%)	1290 (30.5%)	2188 (51.7%)	4234
60–64	623 (14.3%)	75 (1.7%)	1500 (34.4%)	2162 (49.6%)	4360
65+	1267 (9.7%)	134 (1.0%)	5002 (38.3%)	6662 (51.0%)	13 065
Educational attainment, n (%)
Primary	900 (18.4%)	77 (1.6%)	1805 (37.0%)	2101 (43.0%)	4883
Lower secondary	1667 (21.3%)	148 (1.9%)	2276 (29.1%)	3741 (47.8%)	7832
Upper secondary	2859 (19.2%)	408 (2.7%)	4037 (27.2%)	7550 (50.8%)	14854
Post secondary	717 (19.0%)	109 (2.9%)	1114 (29.6%)	1824 (48.5%)	3764
Short tertiary	713 (16.0%)	124 (2.8%)	1257 (28.2%)	2365 (53.0%)	4459
Bachelors	1099 (9.4%)	386 (3.3%)	3208 (27.4%)	7025 (60.0%)	11718
Postgraduate	265 (5.4%)	136 (2.8%)	1336 (27.4%)	3141 (64.4%)	4878
Deprivation decile, n (%)
1—most deprived	905 (28.6%)	87 (2.7%)	832 (26.3%)	1340 (42.4%)	3164
2	644 (20.7%)	71 (2.3%)	904 (29.1%)	1492 (48.0%)	3111
3	582 (16.7%)	72 (2.1%)	995 (28.5%)	1844 (52.8%)	3493
4	523 (16.4%)	84 (2.6%)	894 (28.1%)	1683 (52.9%)	3184
5	490 (15.0%)	69 (2.1%)	1006 (30.8%)	1703 (52.1%)	3268
6	449 (14.1%)	77 (2.4%)	939 (29.5%)	1717 (54.0%)	3182
7	385 (12.3%)	89 (2.8%)	987 (31.5%)	1674 (53.4%)	3135
8	375 (12.9%)	84 (2.9%)	894 (30.8%)	1547 (53.3%)	2900
9	310 (12.2%)	77 (3.0%)	752 (29.5%)	1410 (55.3%)	2549
10—least deprived	204 (10.2%)	69 (3.5%)	590 (29.6%)	1130 (56.7%)	1993

Are-based deprivation data was available for years 2016–2019 only (n = 29 984).

Multinomial Logistic Regression Estimates

We observe significant variation in smoking status across each SES indicator as indicated by independent logistic regression estimates (Table 2.). A clear pattern of smoking across the education categories is observed, with daily smoking significantly more likely among the least educated, and less likely among the most educated individuals (Figure 1). Individuals with a primary level of education (compared to a postgraduate level) had over 11 times higher odds of using tobacco products daily (OR = 11.62; 95% CI = 9.91, 13.63). Similarly, those with primary level education had twice the odds of using tobacco products occasionally (OR = 2.08; 95% CI = 1.52,2.85), compared to individuals with postgraduate level education. However, the pattern for former smoking status across each education category, despite being statistically significant, is less clear. Former smoking is most likely for individuals with primary level education (OR = 1.61; 95% CI = 1.45, 1.77) and those with a Bachelor’s level of education had the second lowest odds of former smoking (OR = 1.11; 95% CI = 1.03, 1.20).

Table 2.

Multinomial Logistic Regression Estimation Results by SES Indicator (2015–2022)

Smoking status	Daily smoking	Occasional smoking	Former smoking
Educational attainment
Primary	11.62 [9.91, 13.63]	2.08 [1.52, 2.85]	1.61 [1.45, 1.77]
Lower secondary	9.29 [8.05, 10.73]	1.54 [1.19, 1.99]	1.48 [1.36, 1.62]
Upper secondary	5.53 [4.84, 6.33]	1.42 [1.16, 1.74]	1.27 [1.18, 1.37]
Post secondary	4.96 [4.25, 5.78]	1.41 [1.09, 1.83]	1.41 [1.27, 1.55]
Short tertiary	3.91 [3.36, 4.55]	1.29 [0.99, 1.65]	1.29 [1.18, 1.42]
Bachelors	1.96 [1.69, 2.25]	1.30 [1.06, 1.59]	1.11 [1.03, 1.20]
Postgraduate	1	1	1
N	51 433
Deprivation decile
1—most deprived	4.23 [3.55, 5.04]	1.34 [0.96, 1.87]	1.11 [0.97, 1.28]
2	2.68 [2.24, 3.20]	0.97 [0.69, 1.38]	1.03 [0.90, 1.18]
3	1.94 [1.62, 2.32]	0.81 [0.57, 1.14]	0.92 [0.81, 1.05]
4	1.88 [1.57, 2.26]	0.99 [0.71, 1.38]	0.93 [0.81, 1.06]
5	1.77 [1.47, 2.13]	0.81 [0.57, 1.16]	1.02 [0.90, 1.16]
6	1.57 [1.31, 1.89]	0.91 [0.64, 1.27]	0.95 [0.84, 1.09]
7	1.38 [1.14, 1.66]	1.06 [0.76, 1.48]	1.03 [0.91, 1.18]
8	1.41 [1.20, 1.67]	1.02 [0.73, 1.43]	1.02 [0.90, 1.17]
9	1.24 [1.02, 1.51]	0.97 [0.69, 1.36]	0.98 [0.85, 1.12]
10—least deprived	1	1	1
N^	29 319

“Never smoking” category is the reference category. Reported are adjusted Odds ratios (OR), adjusted for sex × age (interaction), survey year and 95% Confidence Intervals in parentheses from independent multinomial logistic regressions for each SES indicator. ^Analysis of deprivation decile was limited to survey years 2016–2019 (n = 29.984). OR = 1 is the reference category.

Figure 1.

Adjusted marginal probability plots of smoking status by education level (2015–2022) and area-based deprivation (2016–2019).

Across deprivation deciles, we observe statistically significant results for daily smoking only (Table 2.). A steady increase in daily smoking across the most deprived deciles was observed (Figure 1). The odds of the most deprived individuals (decile one) were four times higher for daily smoking (OR = 4.23; 95% CI = 3.55, 5.04), compared to the least deprived (decile ten). A similar pattern for occasional smoking is not observed.

Relative Inequalities

Smoking prevalence changes across each SES indicator suggested a clear decline in smoking prevalence for the population over time (Appendix I). Smoking prevalence decreased from 36% (2015) to 31% (2022) and 15% (2015) to 10% (2022) for individuals with primary and postgraduate education, respectively (Figure A1, Appendix I). Similarly, smoking prevalence decreased from 44% (2016) to 38% (2019) and 23% (2016) to 15% (2019) among the most and least deprived individuals, respectively (Figure A2, Appendix I). However, the relative inequality index (RII) provides a different view (Table A1, Appendix II). Significant variation was observed, where relative inequalities were more persistent among the least educated daily tobacco product users across each survey year (Figure 2). Additionally, RII over the entire survey period, suggests that relative inequality among the least educated continued to widen (RII = 2.86, 95% CI = 2.63,3.09; p-value < .01).

Figure 2.

Age-adjusted relative index of inequality by education level (2015–2022) and area-based deprivation (2016–2019) for daily smoking versus never smoking.

Similarly, relative inequalities were more persistent among the least deprived (Table A1, Appendix II). However, we observe a slight reduction in relative inequality for daily smoking among the most deprived (Figure 2). Despite this reduction, daily smoking prevalence among the most deprived continued to persist over the period 2016–2019 (RII = 2.64, 95% CI = 2.36, 2.93; p-value < .01).

Finally, our test of trends over the survey period shows that the variation in smoking status (and smoking inequalities) across each SES indicator was largely statistically significant (except for the former smoking group by deprivation), represented by the statistically significant Odds Ratios (Table A2, Appendix III). This indicates that variation in smoking increased across different population social groups in Ireland.

Discussion

This study was the first to explore socioeconomic variation among adult smoking individuals in Ireland based on education and area-based deprivation over time. We found significant socioeconomic patterning of daily smoking by education and area-based deprivation in recent years in Ireland, and socioeconomic patterning of occasional and former smoking by education only. Additionally, we identified that significant relative smoking inequalities over the observation period continued to increase, among the least educated and individuals living in the most deprived areas, despite the generally reducing smoking prevalence in Ireland. These findings were reinforced by statistically significant test of trends over time.

Our results are novel from an Irish perspective, and contribute to existing evidence highlighting significant smoking variation across various socioeconomic groups. A study examining such differences across EU countries found that the magnitude of smoking status differences was significantly higher for lowest socioeconomic status population groups.⁴⁶ Similarly, lower educational attainment was associated with significantly higher smoking prevalence.^5,11,14,18 Additionally, in line with our findings across deprivation categories, smoking prevalence was significantly greater among the most deprived populations,¹⁷ and four times higher in the United Kingdom.⁴⁷ Similar findings were reported in Hungary⁵ where educational attainment (compared to income and employment status) provided larger differences in smoking prevalence. This could be explained by the nature of each SES indicator.

Our examination of relative smoking inequalities, that is, RII over time suggest statistically significant results for the daily smoking individuals. Our findings suggest that higher daily smoking prevalence continues to persist among the least educated, and continued to increase, relative to most educated. Similarly, we observe significant relative inequalities among the most deprived daily smoking individuals, relative to the least deprived. However, our findings reporting the presence of significant relative inequalities in smoking over time mirror those of many other studies. Increasing smoking inequalities among the least educated (RII = 0.84¹⁷; RII = 1.74²⁰; RII = 3.5¹⁸) and among the most deprived (RII = 1.15) populations¹⁷ have also been reported. This highlights that many countries continue facing similar challenges, as a result of declining smoking prevalence among the higher socioeconomic groups, and increasing prevalence among lower socioeconomic groups.²⁶

Quantifying inequalities using index measures such as RII provides additional insights into the inequalities across SES groups. RII allows for comparison across outcomes at different scales, that is, SES groups.⁴⁰ However, such measures are defined based on the available data. Thus, in some instances, the precision and interpretation of inequalities may not always be completely accurate. As a result, this may overshadow the overall importance and the resulting implications of how inequality is interpreted and the necessary steps required to tackle such. Additionally, inequality measures are frequently focused on identifying the presence of inequalities among lower SES groups.^16,25 Conversely, in our analysis, despite the overall decrease in smoking prevalence over time, our results highlight the importance of maintaining the continued efforts in targeting smoking populations across the highest smoking groups—and this would remain true even if the RII showed a decrease in inequality.

Despite having incorporated age adjustments into the RII estimation, there may be other explanations for the relative inequality differences in daily smoking between education and deprivation. For instance, education captured in the survey is based on each individual response, making it a more accurate measure of socioeconomic status. In contrast, the HP area-based deprivation index is captured at a small area level, prior to categorizing it into deciles.^37,38 It is possible that during the categorization process, the aggregation of the small-area indices, may have led to slightly reduced variation of deprivation across some of the deciles. Therefore, some individuals who may be more/less deprived may have been categorized into deprivation deciles that are not truly reflective of their true deprivation status.

These findings have significance for tobacco policy and practice in Ireland. Ireland was one of the first countries in the world to set a “tobacco endgame goal, and is on track to be the first to fail to meet this target date in 2025. While overall progress to tobacco endgame is slow and off-track, the increasing relative inequalities in smoking delineated in this study underline a real risk that people in more disadvantaged socioeconomic groups will be left behind in a “Tobacco Free Ireland.”²⁹ The current plan to tackle smoking through the health services in Ireland is focused on scaling up and intensifying stop smoking care to better meet the needs of people in more disadvantaged socioeconomic groups. Specifically, stop-smoking services, which are universal services in Ireland, have already been augmented for communities with concentrated socioeconomic disadvantage known as “Sláinetcare Healthy Community Programme Areas,” and implementation of financial incentives are currently being piloted to intensify service effectiveness in these areas.^48–52 This is in line with a proportionate universalism to tackling health inequalities,⁵³ and supported by evidence that targeting and tailoring of stop-smoking care to better meet the needs of people in lower socioeconomic group has a potential pro-equity impact.⁵⁴ While these user-focused approaches are important, there is an urgent need in Ireland to position these tactics in the context of a wider plan which also focuses on tackling the product (eg, regulation of nicotine content in tobacco products) and supply (eg, upward adjustment of age-restrictions and shrinking retail footprint).³¹ This comprehensive and integrated approach to tobacco endgame has been illustrated to lead to dramatic reductions in health inequalities.⁵³

We acknowledge several limitations in this study. The repeated cross-sectional nature of the survey data limited our analysis, as we were unable to follow the same respondents over time. Additionally, the socioeconomic indicators used in our study were limited to the socioeconomic information captured in the survey data. In addition, smoking status categories used in our analysis were based on self-reported data from respondents. Thus, this may over or under-represented the smoking status over the 7-year period. Furthermore, caution should be taken when interpreting and comparing our results for the “former smoking” group. Our definition of “former smoking” was based on self-reported data under the assumption that all individuals had stopped using tobacco products in the 12 months prior to the survey. This definition is different to one often used in other population health surveys and studies, where “former smoking” often refers to individuals who have smoked more than 100 cigarettes in their lifetime.^54–56 As noted above, the surveys were not conducted in 2020 due to the COVID-19 pandemic. Therefore, we are not capturing smoking status and other socioeconomic information across a continued period from 2015 to 2022. Also, the method of measuring and capturing deprivation deciles changed post 2015³⁷ which may have varied slightly relative to the other survey years. Additionally, no deprivation data was available in 2022, which limited our deprivation analyses over a shorter period 2016–2019, relative to education. Finally, we captured inequality.

Conclusion

Socioeconomic disparities among the least educated and most deprived adult smokers in Ireland continue to persist. This highlights the need for finding approaches that will target and support the lower socioeconomic status groups in addressing smoking-related health inequalities. It is important that the potential effects across these groups are considered, and that policy makers address the issue of smoking inequalities, often disregarded in current policies, when developing future tobacco control policies.

Acknowledgments

We thank the Department of Health for providing access to the Healthy Ireland survey data research micro files (RMFs) used to produce the outputs in this study. I. The Healthy Ireland Survey has been funded by the Department of Health. II. The Department of Health takes no responsibility for the views expressed or the outputs generated from the research undertaken on the RMF(s).

Author Contributions

Gintare Valentelyte (Conceptualization [equal], Formal analysis [equal], Software [equal], Writing—original draft [equal]), Aishling Sheridan (Validation [equal], Writing—review & editing [equal]), Paul Kavanagh (Investigation [equal], Resources [equal], Validation [equal], Writing—review & editing [equal]), Frank Doyle (Conceptualization [equal], Investigation [equal], Resources [equal], Supervision [equal], Validation [equal], Writing—review & editing [equal]), and Jan Sorensen (Conceptualization [equal], Investigation [equal], Methodology [equal], Resources [equal], Supervision [equal], Validation [equal], Writing—review & editing [equal])

Declarations of Interests

None declared.

Funding

This research was funded by the HSE Tobacco Free Ireland Programme [HEFISSS2022], as part of a collaboration with the Healthcare Outcome Research Centre, School of Population Health, RCSI University of Medicine and Health Sciences, Ireland. This publication has emanated from research supported in part by a Grant from Science Foundation Ireland under Grant number [22/RP/10091].

Ethics

The Research Ethics Committee at the Royal College of Physicians of Ireland provided ethical approval for using the Healthy Ireland Survey data. All personal data used and collected for the survey is stored by Ipsos in data centers and servers within Ireland, the United Kingdom and the European Economic Area. This is done in compliance with the General Data Protection Regulation (GDPR). Ipsos only retains personal data for as long as is necessary to support the research project and findings.

Data Availability

Data not publicly available.