Is tobacco use increasing among medical students of Bengaluru city, India? Evidence from two cross-sectional studies

Introduction

Globally, behaviours and lifestyles are in a state of flux and tobacco companies are shifting their focus to newer markets and products (such as e-cigarettes). ^[1] It is expected that increasing numbers of youth will begin to use tobacco products and that young women will also be encouraged to smoke. This is likely to alter the epidemiology of tobacco use in countries such as India currently characterized by a preponderance of smokeless tobacco use and beedi smoking rather than cigarette smoking as also predominant tobacco use during middle-age (late 40s) among adult males rather than in young adulthood (20s to 30s) as seen in industrialized countries. ^[2] Among youth, tobacco use among medical students is a key indicator of tobacco control activities in a country or region. It is therefore important to track tobacco use among medical students over time given their role as future doctors in tobacco cessation counselling and advocacy for stronger tobacco control policies. We estimated the self-reported prevalence of tobacco use, across a 5-year interval, among medical students of a southern Indian city.

Methods

Study setting and population

The studies were undertaken in Bengaluru, a metropolitan city in southern India. ^[3] The demographic characteristics of the city are: population 8.4 million, literacy rate 90% and sex-ratio 914 females per 1000 males. ^[4] The studies were conducted in four medical colleges (one government and three private). All third year (6th term) students were the eligible population. ^[5]

Study design, instrument and data collection

Two cross-sectional studies were done: the first survey was in 2007 ^[6] and the second in 2013. A self-administered questionnaire was used to obtain information on tobacco smoking and chewing as per the Global Health Professionals Study (GHPS) methodology. ^[5] Ethical approval was obtained from the Institutional Review Board of St John′s Medical College, Bengaluru. Permission to conduct the study was obtained from the heads of the four participating institutions. Written informed consent was obtained from the participating students. Questionnaires were distributed to the students in a classroom setting and collected after completion; anonymity was respected.

Study definitions and analysis

Definitions of tobacco use status were based on self-reported smoking of cigarettes/beedis and chewing of tobacco. Never-users were defined as persons who had never smoked or chewed tobacco even once in their life-time; experimental users were defined as those who had smoked or chewed tobacco once or twice but never regularly; current users were defined as those who had smoked or chewed tobacco in the last 30 days preceding the survey, either occasionally (at least once a week) or daily (at least once per day). ^[6] Data were entered and analysed using SPSS 21.0. Simple descriptive analysis was done for the variables of interest; a p value <0.05 was considered significant.

Results

The participation rates were 82% (323/395) in 2007 and 78% (253/324) in 2013 (p=0.2). In different colleges it ranged from 61% to 95% in 2007 and from 72% to 95% in 2013. The proportion of students from the government college was 40% in 2007 and 23% in 2013 (p<0.001). Male students comprised 54% of all participants in both surveys. Among males, experimental use of tobacco was reported by 24% (41/172) in 2007 and 42% (56/135) in 2013; this difference was statistically significant (p=0.001; [Figure - 1]). This appeared to be driven predominantly by a significant change in the private sector (24% v. 43% [p=0.01]) and a non-significant change in the government sector (23% v. 38% [p=0.1]). Current smoking increased from 3.5% (6/172) in 2007 to 8.9% (12/135) in 2013 (p=0.05); the maximum prevalence in a college was 16% in 2007 and 22% in 2013. There was no difference by type of college: change in government sector (1.1% v. 2.5% [p=0.5]) and change in private sector (6% v. 15.6% [p=0.2]). Daily smoking was reported by 1.7% in 2007 and 3.7% in 2013 (p=0.3).

Fgure 1. Changes in prevalence of reported smoking among male and female medical students in Bengaluru in 2007 and 2013

Among females, experimental use was reported by 3.3% (5/151) in 2007 and 11.2% (13/116) in 2013; this difference was statistically significant (p=0.01). There was no difference by type of college: change in government sector (2.2% v. 5.6% [p=0.5]) and change in private sector (5% v. 12.2% [p=0.08]). Current smoking was reported by ≤1% of female students in both the surveys.

Among males, current chewing habit was reported by 1.8% (2/171) and 3.7% (5/135) in the two years, respectively. Mixed users (smoking and chewing) were 1.2% in 2007 and 2.2% in 2013. None of these differences were statistically significant. Among females, current chewing was reported by 4% (6/146) in 2007 and by 0% in 2013 (p=0.04, [Figure - 2]).

Figure 2. Changes in prevalence of reported chewing (alone and with smoking) among male and female medical students in Bengaluru in 2007 and 2013

Discussion

Studies investigating secular changes in tobacco use are rare in India. While three data points may be more suggestive of a trend than the two data points that we have, our study is still useful in showing the overall direction of change. Smoking and chewing rates in Bengaluru were consistently lower than those reported from other colleges in India. ^[7],[8],[9] Our two separate cross-sectional studies have shown that while regular smoking or chewing habits among male and female students have remained the same, experimental use among male and female students has increased significantly over this time period. These observed differences could be real or due to selection bias (with non-respondents being systematically different from respondents) or information bias (due to tobacco use misclassification). ^[10],[11] Overall, the participation rate in our study was high and consistent with that observed in earlier studies; ^{[9],[12],[13]} further, it is suggested that non-response bias may be less of an issue in surveys of health professionals than of the general public. ^[14] We, therefore, believe that our findings provide a realistic assessment.

However, college student smokers are not a homogeneous group. Different categories have been identified earlier with different smoking trajectories ^[15] and different implications for nicotine dependence ^[16] and health and other outcomes. ^[17] It is unclear whether students who have experimented with smoking in the past, either in school or college, will transition to regular use in the future in the face of the efforts of the tobacco industry to reach new markets with the aid of multiple traditional gimmicks to lure youth as well as through the introduction of e-cigarettes. ^[1],[18] An increasing trend of smoking was noted in the east European region about a decade back ^[19],[20] following large tobacco companies looking for newer markets. Increased psychosocial risk, key socioeconomic factors, external environment and continuing lack of access to tobacco control education could all potentially lead to greater vulnerability among students who have experimented with smoking in the past.

While it is encouraging to know that there was no increase in current smoking or chewing among medical students in Bengaluru over the 5-year period, any tobacco use is still a cause for concern among doctors of tomorrow. It is therefore important for schools and colleges to increase access to tobacco control education so that doctors in training are better prepared to deal with the challenges of the future with regard to their own tobacco use as well as that of their patients.

Acknowledgements

The author is grateful to the deans of the participating institutions for granting permission and facilitating the conduct of the surveys. Students′ participation is also acknowledged. The participating institutions were St John′s Medical College, Bangalore Medical College, M.S. Ramaiah Medical College and Kempegowda Institute of Medical Sciences. The assistance of Anu John, Anita Nath, Smitha K. and Vivek Ramesh in data collection/management is also acknowledged.