Functional beliefs and risk minimizing beliefs among Thai healthcare workers in Maharaj Nakorn Chiang Mai hospital: its association with intention to quit tobacco and alcohol

Health beliefs about smoking and alcohol drinking were evaluated separately using a nine-item questionnaire derived from previous literature [18, 19]. The first five items were related to functional beliefs of smoking or alcohol use. The last four items referred to risk minimizing beliefs about the risk and harms of smoking or alcohol use (Tables 1, 2). Participants rated their agreement with each functional and risk minimizing belief. Agreement scores ranged from one to five. A score of five indicated that the participant totally agreed with the statement and a score of one indicated that the participant totally disagreed with the statement.

The history of each substance used was initially categorized into four groups. The first group was for those who had never used substances. The second group was those who were former users defined as having stopped using the substance for longer than one year. The third group was those who were recent quitters defined as having recently stopped within one year and the last group of users were those who categorized themselves as currently using the substance (within the past 3 months). If participants were currently using a particular substance, they were asked about their intention to quit each particular substance using a four-item response: 1-no intention to quit, 2-intention to quit in the next six months, 3-intention to quit within six months and 4-intent to quit within 30 days. If participants indicated that they planned on give up a particular substance, they were also asked about their confidence to quit each particular substance by using a five-item response of success: 0-not at all confident, 1-not confident (< 25% chance of success), 2-moderately confident (25-50% chance of success), 3-confident (50-75% chance of success and 4-very confident (> 75% chance of success).

Participants who were current users and those who had stopped within the past year were used in the analysis for each substance. Separately for tobacco and alcohol consumption, a factor analysis with orthogonal rotation (varimax) was used to examine whether the nine-item questionnaires formed the two coherent health beliefs (functional beliefs and risk minimizing beliefs) as hypothesized. A loading factor of at 0.4 was used as a cutoff point [27]. Kruskal-Wallis test was used to determine the association between the agreement score for each item with intention to quit. Using predicted factor scores from factor analysis, the relationship between factor scores for each of the two beliefs and intention to quit and confidence to quit were tested using ANOVA. These associations were further adjusted for age and sex using linear regression, as it was an a-priori belief that these two factors could be considered as potential confounders. Lastly, functional belief scores and risk minimizing belief scores were categorized into quartiles and its association with recent cessation of smoking and alcohol use using logistic regression were examined. Sensitivity analyses were performed by excluding recent quitters from the analyses. A p-value of ≤0.05 was considered statistically significant. All analyses were conducted using STATA version 12.0.

Using factor analysis with orthogonal rotation (varimax), there were four potential factors that could be derived from the nine-item questionnaire. However, only in the first two factors the individual items had a factor loading of greater than 0.4 (Additional file 1: Table S3), thus supporting that the nine-items form two types of coherent beliefs (Table 1). Results suggest that current smokers and recent quitters engaged in a number of functional and risk minimizing beliefs about smoking. The most common functional belief was that “Smoking calms you down when you are stressed or upset”, found in about 33.7% of smokers and recent quitters. The most common risk minimizing belief was that “The medical evidence that smoking is harmful is exaggerated”, was found in 43.8% of smokers and recent quitters (Additional file 1: Table S4). Results from Table 1 suggest that each of the functional beliefs were associated with intention to quit. Higher agreement in each functional belief about smoking was inversely associated with the intention to quit. Only three of the four risk-minimizing beliefs were associated with intention to quit (Table 1).

Using factor scores derived from factor analysis, only the factor score for functional beliefs was associated with intention to quit smoking and the confidence level in quitting. Adjusted for age and sex, individuals with higher functional beliefs of smoking were less likely to quit smoking (Fig. 1) and were less confident of being able to quit (Fig. 2). The factor score for risk minimizing beliefs was not associated with intention to quit smoking (Fig. 1) and the confidence level in quitting (Fig. 2). The associations with intention to quit did not materially change when recent quitters were excluded from the analyses (Additional file 2: Figure S1).

In a multivariate regression model, there was some weak evidence of a gradient in associations between quartiles of functional belief scores and recent smoking cessation (Wald test statistic = 3.02,df = 1, p = 0.08) as well as between quartiles of risk minimizing belief scores and recent smoking cessation (Wald test statistic = 6.35,df = 1, p = 0.02). Those with functional beliefs scores and risk minimizing scores in the highest quartile were less likely to have recently quit smoking (Table 3).

Similar to results for smoking, although there were four potential factors that could be derived from the questionnaire (Additional file 1: Table S5), the loading factors were aggregated towards two types of coherent beliefs (Table 2). For functional beliefs, approximately 20% agreed that “Drinking calms you down when you are stressed or upset” and that “Drinking makes it easier for you to socialize”. The most common risk minimizing belief was that Medical evidence that drinking is harmful is exaggerated” (Additional file 1: Table S6). Displaying a similar pattern to the intention to quit smoking, higher agreement in each of the functional beliefs of alcohol was inversely associated with the intention to quit (Table 2). All four risk minimizing beliefs of alcohol were associated with an intention to quit alcohol (Table 2). Adjusting for age and sex, factor scores of functional beliefs and risk minimizing beliefs were each inversely associated with intention to quit alcohol drinking (Fig. 3) and confidence to quit drinking (Fig. 4). Those with higher scores in functional and risk minimizing beliefs about drinking were less likely to quit drinking and had a lower confidence in quitting. Some power was loss in the sensitivity analysis but the associations with intention to quit did not materially change when recent quitters were excluded from the analyses (Additional file 3: Figure S2).

In a multivariate regression model, there was a gradient in associations between quartiles of functional belief scores and recent cessation of alcohol use as well as between quartiles of risk minimizing belief scores and recent cessation of alcohol use. Compare to those in lowest quartile of functional belief scores, the odds ratio for recent cessation of alcohol use was 0.22 (95% CI 0.15 to 0.33) for those with score in the 2nd quartile, 0.11 (95% CI 0.07 to 0.16) for those in the 3rd quartile and 0.04 (95% CI 0.03 to 0.07) for those with the highest functional belief scores (Wald test statistic = 178.2,df = 1, p < 0.01). This gradient was also demonstrated with risk minimizing beliefs scores. Compared to those in the lowest quartile of risk minimizing beliefs scores, the odds ratio for recent cessation of alcohol use was 0.50 (95% CI 0.34 to 0.74) for those with score in the 2nd quartile, 0.32 (95% CI 0.21 to 0.48) for those in the 3rd quartile and 0.21 (95% CI 0.13 to 0.32) for those with the highest functional belief scores (Wald test statistic = 53.4,df = 1, p < 0.01) (Table 4).

When considering the intention and confidence to quit smoking, the highest agreement on risk minimizing beliefs from our sample was the skeptical belief that “Medical evidence that smoking is harmful is exaggerated”. This result was similar to a previous study in Australia [19] which reflected lack of understanding in the harm or health outcomes from smoking and its consequences. This belief can usually be found in the smokers, which was also reflected in our study. However, using factor scores, our study did not find an association between risk minimizing beliefs and intention to quit and confidence to quit smoking. Risk minimizing beliefs can be considered “weak beliefs” as they can be easily influenced and changed [28]. This potential fluctuation in risk minimizing beliefs and our small sample of current smokers and recent quitters may be reasons why this study could not detect any association between risk minimizing beliefs and the intention to stop smoking.

Higher agreement in functional beliefs about smoking was inversely associated with intention to quit and confidence to quit smoking, which has been observed in a previous study [29]. Smokers have dissonant reduction when they attempt to enhance the functional beliefs in smoking [18]. When exploring each functional belief individually, the functional belief that was most commonly found was that “Smoking can calm you down when you are stressed or upset”. This is not surprising as nicotine can modulate pathways involved in stress response, depression and anxiety [30]. Our study found that the few smokers had functional beliefs about ‘Smoking for social enhancement’. This is potentially due to the regulatory environment in Thailand that restricts smoking areas and has banned smoking advertisements on television and radio [20].

Similar to results for smoking, this study found an inverse association between functional beliefs and the intention to quit alcohol drinking. This may be because concurrent use of alcohol and smoking are common [25]. However, when exploring each functional belief individually, there were different patterns of functional beliefs between smoking and alcohol drinking. Firstly, in contrast to smoking, functional beliefs of drinking for “social enhancement” was quite common. In Thailand, alcohol use is integrated into social norms and also traditional rites. In additional, drinking is rarely perceived as a social problem [31, 32].

Risk minimizing beliefs about drinking were significantly associated with the intention to quit and confidence to quit alcohol drinking, while this association was not found in smoking. Currently drinkers tend to have risk minimizing beliefs as most of Thai people who use alcohol are at low to moderate risk of harm, which may not show the serious health outcomes [33, 34]. Furthermore, there is conflicting evidence regarding the potential protective effect of alcohol against coronary heart disease [34, 35], which may be why risk minimizing beliefs are associated with alcohol use and inversely associated with the intention to quit.

The present study has some limitations. The response rate of our study was 60%, which may introduce some selection bias. However, in a previous publication, we have demonstrated that our sample was representative of the source population in terms of age, sex and education level [26]. Because of the data was based on self-report, health beliefs and the intention to quit smoking and alcohol drinking may be vulnerable to social desirability bias. However, voluntary participation, assurances of confidentiality in this study may have reduced some of the impact of social desirability bias. This study was a cross-sectional data, we can only assume the temporal relationships between decreasing functional beliefs and risk minimizing beliefs with subsequent changes in willingness to quit smoking or alcohol use. However, some evidence from prospective studies of smoking have supported this notion [13, 18]. The number of smokers enrolled in this study was small, thus the estimates were imprecise and could be underpowered to detect the association between minimizing beliefs and the intention to quit and confidence to quit smoking. As other health beliefs, in particular positive or protective beliefs, were not explored in this study, we could not provide evidence on what beliefs may be protective or promotes intention to quit. Nonetheless, this study also has some strengths. Factor analysis was utilized to derive factor scores of beliefs rather than just a single question which had previously been common in past literature [36]. It is also one of the first studies to report findings for both alcohol and tobacco use.