Khat chewing and health related quality of life: cross-sectional study in Jazan region, Kingdom of Saudi Arabia

This was a cross-sectional survey study of male khat chewers and non-khat chewers who attended primary healthcare centers in Jazan province in the Kingdom of Saudi Arabia. The inclusion criteria for the participants of the study were male ≥18 years old attending selected primary healthcare centers located in the Jazan region, no disabilities that would prevent understanding the questionnaire, no consumption of substances other than khat and tobacco, and no associated co-morbidities and history of chronic diseases. Because of cultural sensitivities and the domination of men in khat chewing, we excluded the female population from the study sample.

A total of 630 participants from two independent populations of khat and non-khat chewers were selected to participate in this study from March 2 to July 31, 2012. The questionnaires were designed to be self-administered, but in some cases a trained research assistant read the questionnaire to illiterate subjects and recorded their responses. The sample required to compare the two population means of khat chewers and non-khat chewers was 608 (304 subjects in each group) using the following formula[37, 38]:

where α = 0.05 and β = 0.20. The values of z_1 - a/2 and z_1 - β are 1.96 and 0.84, respectively, 2(z_1 - a/2 + z_1 - β)² = 15.68, and  d² = δ/σ (δ is the difference between the two population means and σ is the standard deviation of the measurements; both δ = 5 and σ = 22 were found from a pilot test). Therefore, the sample size required in each group was calculated to be 304 patients and the two samples scenario required a total of 608 subjects. We allowed 5% for incomplete data or missing values in the data, and the required sample size was increased to be 630 participants. The sampling was done at in two stages. In the first stage a random sample of four out of 13 primary healthcare centers were selected. In the second stage, within each selected medical center, we invited all patients who visited the clinic during the research duration to participate in our study, and collected the information required until we obtained the sample size required. A sample of 630 (94%) who completed the questionnaires were used for the analysis. 40 (6%) patients refused to participate (27 [4%]) in the study or provided incomplete data and (13 [2%]) were excluded.

This study was reviewed and accepted by the ethics committee at Jazan University. All of the participants of the study signed a written informed consent and agreed willingly and voluntarily to participate in the study.

The SF-36 is currently the most commonly used health status measure worldwide[38]. This instrument is a standardized generic instrument developed by the RAND Corporation (USA). The SF-36 provides a simple descriptive summary and a single index value for health status[39]. The SF-36 consists of 36 questions divided into eight domains: physical functioning (PF), bodily pain (BP), social functioning (SF), general health (GH), role-physical (RP), role-emotional (RE), vitality (VT), and mental health (MH). The item scores in each of the factors can be coded, summed, and transformed into a scale ranging from 0 (the worst imaginable health state) to 100 (the best imaginable health state). This instrument is easy to administer, not time consuming, available in Arabic translation, and valid (Cronbach’s alpha value was 0.65)[39, 40].

The SF-36 questionnaire was used to measure the HRQoL of khat chewers in the Jazan region. Even though the SF-36 questionnaire is a self-administered instrument, six research assistants were recruited and trained by the research team to help patients having difficulty understanding the questions. These trained research assistants were available during the research period at every clinic to assist the study participants in case the participants have any problem in understanding the context of the questions. These research assistants are suppose to provide guidance only and do not influence the participants opinion in any way. Figure 1 shows the conceptual framework used with the SF-36 instrument in this study. The eight components of the SF-36 score were classified into two main groups: physical health and mental health questions. The physical health section includes questions on PF, RP, BP, and GH, whereas the mental health section questions are on MH, RE, SF, and VT.

Descriptive statistics for each dimension reported for the SF-36 questionnaire included the mean, standard deviation, coefficient of variation, skewness, kurtosis, and range for the scales. The demographic variables of the respondents were also reported with the age of the respondents in both groups categorized into four groups: 18–24 years, 25–34 years, 35–44 years, and ≥ 45 years. The odds ratio test was used to identify the association between the demographic characteristics of the respondents and their status of khat chewing. Mann–Whitney (MW) and Equality of mean tests with 95% CI were used to compare the differences in HRQoL between khat chewers and non-khat chewers. MW test is a non-parametric test used when the dependent variable cannot be assumed to be normally distributed. The Kolmogorov-Smirnov test was used to evaluate the normality of the study data collected using the SF-36 instrument. The kurtosis and skewness statistics of the two samples indicated non-normally distributed data. The coefficient of variance of the two groups reflected that the data had an acceptable degree of variance and homogeneity[41, 42]. Cronbach’s alpha and the intraclass correlation coefficient (ICC) were used to evaluate the reliability and consistency of the SF-36 instrument. The ICC approach used was two-way mixed effect model with consistency type. A reliability of 0.6 is considered as acceptably reliable, while 0.8 or higher is considered a good reliability; and for the ICC a value greater than 0.3 is acceptable. Finally, a pilot test of 60 and 56 respondents on each of the two independent groups of khat chewers and non-khat chewers was performed, respectively, to calculate the δ (difference between the two population means) and σ (standard deviation of the measurements). All statistical analyses were performed using SPSS 16.0. A statistical value of P < 0.05 was taken as significant.

The SF-36 questionnaire was distributed to 670 participants from two independent populations of khat and non-khat chewers. Finally, a sample of 630 (94%) who completed the questionnaires were used for the analysis. The participants completed the questionnaire in 10–30 min, with 80% completing the questionnaire in 15 min or less. A total of 40 (6%) patients either refused to participate in the study or provided incomplete data (27 [4%] and 13 [2%] patients, respectively).

The descriptive statistics for the SF-36 data collected for both groups are given in Table 1. The Kolmogorov-Smirnov test for normality provided evidence that the distributions of the two populations of khat chewers and non-khat chewers were significantly different from each other (P < 0.001); indicating non-normally distributed data. Although the normality assumption is violated, we still interpret the mean values and a coefficient of variation of the data, reporting the mean for interpretation purposes is required, even when the distribution is not normal. In the sample of non-khat chewers, the average values for the eight components of the SF-36 were between 88.71 (GH) and 96.47 (SF), whereas for the sample of khat chewers, the average values of the eight components of the SF-36 were between 52.83 and 78.60. Non-khat chewers had higher mean scores for all the eight subscales of the SF-36 compared with khat chewers (Table 1). The overall mean score of HRQoL for the non-khat chewer group was higher (p < 0.001) than that in the khat chewer group (92.7% vs 63.5%).

Cross-tabulation analysis was performed to determine the association of some selected demographic variables versus the status of khat chewing. Four demographic factors were selected for this analysis; marital status, employment status, age classified into groups, and education. Table 2 shows the cross-tabulation results of socio-demographic variables and the status of khat chewers and non-khat chewers. Among the respondents from the khat chewer group, 148 (39.2%) khat chewers were married, whereas 167 (66.5%) khat chewers were not married. In contrast, among the respondents from the non-khat chewer group, 230 (60.8%) non-khat chewers were married, whereas 85 (33.7%) non-khat chewers were not married. This result showed that the odds of being a non-khat chewer being married were 67.2% higher than the reference group of unmarried.

With regard to employment status, the majority of self-employed, government-employed, and privately employed respondents were reported to be non-khat chewers, while the majority of unemployed and students were reported to be khat chewers. The odds of being a khat chewer among government employees was 1.56 (95% confidence interval [CI]: 0.845, 2.869) times higher than the reference group of the self-employed respondents, but this was not significant. The odds of being a khat chewer among the unemployed and student participants was 6.68 (95% CI: 3.566, 12.542) times and 16.640 (95% CI: 9.436, 29.34) times higher, respectively, than the reference group of self-employed.

Older respondents were more likely to be non-khat chewers than younger respondents. The probability of khat chewing among the age group of 25–34 years was identical to that of the reference age group (odds ratio: 1.13; (95% CI: 0.717, 1.790). The older age groups of 35–44 years and ≥ 45 years had a lower probability of being a khat chewer compared with the reference age group, 0.322 (95% CI: 0.207, 0.501) and 0.167 (95% CI: 0.101, 0.276), respectively. Finally, the respondents with an education level of secondary school or higher were less likely to be khat chewers than respondents with a lower education level. The odds of being a khat chewer/non-khat chewer among respondents with a primary education were not different from the respondents with no formal schooling (odds ratio: 1.091; 95% CI: 0.518, 2.300).

The mean differences of the physical health perceptions were between 13.71 for bodily pain (BP) and 24.29 role-physical (RP); whereas, the mean differences of the mental health perceptions were between 37.47 Mental Health (MH) and 42.33 Role Emotional (RE). On average, the health gap (mean difference) between khat and non-khat users for mental health perceptions was larger than the health gap for physical perceptions. This indicates that khat chewers complain more about mental and social problems rather than physical problems (Table 3). The Mann–Whitney U-test showed that non-khat chewers had significantly higher mean scores of HRQoL across all of the domains of the SF-36 than khat chewers (P < 0.001).

Table 4 shows that all of the eight subscales of the SF-36 instrument demonstrated high internal consistency, as measured by Cronbach’s alpha, which approximately exceeded 0.7 (range, 0.73–0.95). The table also shows that the ICC ranged from 0.312 to 0.769, which is acceptable and indicates consistency of the data[41]. Generally, respondents from the sample of khat chewers had higher ICC values and Cronbach’s alpha than respondents who were non-khat chewers.

Similar to other self-reported questionnaires, this study was subject to recall bias of patients and selection bias of medical centers. Additionally, instead of using a generic instrument (SF-36 questionnaire) it may be more appropriate to use a specifically designed instrument for the effects of khat. The sample representativeness is questionable; meaning any inferences from this study population to the whole population was not appropriate. The HRQOL was not verified in this study with medical records, which could have added value to results. Tobacco smoking was not investigated; this could have confounded the results. The social desirability in this study may have been eliminated through using the self-administered questionnaire though we do not know the number of participants who self-administered questionnaires and those who provided data through face to face interviews. The implication of this study for future research with respect to validation of these study findings and answering relevant questions that include the health economic impacts (cost analysis) of khat chewing is appropriate to be reported within the discussion. Future research is to consider data from female khat chewers as well.

Measuring the quality of life is a broad topic and it is impossible to form a definite conclusion as to why non-khat chewers had higher rates of physical and mental health compared with their khat chewing counterparts. Some chemical ingredients in khat might have caused chewers to report lower levels of quality of life. Another possibility is that users with a low quality of life use khat more frequently. These possibilities need to be explored further, and could be a focal point for future studies.