Protective Motivation Theory and Intention to Quit Smoking: Results Based on Structural Equation Modeling and Mediation Analysis | BMC Public Health

Procedure

All study procedures were approved by the Peking University Health Sciences Center Ethics Board (Ethics approval number: IRB00001052-18,055). This was a cross-sectional study conducted from July to August 2020. Peking University Health Science Center School of Public Health sent investigators to a total of 26 provinces. Our study covers the majority of the 31 provinces of mainland China, except for Shanghai, Tianjin, Xinjiang, Tibet and Qinghai. Participants were randomly selected on the street in major business districts and residential areas in urban regions. Data was collected through face-to-face interviews. The survey lasted about 8-10 minutes for each participant. We have not provided any incentive to participate.

Speakers

We recruited participants who met the following criteria: (1) were between the ages of 18 and 60; (2) were daily smokers, with smoking duration ≥ 1 year; (3) smoking cigarettes; (4) lived in current location for ≥ 5 years; and (5) agreed to participate in the survey.

Sample size

The sample size calculation was based on a cross-sectional design. According to the 2018 China Adult Smoking Survey [21], only 5.6% of smokers intended to quit, so we assumed that 6.0% of the population would intend to quit; a value of 0.80 (beta = 0.20) was used for potency and 0.05 was used for alpha. Therefore, the minimum sample size was 550.

Data collection and informed consent

A four-step method was applied in this investigation. In the first stage, experts from Peking University designed the standardized questionnaires. In the second step, an online survey system was developed and a specific web link was generated. In the third step, the investigators presented the content and purpose of the survey according to specific presentations to potential participants and obtained verbal informed consent from all participants. All participants were informed that the statistical analyzes would be conducted anonymously and that their information would be used for research purposes and published. In the final step, all interviewers enter information via the Internet after completing the face-to-face interviews.

The online inquiry system had a self-checking function and can automatically identify missing data, logical errors and illegal characters.

Measurements

Measurements of PMT constructions

PMT constructs were assessed using the PMT scale, based on the work of Xu et al. [14]. We also improved and adjusted some of the questions to fit measures of intention to quit. Tests of the validity and reliability of the questionnaire have been carried out and have been reported elsewhere [15]. Specifically, the scale consisted of 21 items and used a 7-point Likert-type response scale ranging from 1 (strongly disagree) to 7 (strongly agree). Each construction subscale included three items and we calculated the mean as the subscale score.

Perceived severity was measured by three items: “The earlier a person starts smoking, the greater the damage”, “More smokers get sick than non-smokers”, and “Smokers die earlier than non-smokers”. smokers “. Perceived vulnerability was measured by three items: “I would become addicted if I smoked”, “I would get sick if I smoked” and “If I smoked, I could die earlier”. Intrinsic rewards were measured by three items: ‘Smoking makes people feel comfortable’, ‘Smoking helps people concentrate’ and ‘Smoking improves brain work’. Extrinsic rewards were measured by three items: ‘Smokers look cool and fashionable’, ‘Smoking is good for social media’ and ‘A smoker’s life is happier than a non-smoker’. -smoker”. Self-efficacy was measured by three items: “I am convinced that I can successfully quit smoking”, “I have the ability to quit smoking” and “I think quitting smoking is easy for me “. Response effectiveness was measured by three items: “People will feel good not smoking”, “People will be less likely to get a disease if they don’t smoke”, and “Quitting smoking is good for you.” the cure of the disease”. The response cost was measured by three items: “A person can be isolated if he quits smoking”, “Refusing a cigarette offer is very impolite” and “We will miss the pleasure if he quits smoking”.

Threat Assessment and Adaptation Assessment Measures

We created two PMT pathway scores based on the seven PMT constructs described above. The threat rating pathway was calculated as the difference between perceived threat (calculated as the mean of vulnerability and severity items) and perceived rewards (calculated as the mean of extrinsic and intrinsic reward items). The coping assessment pathway was calculated as the difference between perceived efficacy (calculated as the average of response efficacy and self-efficacy items) and perceived cost (equal to construct score the cost of the response).

Measures of intention to quit

Intention to quit smoking was assessed using two questions: “I plan to quit smoking next month” and “I plan to quit smoking next month”. We also used a 7-point Likert-type response scale ranging from 1 (strongly disagree) to 7 (strongly agree). This has been used by other researchers [16]. We calculated the mean as a subscale score.

Other Variables

We also collected several individual characteristics variables, such as socio-demographic information, including gender, age, BMI, marital status, ethnicity, education, annual income, chronic diseases and employment. Smoking status, including number of cigarettes per day, duration of smoking, withdrawal symptoms, awareness of the harms of smoking, and e-cigarette use, was assessed in the second section of the questionnaire. Information about quitting smoking, including the questions “Have you ever tried to quit smoking?” » ; ‘What is the longest time you have quit smoking?’; and ‘When do you want to quit now?’ in the third section of the questionnaire.

Data analysis

Our data analysis proceeded in three stages. First, we assessed the reliability of individual PMT constructs. Second, we tested how well the PMT constructs predicted intention to quit smoking. Next, we performed a mediation analysis to assess possible mediators. SPSS 19.0 was used for reliability testing and AMOS 24.0 was used for structural equation modeling (SEM) and mediation analysis.

Step 1: Assess the reliability of individual PMT constructs

Cronbach’s alpha and interclass correlation coefficient (ICC) were used to assess the reliability of individual PMT constructs. A Cronbach’s alpha > 0.7 indicates good internal consistency [18].

Step 2: Test the ability of PMT constructs to predict quit intention

We applied SEM to test the ability of PMT constructs to predict intention to quit smoking. SEM was a better method for our study than correlation or multiple regression analyzes because it could be used to test global models rather than individual coefficients and because it incorporates several dependent and mediating variables. [22, 23].

We used the following model fit statistics which were found to be significant in SEM [23, 24]: Bentler’s comparative fit index (CFI): recommended > 0.90; Tucker-Lewis Index (TLI): recommended > 0.90; Standard Fit Index (NFI): recommended > 0.9; Root Mean Squared Error of Approximation (RMSEA): Recommended ≤ 0.08.

Step 3: Mediation analysis

We then performed a bootstrap test to estimate the extent to which the influence of PMT constructs on threat assessment and coping assessment was affected by various mediators. This method is widely used to assess whether a factor mediates a relationship between two variables [25, 26]. Moreover, as a complementary method, we also tested the mediating effects using a Sobel test [25]. As the result was almost the same, here we only report the results of the bootstrap test.

Sharon D. Cole