provides weighted characteristics of the sample by country. About 12% of the total Wave 4 sample experienced FS. The UK had the lowest (6%) and the US had the highest (13%) percentage of smokers who reported FS. Canada had the highest (76%) and the UK had the lowest (64%) percentage of smokers who had an interest in quitting. Percentage of smokers who made a quit attempt was highest in Australia (42%) and lowest in the US (33%). Successful cessation rate was highest in the UK (37%) and lowest in the US (26%).
shows the characteristics of smokers with FS and the bivariate associations of FS with the outcomes and other variables. Country data were pooled as there was no evidence of interaction between FS and country in their association with the outcomes. About 83% of smokers with FS and 69% without FS had an interest in quitting. At the bivariate level, there was no association between FS and quit attempts. Among smokers who attempted to quit, 20% of those with FS and 31% of those without FS were successful. Smokers with financial stress were more likely to be younger and have lower income than other smokers.
shows adjusted odds ratios for the association of covariates with interest in quitting in the same wave, and quit attempt and quit success at the next wave. The odds of having an interest in quitting were 71% (p < 0.004) higher in smokers with FS than others. Other factors that were associated with a higher probability of interest in quitting were: spending too much money on cigarettes, higher self-efficacy, being worried about the effect of cigarettes on health and quality of life, reporting that smoking has lowered one's quality of life, lower levels of nicotine dependence, being female, being under 40 years old, not having low education, and higher income. The probability of interest in quitting was highest in Australia and lowest in the UK. We note that when interest in quitting was defined as planning to quit within the next six months (as compared to beyond six months or no plan at all), financial stress did not have an association with interest in quitting. Furthermore, among those who had an interest in quitting, there was no evidence of an association between financial stress and whether a respondent had an interest in quitting within the next six months or sometime in the future.
While FS and subsequent quit attempts were not associated at the bivariate level, after controlling for interest in quitting and the respondents' reports on whether smoking had lowered their quality of life at the baseline (Wave 4), the data provided considerably more evidence for an association. In the multivariate analysis, the odds of making a quit attempt by the next wave were 26% (p = 0.023) lower in smokers with FS than others. Interest in quitting, being worried about the effect of cigarettes on health and quality of life, reporting that smoking has lowered one's quality of life, lower nicotine dependence, being under 25 years old, and not having low income were independently associated with a higher probability of making a subsequent quit attempt. The probability of a subsequent quit attempt was highest in Australia and lowest in the US.
Among those who tried to quit smoking, the odds of a successful attempt were 48% (p = 0.012) lower in smokers with financial stress than others. Lower nicotine addiction was independently associated with a higher probability of quit success. The probability of quit success was highest in the UK and lowest in Canada.
There was no interaction between FS and country in any of the regression models.
Whether a person had an interest in quitting, made a quit attempt, or had quit smoking in Wave 4 was not associated with being lost to follow-up in Wave 5, with p-values of 0.797, 0.640, and 0.701, respectively. Lower age and education in Wave 4 were associated with being lost to follow-up in Wave 5 with p values of <0.001 and 0.006, respectively. None of the other covariates were associated with attrition. The exact impact of attrition on results cannot be assessed. However, by including the correlates of attrition (ie age and education) in the multivariate analyses, we have partially controlled for bias.
Supplementary analyses showed that there was no evidence of an association between financial stress and three variables in the ITC dataset that pertain to the use of services or products to help cessation: use of “stop-smoking medications” (p = 0.665), doctors' “help or referral to another service” as an aid to cessation (p = 0.791), or use of telephone quitlines (p = 0.667) in wave 5. Consequently, including these variables in the quit attempt and quit success analyses resulted in negligible change in the effect of financial stress. For example, when these three variables were included separately in the quit success regression model (), the odds ratio for the effect of financial stress changed from 0.5338 (p = 0.012) to 0.5343 (p = 0.012), 0.5083 (p = 0.007), and 0.5336 (p = 0.012), respectively.
In further supplementary analyses, we included health (“In general, how would you describe your health? Is it … 1. Poor, 2. Fair, 3. Good, 4. Very good, or 5. Excellent”) in the models. This variable was not associated with interest in quitting (p = 0.204), quit attempt (p = 0.599), or quit success (p = 0.312). Consequently, the inclusion of health in the regression models did not make an appreciable change in the effect of financial stress. For example, after health was included in the quit success model, the odds ratio for financial stress changed from 0.534 (p = 0.012) to 0.533 (p = 0.012). Thus, we did not include health in the analyses.