Perceived Health and Capacity to Cope With Stress in Recent Ex-smokers: Impact of Vaping Versus Quitting All Nicotine

Abstract

Introduction

Little is known about the continued use of nicotine following smoking cessation on perceived well-being in comparison to complete cessation of nicotine use.

Aims and Methods

To explore aspects of perceived well-being and coping among recent ex-smokers as a function of vaping status. Ever-daily smokers in the International Tobacco Control 4 country smoking and vaping surveys in 2016 (w1 N = 883) and 2018 (w2 N = 1088). Cross-sectional associations and longitudinal samples for those who quit between waves and those who quit at w1 and maintained abstinence to w2. Main outcome measures were: Past 30 days of depression symptoms, perceived stress, stress management since quitting, and change in perceived day-to-day health.

Results

In the cross-sectional analyses vapers were more likely to report both improved stress management (aOR = 1.71, 95% CI 1.23–2.36) and perceived day-to-day health (aOR = 1.65, 95% CI 1.26–2.16) than nicotine abstainers. In the longitudinal analyses, smokers who switched to vaping between waves (n = 372) were more likely to report depression symptoms at w2 (aOR = 2.00, 95% CI 1.09–3.65) but reported improved perceived health (aOR = 1.92, 95% CI 1.16–3.20). For the past daily smokers who remained quit between waves (n = 382), vapers were more likely to report improved stress management relative to abstainers (RRR = 5.05. 95% CI 1.19–21.40). There were no other significant differences between vapers and nicotine abstainers.

Conclusions

There is little evidence to support the view that perceptions of well-being deteriorate in vapers compared to complete nicotine abstainers in the immediate years after smoking cessation.

Implications

This study could find no conclusive evidence that the continued use of nicotine via e-cigarettes was detrimental to health compared to completely stopping nicotine intake altogether. Our results would suggest that continuing to use nicotine may even result in some benefits in the short term such as improved stress management, however further longitudinal studies are required to examine if these effects are restricted to the early post-quitting phase and whether other positive or negative effects on psychosocial health emerge in the future.

Introduction

Nicotine vaping products (NVP) have become popular among some smokers and recent ex-smokers, with most users citing smoking cessation and perceived harm reduction as their primary reasons for use.¹ Evidence is emerging that NVP are more effective than conventional nicotine replacement therapy as a smoking cessation aid.^2–5 However, many of those who successfully quit smoking with the aid of vaping continue to vape, and vaping among long-term ex-smokers has increased in some countries over time, eg, England⁶ as a means of relapse prevention and/or for the experienced value of the nicotine use. Currently, there is limited evidence on whether those moving to vaping will continue to vape long-term or eventually stop using all nicotine. Thus, the benefits of smoking cessation may differ by any differential effects of ongoing vaping.

Quitting smoking has rapid beneficial effects on physical health and there is increasing evidence that it also leads to overall improvements in mental health, stress levels, and quality of life or well-being.⁷ Given its popularity,⁶it is important to find out whether continued vaping after smoking cessation leads to increased or reduced benefits in these domains.

NVPs are perceived by many ex-smokers to be a satisfying alternative to cigarettes as they fulfill some or all of the psychological functions that cigarettes deliver with the additional benefit of lower perceived risk.^1,8,9 Indeed, there is some evidence that nicotine has positive effects on certain cognitive domains including working memory and executive function,^10,11 and may assist coping and affect regulation.^12–14 Smokers also cite stress relief as one of the major reasons for smoking. Some of these effects are a result of misattributed withdrawal relief^15,16 and the perceived effects may be illusory. However, since stress is an important risk factor for smoking relapse,^17,18 understanding the effects on experienced stress levels as a function of continued use of nicotine post-smoking cessation is important.

Beyond nicotine delivery, qualitative work shows that vaping has characteristics that make it a more likely substitute than other nicotine products since it fulfills many of the physical, psychological, social, and cultural related dimensions of smoking.¹⁹ In particular, the social aspect of vaping is reported as a major source of appeal for ex-smokers.²⁰

Most adult vapers are former or current smokers.²¹ A recent scoping review of health effects of vaping²² found that very few studies were sufficiently rigorous to establish causality and form conclusions on health risks. A small number of human clinical studies that have specifically investigated regular smokers who have switched to exclusive vaping have demonstrated cardiovascular benefits in reduced blood pressure²³ and endothelial function,²⁴ and respiratory benefits with heavy smokers experiencing improved chronic obstructive pulmonary disease symptoms after switching to vaping.^25,26 These positive studies require replication and long-term follow-up. There has also been numerous animal and in vitro studies that reveal potential for adverse health effects,^27–30 but they may not be generalizable,^22,28 and toxic chemicals found in commercial combustible cigarettes are orders of magnitude greater than NVPs.^31,32 Critics of studies demonstrating the health benefits of switching to vaping, concede that the potential harms may largely fall on teenagers and young adults, who are at risk of becoming lifetime addicted to nicotine, but there may be actual health benefits to older smokers who make the switch to vaping.³³ Regardless of viewpoint, it is clear that well-designed longitudinal studies are required to assess the long-term effects of switching to vaping.

This paper explores the impact of ongoing NVP use (vaping) in comparison to quitting all nicotine use (abstainers) on perceived stress, capacity to cope with stress, having depression symptoms, and perceived change in health, in a population-based sample of ex-smokers. The main research questions are: (1) Do ex-smokers who vape differ from ex-smokers who do not vape (abstainers) on these health-related measures? and (2) Are there any differences by length of time since they quit smoking?

From the perspective of the model that nicotine use in former smokers confers real experienced benefits, continued vaping should lead to more vapers reporting benefits (or fewer losses) in terms of stress reduction and better mental well-being in comparison to complete cessation of nicotine use, with no adverse effects on perceived change in health. By contrast, if nicotine is contributing to psychological or physical pathology, we would expect fewer benefits or even worsen for those who continue to vape.

Methods

Sample

Data came from waves 1 (w1) (2016) and 2 (w2) (2018) of the ITC Four Country Smoking and Vaping Survey (ITC-4CV1 and ITC-4CV2) conducted in Australia (AU), Canada (CA), England (EN) and the United States.

The present analysis was limited to ever-daily smokers. There were three samples of interest (see Figure 1). For the two independent cross-sectional analyses, eligible participants were those who quit at the relevant wave for less than 2 years but had quit for at least 1 month. We also excluded 154 current (at least weekly) users of nicotine replacement therapy, 175 alternative tobacco product users (eg, smokeless tobacco or cigars) in the last 30 days, and a further 75 current users of a vaping product containing no nicotine. The latter were excluded because of the difficulty in separating effects attributable to the act of vaping from effects attributable to nicotine administration. There were 883 cases at w1 and 1088 at w2 (summed across waves: Australia: n = 245; Canada: n = 615; England: n = 548; United States: n = 439) with 123 individuals contributing to both.

Figure 1.

Cross-sectional and longitudinal samples. w1 = wave 1, w2 = wave 2, NVP = nicotine vaping products, NRT = nicotine replacement therapy.

The longitudinal sample (n = 804) (Figure 1) includes two sub-samples: (1) Those smoking at w1 (daily or weekly), who quit at w2 for at least 1 month (n = 372), and (2) those who were quit at both waves (ie, had remained quit for more than 2 years at w2) (n = 382). Those who quit at w1 but who relapsed and recovered between waves (n = 50) were excluded. Apart from the first sub-sample which included smokers at w1, all the same exclusions relevant to the cross-sectional samples applied to both sub-samples.

Full descriptions of the ITC Project conceptual framework³⁴ and methods^35,36 have been published elsewhere. In brief, the ITC Four Country Smoking and Vaping Wave 1 Survey (4CV1) sample comprised the following cohorts: (1) smokers and ex-smokers who participated in the previous wave of the ITC 4C Project, regardless of vaping status (re-contact sample), (2) newly recruited current smokers and ex-smokers (quit in the past 24 months) from country-specific panels, regardless of vaping status, and (3) newly recruited current vapers (use a vaping device at least weekly) from country-specific panels. Respondents for the ITC 4CV1 Survey were recruited via random-digit-dialing sampling frames, web-based or address-based panels, or a combination of these frames, as an expansion to the previous ITC 4C Project. In Canada, England, and the United States there was over-sampling of those who are aged 18–24 years old and/or those with experience of vaping.

Measures

NVP Status

Respondents were considered to be vapers if they reported that they vaped (nicotine) daily or at least weekly.

Perceived Well-Being Measures

Depression Symptoms

At both waves, used the two-item Perceived Health Questionnaire (PHQ-2)³⁷: “During the last 30 days, have you often been bothered by little interest or pleasure in doing things?” and “During the last 30 days, have you often been bothered by feeling down, depressed or hopeless?” Respondents were dichotomized into those who reported neither (no depression symptoms) versus either or both.

Perceived Stress

Assessed at both waves using a composite of two questions (r = 0.72), taken from Cohen’s Four-item perceived stress scale³⁸: (1) During the last 30 days, how often have you felt that you were unable to control the important things in your life? and (2) During the last 30 days, how often have you felt difficulties were piling up so high that you could not overcome them? Both were answered never, rarely, sometimes, often, very often, or don’t know. For analysis, “don’t know” responses were coded as “sometimes” and the two measures were combined into a continuous score ranging from 2 to 10.

Stress Management Since Quit

At both waves, respondents were asked “Since you quit smoking, has your ability to calm down when you feel stressed or upset improved, become worse, or stayed the same?” coded into those three options. This item has been used in previous ITC studies and is designed to assess the capacity to recover from a negative event.³⁹

Perceived Change in Day-To-Day Health (Last 6 Months)

At w2 only, respondents were asked Now thinking about your day-to-day health . . . In the last 6 months, have you noticed any change in your day-to-day health (such as changes in energy levels, coughs, etc.)? The response options were improved a lot, improved a little, not changed, become a little worse, became a lot worse, and don’t know. Because of low numbers in the latter groups, from “not changed” to “don’t know” were combined for multivariate analysis.

It is important to note that the time frame for the various measures differs including the two change measures and this should affect the interpretation of responses.

Demographic and Smoking-Related Covariates

Demographic measures included age (18–24, 25–39, 40–54, 55, and older years), gender, ethnicity (white, or English speaking in Australia, vs. nonwhite/non-English/identified minority group), education (Low, medium, and high) and Income (low medium, high, and refused to answer); and a measure of financial stress: “In the last 30 days, because of a shortage of money, were you unable to pay any important bills on time, such as electricity, telephone, or rent bills?”

Length of time quit was assessed by “How long ago did you quit smoking?” Categorical responses were grouped into: <1 month ago, 1–3 months ago, 4–6 months ago, 7–12 months ago, and 1–2 years ago. For the longitudinal analysis, we also divided the 1–2 year category, into 12–18 months including all those who are smoking during the previous wave and 18–24 months for those who quit at the previous wave and who reported no intermediate relapse. Pre-quit nicotine dependence (cigarettes per day) was assessed by current consumption at the predictor wave when smoking and reported consumption prior to quitting among those who quit at the predictor wave.

Post-quit nicotine dependence was assessed by urges to smoke: ‘In general, how strong have urges to smoke been in the last 24 hours?' (“I have not felt the urge to smoke in the last 24 hours, slight, moderate, strong, very strong, extremely strong, or don’t know”). After examining the distribution of responses, re-categorized into: No urges in the last 24 hours, slight urges, and moderate or greater urges, with the single “Don’t know” respondent excluded. Finally, self-perceived addiction: “Do you consider yourself addicted to cigarettes?,” “Not at all,” “Yes, somewhat addicted,” and “Yes, very addicted” with “Don’t know” responses (about 4% at both waves) recoded into the middle category.

Data Analysis

We first presented the descriptive analysis of the cross-sectional analysis at each wave. We then conducted regression analyses for each of the well-being outcomes for each wave. Proportions of vapers and non-vapers, with their 95% confidence intervals, are reported for each of the well-being outcomes, as derived from ordered logistic regression (change in perceived day-to-day health, having established it passed the proportionality assumption via the “brant test” Stata module), multinomial logistic regression (stress management), logistic regression (Depression symptoms) and linear regression (Stress). Finally, we conducted longitudinal analyses controlling for baseline measures of stress and depression when predicting w2 levels, meaning for these two outcomes it is effectively changed in levels of these measures that we are testing for. The models adjusted for country and each of the demographic and smoking-related covariates listed above including time since quit. Adjusted odds ratios (aORs), relative risk ratios (RRRs), or beta coefficients from each model, as appropriate, are also reported with their 95% confidence intervals. A p-value <.05 was considered statistically significant. All analyses were conducted using Stata SE Version 16.1 (StataCorp, TX, USA).

Results

Table 1 presents characteristics of the cross-sectional sample at each wave, by nicotine use status. Most had been quit for over 6 months. Over two-thirds (67.4% at w1 and 68.7% at w2) reported no longer experiencing any urges to smoke, yet fewer (42% at w1 and 43% at w2) considered themselves “not at all” addicted to cigarettes. The vaper samples consisted of 86% and 93% daily for w1 and w2 respectively, the remainder being weekly vapers.

Table 1.

Sample Characteristics and Perceived Well-being by Nicotine Use Status of the Wave 1 and Wave 2 Samples: Cross-Sectional Comparisons

	Wave 1 2016						Wave 2 2018
	NVP N = 189	Abstainers N = 694	Total N = 883	Chi	Sig	ES	NVP N = 326	Abstainers N = 762	Total N = 1088	Chi	Sig	ES
Country
Canada	22.8	35.3	32.6	22.2	<.001	0.16	17.2	40.6	33.5	138.5	<.001	0.36
United States	21.7	22.0	22.0				21.8	26.0	24.7
England	45.0	28.1	31.7				51.8	18.0	28.1
Australia	10.6	14.6	13.7				9.2	15.5	13.6
Gender
Male	43.9	46.8	46.2	0.5	.476	0.02	46.6	44.8	45.3	0.3	.569	0.02
Female	56.1	53.2	53.8				53.4	55.2	54.7
Age group
18–24 y	10.1	13.1	12.5	7.2	.066	0.09	6.8	6.7	6.8	16.4	.001	0.13
25–39 y	18.5	24.6	23.3				19.7	25.9	24.4
40–54 y	27.0	27.4	27.3				42.2	24.1	28.5
55 and up y	44.4	34.9	36.9				31.3	43.3	40.4
Ethnicity
White	95.2	87.1	88.8	9.8	.002	0.11	87.7	86.6	86.9	0.2	.632	0.01
Nonwhite	4.8	12.9	11.2				12.3	13.4	13.1
Education
Low	34.4	28.4	29.7	3.8	.152	0.07	29.3	28.3	28.5	11.6	.003	0.14
Moderate	36.5	43.8	42.2				51.7	38.3	41.5
High	29.1	27.8	28.1				19.0	33.5	30.0
Financial stress
Yes	9.0	10.5	10.2	0.4	.539	0.02	10.9	12.6	12.2	2.6	.109	0.02
No	91.0	89.5	89.8				89.1	87.4	87.8
When last quit
2–3 months	11.1	13.5	13.0	1.8	.625	0.04	15.6	16.8	16.5	0.9	.821	0.03
4–6 months	13.8	15.1	14.8				19.3	20.2	19.9
7–12 months	29.6	26.7	26.5				26.1	23.5	24.3
1–2 years	45.5	45.7	45.6				39.0	39.5	39.3
Cigs per day before quit (mean)	18.1	13.2	14.3	27.9	<.001	0.44	16.0	13.0	14.1	14.1	<.001	0.34
Perceived addiction to smoking post-quit
None	33.3	44.6	42.2	10.5	.005	0.11	38.0	45.1	43.0	6.0	.049	0.08
Somewhat	37.0	35.2	35.6				34.1	32.8	33.2
Very	29.6	20.2	22.2				27.9	22.1	23.9
No urge to smoke post-quit	73.9	65.6	67.4	6.1	.301	0.07	67.4	69.2	68.7	4.7	.450	0.03
Depression symptoms	32.7	44.5	42.0	8.5	.004	0.10	46.3	39.6	41.6	4.2	.041	0.06
Perceived stress
(0–10) mean (sd)	4.5 (2.2)	4.9 (2.0)	4.8 (2.0)	5.2	.023	0.19	4.6 (2.0)	4.5 (2.0)	4.5 (2.0)	1.3	.254	0.08
Stress management
Become worse	9.5	19.7	17.6	10.8	.004	0.11	6.8	16.1	13.3	17.2	<.001	0.13
Stayed same/dk	60.3	54.6	55.8				65.8	58.3	60.6
Improved	30.2	25.7	26.6				27.4	25.7	26.2
Perceived change in health*
Became a lot worse							0.3	1.3	1.0	9.5	.050	0.08
A little worse							2.2	4.0	3.4
Not changed							26.5	29.7	28.7
Improved a little							34.8	36.2	35.7
Improved a lot							36.3	28.9	31.1

Cross-sectional unweighted data: Past daily smokers, recent quitters, excluding quit in last month Abstainers = nicotine abstainers; NVP = Nicotine Vaping Products; ES = Effect size (Cramer’s V categorical variables, Cohen’s d continuous measures); *Perceived health not assessed at wave 1.

A number of consistent differences can be observed by nicotine use status in each wave. For example, vapers were more likely to live in England. Abstainers were more likely than NVP users to consider themselves “not at all” addicted when surveyed and smoked fewer cigarettes per day before quitting than vapers, but there was no difference in currently reported urges to smoke. Inconsistent across the two waves was the age distribution, at w1 NVP users were more likely to be older (≥55 years), whereas at w2 over 40% of all vapers were aged 40–54 years and the nicotine abstainers were older. Also inconsistent was ethnicity, the higher proportion of nonwhite vapers in w2, and education level achieved, with a lower proportion of vapers in the high category at w2 (19% compared to 29% in w1).

Overall, around 40% of quitters reported depression symptoms (42.0 at w1 and 41.6% at w2). Reported stress levels (out of 10) were middling (4.8, SD = 2.0 at w1) and (4.5, SD = 2.0, at w2), while more reported improved (26.2%) than reduced (13.3%) stress management, and most reported improved day-to-day health (w2 only, 66.8%) with only 4.4% indicating it had got worse.

Table 2 presents the multivariate analyses for each wave, predicting each health-related outcome controlling for the demographic and smoking-related characteristics listed in Table 1. Cross-sectional model-derived proportions and aORs/RRRs for each outcome by nicotine use status for each wave are presented.

Table 2.

Model-Derived Estimated Proportions for Each Well-Being Outcome, by Nicotine Use Status: Cross-Sectional Samples

Outcomes	Wave	Model n	Nicotine vaping products (NVP)	Predictors Abstainers	Adjusted odds ratio (aOR)	p
Depression symptoms (%, 95% CI)	1	866	34.3 (27.5–41.1)	44.1 (40.5–47.7)	0.64 (0.45–0.92)	.017
	2	1071	45.9 (40.5–51.4)	39.7 (36.3–43.2)	1.32 (0.98–1.78)	.069
Stress (last 30 days) (mean, 95% CI)	1	863	4.65 (4.4–4.9)	4.86 (4.7–5.1)	0.81 (0.59–1.12)	.197
	2	1070	4.54 (4.3–4.8)	4.48 (4.3–4.6)	1.06 (0.82–1.38)	.656
Stress management (since quit) (%, 95% CI)	1	871			1.71 (1.23–2.36)	.001
Became worse			9.8 (5.6–14.1)	19.9 (17.0–22.8)	0.46 (0.26–0.80)	.006
Stayed the same			56.8 (49.5–64.0)	55.3 (51.6–58.9)	ref
Improved			33.4 (26.4–40.4)	24.8 (21.7–28.0)	1.33 (0.90–1.96)	.156
	2	1073			1.55 (1.17–2.07)	.002
Became worse			6.2 (3.6–8.7)	16.5 (13.8–19.2)	0.30 (0.18–0.52)	<.001
Stayed the same			65.0 (59.6–70.4)	58.3 (54.8–61.6)	ref
Improved			28.8 (23.6–34.1)	25.1 (22.1–28.2)	1.03 (0.73–1.45)	.860
Change in perceived health (past 6 months) (%, 95% CI)	2	1069			1.65 (1.26–2.16)	<.001
Not changed or worse			27.9 (23.0–32.9)	35.6 (32.1–39.0)
Improved a little			31.8 (26.6–37.0)	37.0 (33.5–40.5)
Improved a lot			40.3 (34.6–46.0)	27.4 (24.3–30.6)

Table shows the results of three individual models for wave 1 and four individual models for wave 2.

Covariates in each model included country, gender, age group, ethnicity, education attained, financial stress, when last quit attempt, urges to smoke, and perceived addiction.

Cross-sectional unweighted data: past daily smokers, recent quitters, excluding quit in the last month. For stress management, ordinal regression odds ratios are provided for overall model fit and multinomial logistic regression odds ratios are provided for comparison at specific levels.

Depression symptoms were less common in the vapers at w1 (aOR = 0.64, 95% CI 0.45–0.92), but more common at w2, both effects were significant bivariate, although only a trend for the w2 effect in the multivariate analyses (aOR = 1.32, 95% CI 0.98–1.78).

For perceived stress, vapers reported lower stress at w1, but there was no difference at w2 and the w1 effect disappeared when controlling for the covariates.

Abstainers were more likely to report worsened stress management since quitting than vapers at both waves and in both sets of analyses (w1 RRR = 0.46, 95% CI 0.26–0.86; w2 RRR = 0.30, 95% CI 0.18–0.52). There were no significant differences in reported improvement in stress management at either wave.

W2 vapers were more likely to report that their day-to-day health had improved a lot in the last 6 months than the abstainers (40.3% vs. 27.4% of the abstainers) (aOR = 1.65:95% CI 1.26–2.16).

We also replicated this analysis by restricting vaping outcomes to daily vapers (see Supplementary Table A). In most cases the relationships between well-being measures and nicotine use status increased, for example, the odds of w1 stress management becoming worse decreased (aOR = 0.33 [95% CI = 0.17–0.64], p = .001), compared to when both weekly and daily vapers were included (aOR = 0.46 [95% CI = 0.26–0.80] p = .006) meaning that the difference with abstainers was greater for daily vapers.

Longitudinal Analyses

Characteristics of the longitudinal samples are in Supplementary Table B. Longitudinal model-derived proportions and aORs/RRRs for each well-being outcome by nicotine use status. Results for w1 smokers who quit at w2 (recent quitters) are in Table 3A with those who quit at both waves (longer-term quitters) in Table 3B. More details can be found in Supplementary Tables C and D.

Table 3.

Model-Derived Estimated Proportions for Vaping Status Each Wave 2 Well-Being Outcome Controlling for Wave 1 Measures. (Longitudinal Data)

W2 Outcomes	Model n	Nicotine vaping products (NVP) % (95% CI)	Abstainers % (95% CI)	Adjusted odds ratio (aOR)	p
A. Smokers at wave 1 who were quit at wave 2 n = 372
Depression symptoms (last 30 days)	362	49.1 (39.1–59.1)	35.9 (30.7–41.1)	2.00 (1.09–3.65)	0.025
Stress (last 30 days) (mean, 95% CI)	364	3.94 (3.6–4.3)	4.14 (4.0–4.3)	0.81 (0.54–1.23)	0.330
Stress management (since quit)				1.20 (0.70–2.08)	0.506
Became worse	363	6.0 (1.2–10.6)	13.2 (9.3–17.4)	0.37 (0.13–1.03)	0.058
Stayed the same		70.6 (61.3–79.8)	62.2 (56.6–67.8)	ref
Improved		23.4 (14.8–32.2)	24.5 (19.5–29.4)	0.81 (0.41–1.59)	0.536
Change in perceived health (past 6 months)				1.92 (1.16–3.20)	0.012
Not changed or worse	363	25.9 (16.6–34.9)	36.3 (30.7–42.0)
Improved a little		32.0 (20.9–41.1)	35.1 (29.8–41.1)
Improved a lot		42.1(32.4–54.0)	28.6 (23.0–33.5)
B. All who were quit at w1 and sustained to w2 (NB. Excludes 50 quit at w2 who relapsed but recovered at w2) n = 382
Depression symptoms (last 30 days)	356	30.2 (12.0–48.3)	36.4 (30.1–42.1)	0.69 (0.19–2.53)	0.576
Stress (last 30 days) (mean, 95% CI)	356	4.01 (3.4–4.6)	4.08 (3.9–4.3)	0.93 (0.45–1.91)	0.847
Stress management (since quit)				3.90 (1.42–10.71)	0.008
Became worse	356	2.9 (1.2–6.9)	13.7 (8.5–18.9)	0.34 (0.04–2.86)	0.320
Stayed the same		43.6 (22.5–64.6)	62.9 (56.6–69.2)	ref
Improved		53.6 (31.9–75.3)	23.4 (19.0–27.8)	5.05 (1.19–21.40)	0.028
Change in perceived health (past 6 months)				1.26 (0.47–3.35)	0.650
Not changed or worse	356	51.8 (32.5–71.1)	57.5 (51.1–63.9)
Improved a little		27.7 (9.1–46.2)	21.0 (15.8–26.2)
Improved a lot		20.5 (4.5–36.5)	21.5 (16.2–26.7)

Covariates in each model included country, gender, age group, ethnicity, education attained, financial stress (unable to pay bills), the strength of urges to smoke, perceived addiction to smoking, wave 1 depression, wave 1 stress, pre-quitting, or current cigarette consumption (cigarettes per day), wave 1 smoking frequency (daily, weekly), wave 1 vaping frequency (daily/non-daily), time since quit (months). Table 3B relapsed between w1 and w2 excluded. For stress management, ordinal regression odds ratios are provided for overall model fit and multinomial logistic regression odds ratios are provided for comparison at specific levels.

About half of all smokers who switched to vaping reported being depressed at w2 compared to about a third of the abstainers (51% vs. 35%, p = .006). In contrast, for the longer-term quitters (quit at both waves), only 23% of vapers reported being depressed at w2 compared to 39% of abstainers, (p = .011). In the sample of those smoking at w1 (and thus with stress and depression measured pre-quitting) (Table 3A), vapers were significantly more likely to have depression signs at w2 than abstainers (aOR = 2.00, 95% CI 1.09–3.65) but were more likely to report improvement in their day-to-day health over the previous 6 months (aOR = 1.92, 95% CI 1.16–3.20). By contrast, for those longer-term quitters who had sustained smoking abstinence before w1 (Table 3B), the pattern was notably different. There were no significant differences in depression signs, stress, and perceived changes in day-to-day health at follow-up, but net improved stress management among the vapers (aOR = 3.90, 95% CI 1.42–10.71). It is notable that the reported effects for depression and stress management in the abstainers did not change between the two groups, but the longer-term quitting vapers were more likely to report improvement in these areas, particularly in stress management than the more recent abstainers. When the whole sample of smokers and recent quitters at w1 and who quit at w2 were examined together (Supplementary Table D) it is clear that there are only small differences between abstainers and vapers, with the only statistically significant finding of stress management less likely to deteriorate for vapers.

Further analyses of the longitudinal sample indicated that w1 stress levels did not affect relationships for stress management or day-to-day health. The same was true for w1 depression levels, with the exception that the relationship between change in perceived health and vaping for smokers at w1 who quit at w2 (Table 3B) became greater if depressed, n = 144 (aOR = 2.84, 95% CI 1.24–6.46) compared to being non-depressed, n = 219 (aOR = 1.76, 95% CI 0.84–3.67) at w1.

Discussion

This study found little evidence consistent with a vaping harms model for any of our outcomes, and none were clearly supportive. By contrast, we found some evidence consistent with the positive effects of vaping, for both stress management and reported day-to-day health. The main exception is inconsistent results concerning depression symptoms, making them difficult to interpret. At w1 the vapers were less likely to report depression, but there was a reverse trend at w2, which effectively became significant for those smoking at w1 when controlling for signs at w1 (ie, in the longitudinal analysis). This pattern was not expected and is not readily explicable by any causal model of quitting. We cannot differentiate between shifts before quitting (ie, wave 2 depression occurred before the quit attempt) or after (potentially as a consequence), or too depressed smokers being more likely to manage to stay quit short term. The distinctly different pattern across the two waves makes an explanation based on sample differences seem plausible, but why such differences would occur in a study where the two waves were recruited similarly, and overlap considerably, is unclear. Although controlled for, this may be partly a function of the length of time after quitting. More data are needed before any clear interpretation is worth canvassing.

We found no evidence that reported current stress levels differed across vaping versus abstaining in the ex-smokers. However, we found clear, albeit not totally consistent, evidence that vapers were perceiving themselves as doing relatively better at stress management than abstainers. This is clearest for the longer-term quitters but exists in the cross-sectional samples (all quit less than 2 years), and there was a strong trend in the longitudinal analysis for fewer reporting negative effects.

In both samples, vapers were less likely to report worsened stress management, but only those in the longitudinal sample reported more improved coping. None of the control variables reduced the associations, so it may be that the lack of effect of improvement in the quit-between-waves sub-group is because such effects not appearing until some time after quitting. As only a small proportion of the ex-smokers reported a deterioration in stress management, effects on coping may be restricted to a small segment, rather than being a more general phenomenon. The finding of more positive effects on coping among vapers than abstainers is consistent with evidence that nicotine does have positive cognitive effects on at least a proportion of people,⁴⁰ but this may only become apparent after having quit for some time.

The finding that the vapers who quit between waves were more likely to report improved day-to-day health was unexpected, but not that there was no difference in those who quit for longer (Longitudinal sample). As the time frame was within the last 6 months it suggests that this perception develops in the months after quitting but after a year or more there was no differential effect. This is what would be expected if there was a short-term gain. It is difficult to conceive of any longer-term differential benefit of vaping on day-to-day health so finding no longer-term effect is unsurprising. Indeed, previous work demonstrated that having a health condition or reduced health associated with smoking was not systematically associated with increased vaping or increased likelihood of using vaping as a quitting strategy.⁴¹ Furthermore, almost a quarter of smokers with self-reported chronic health conditions denied smoking had even damaged their health.⁴² We think it unlikely that the differential response by time since quitting is because of a generalized desire to show switching to vaping in a positive light, but is more likely because of differential experiences as the effect disappeared in the period when there was no recent change in smoking status. It seems unlikely that vaping per se would improve physical health, but if it was associated with improved psychological functioning, then it is possible that vapers feel better overall. It is also possible that the presumed reduction in withdrawal-related effects associated with quitting might contribute to vapers feeling better than abstainers in the months following quitting. Regardless of whether there is a true effect, it is clear that vapers see themselves as having benefitted, and thus likely to be skeptical about broad claims of the harmfulness of vaping. Feeling better and more able to cope with stress are real benefits for those who experience them, and short of evidence that they mask considerably greater longer-term health negatives, should be considered in a cost-benefit assessment of vaping.

Overall, we think it unlikely that there are systematic social desirability effects contributing to the relatively better perceptions among vapers because, as noted above, we found no clear effects for reported stress levels and reverse effects for reported depression symptoms across the two waves as might be expected if the vapers were keen to highlight benefits of vaping.

There were differences while smoking between those who quit as vapers and those who abstained without vaping, so we cannot completely rule out some of the effects being because of preexisting differences, but as controlling for such differences had no effect, we think this is unlikely. Vapers reported considerably higher pre-quitting smoking cigarette consumption and perceive themselves as more addicted to cigarettes than the quitters. However, the effects were in relation to perceived changes so baseline effects are less directly relevant, and we found no evidence that controlling for these factors affects the strength of the associations. What limited evidence we have is that those who became vapers were somewhat more dependent⁴³ and being more dependent, had more room to improve, but this implies a positive effect. Even if the vapers are not more dependent on nicotine, it is reasonable to assume they are more interested in continuing to enjoy the experiences of nicotine use.

Finally, we found marginally stronger effects when we restricted the cross-sectional analyses to daily vapers (ie, excluding those vaping weekly), suggesting it may be the replacement for the nicotine that is at least partly responsible for the findings, but this is speculative.

This study has important limitations. As we make no claims about actual health effects, reliance on perceived effects is not a limitation per se, however, it would be useful to conduct studies relating self-report measures, such as those used here, to more objective indicators of health and of risk for future health problems. The two approaches may differ,^44,45 particularly for physical functioning, while for mental health, where experiences are central to diagnosis, mismatches seem unlikely. Perceptions are important, independent of any differences with underlying reality, as they are influences on the choices people make and the stories they tell others about their experiences.

The study is also restricted to ex-smokers of no more than 4 years duration and its representativeness is not clear. However, we are not claiming prevalence as outcomes, but focus on relationships between measures. While we found perceived stress management was improved in vapers among those who quit for more than 18 months, we found no clear additional benefits on perceived day-to-day health after this time. The study was also restricted to vaping, and we do not know if the effects would extend to the use of other forms of nicotine. Strengths of the study included the large cross-sectional sample size and the longitudinal study design, albeit with a smaller sample size, which permitted assessment of replication of findings at different waves. Finally, some of the effects were only marginally significant, so replication is needed before great confidence can be had in those findings.

It is important to better understand the role continued nicotine use may have on health. Our findings are consistent with studies on the use of snus in Scandinavia, that use of relatively clean forms of nicotine has few and smaller negative health effects compared to continued smoking,⁴⁶ and thus we might expect few short-term adverse effects. We cannot make any claims about long-term effects and our findings are restricted to ex-smokers. It will also be important to extend this work to switching to heated tobacco products as they become more available, as it is likely that adverse effects of switching relative to complete cessation will be detectable at some level of toxin delivery,⁴⁷ and discovering what levels are safe enough should be a priority. It will take far longer to detect any adverse effects of vaping on those who have not smoked regularly as compared with never smokers.

In conclusion, the findings are more suggestive of the positive effects of vaping on perceived health and stress management than they are of any negative effects, but it remains premature to claim clear health benefits of continuing to vape either on day-to-day health or the ability to cope with stress. The tentative conclusion about possible benefits only covers the early years of cessation, as it seems likely that some differential benefits may be relatively short-lived, or at least they do not increase beyond around 2 years after quitting smoking.

Supplementary Material

A Contributorship Form detailing each author’s specific involvement with this content, as well as any supplementary data, are available online at https://academic.oup.com/ntr.

Funding

The ITC Four Country Smoking and Vaping Survey in the United States, Canada, and England was supported by grant P01 CA200512 from the US National Cancer Institute, and a Foundation Grant (FDN-148477) from the Canadian Institutes of Health Research. The ITC Australia Project was supported by the National Health and Medical Research Council of Australia (GNT1106451).

Declaration of Interests

No potential conflict of interest was reported by the authors. No conflict of interest exists in the submission of this manuscript, and manuscript is approved by all authors for publication. This work was original research that has not been published previously, and is not under consideration for publication elsewhere. The authors listed have approved the manuscript that is enclosed. The paper was initiated and an early draft of parts was prepared by JB, before his sudden death in 2020. This paper is dedicated to his memory.

Dedication

This paper is dedicated to the memory of Dr James Balmford who initiated the study and had prepared the first draft before his untimely death in March 2020.

Data Availability

Data from the International Tobacco Control Policy Evaluation (ITC) Project are available to approved researchers 2 years after the date of issuance of cleaned data sets by the ITC Data Management Centre. Researchers interested in using ITC data are required to apply for approval by submitting an International Tobacco Control Data Repository (ITCDR) request application and subsequently signing an ITCDR Data Usage Agreement. To avoid any real, potential, or perceived conflict of interest between researchers using ITC data and tobacco-related entities, no ITCDR data will be provided directly or indirectly to any researcher, institution, or consultant that is in current receipt of any grant monies or in-kind contribution from any tobacco manufacturer, distributor, or other tobacco-related entity. The criteria for data usage approval and the contents of the Data Usage Agreement are described online (http://www.itcproject.org).