Anwar T. Merchant ( Department of Community, Health Sciences, Aga Khan University, Karachi. )
Stephen P. Luby ( Department of Community, Health Sciences, Aga Khan University, Karachi. )
Ghazala Perveen ( Department of Community, Health Sciences, Aga Khan University, Karachi. )
Approximately half of the families in Pakistan reported having atl east one smoker in 1992. Smokers were less educated, poorer and more likely to come from a rural background than non-smokers. The proportion of family earnings in poor families with marginal incomes, may be substantial. We conducted this survey to determine the prevalence of cigarette smoking in males over 15 years of age in Azam Basti, an urban squatter settlement of Karachi where 31% of the children less than 5 years old were malnourished. A pretested, structured questionnaire was administered to males aged fifteen years and above, from randomly selected homes in Azam Basti. In our sample of 102 persons the respondents were 38 years old on average, and earned about Rs.4,500 (US$130) per month. Persons with 10 or more years of education were thrice as likely to have never smoked as compared to those with less than ten years of schooling, (OR=3.6, 95% confidence interval 1.2,11.4). Current smokers were more than twice as likely as non-smokers to have monthly household incomes less than Rs.5000 (US $ 140) (OR= 2.4, 95% confidence interval 0.8, 7.3). Smoking is common in urban squatter settlements in Pakistan. Parental smoking and its relationship with malnutrition in children under five is not well documented or publicized, even though there is evidence that it has a contribution. We propose that primary health care programs consider smoking prevention and cessation as community based interventions (JPMA 48:62, 1998).
Smoking is on the increase in developing countries1 and Pakistan is no exception2. The Pakistan Health Education Survey estimated that approximately 55% of Pakistani families reported that atleast one member smoked3. The Ministry of Health, Government of Pakistan conducted the survey between October, 1991 and November, 1993, taking a representative sample froni all the households in the country3, the primary source of information being mothers with atleast one child aged two years or younger. Families living in rural areas were twice as likely to report of having a smoker admist them as families in urban areas. Individuals with no education were three times as likely to report having a smoking family member as compared to individuals with atleast a secondary school education. Smoking steadily decreased with increasing income3. We conducted a survey to determine the prevalence of cigarette smoking in males over 15 years of age in a low income area of Karachi, the demographic profile of smokers, details of money spent. and their knowledge related to smoking.
A list of all households in Azam Basti, an urban squatter settlement situated in Karachi South was procured from the AgaKhanUnviersity Primary Health Care Center. The total of the households listed was divided by ten to obtain ten clusters. A pair of interviewers went into each cluster. The first house to be visited in the cluster was selected by a random number obtained fmm a ten rupee note in the interviewers pocket. Subsequent houses were selected by knocking at the nearest adjacent door. The questionnaire was administered to males aged fifteen years and above, on a working day between four and seven o’clock in the evening in April 1994. If a household had more than one eligible subject, the one to be interviewed was selected by lottery. If a household did not have an eligible subject it was eliminated and the next house visited. There were no revisits. A pretested, structured questionnaire was used. Smokers were labelled as persons who had smoked at least once per day for in the preceding month. Those who had smoked at least once per day for one month but currently did not smoke were ex-smokers and those who had never smoked daily for one month, never smokers. The data was entered and analyzed on Epi Info 6.01 . We did an ANOVA to compare continuous data, X2 to compare discrete data, and odds ratios using never smokers as the reference gmup.
One hundred and twenty households were visited. In 18, the house was locked or there were no eligible subjects present leaving a sample of 102 persons. Nobody refused the interview. The mean age of the respondents was 38 years and monthly earnings of Rs.4,500 (approximately US$ 130 at the then current exchange rate of Rs,35=US$ 1). The average age of the men was 19 years, when they became regular smokers
Table I describes men who were smokers, ex-smokers and never smokers. Ex-smokers were older (48 versus 37 years p<0.0l) than smokers; they were more likely to report suffering from a medical problem than never smokers (25% versus 23%) but this may be a chance effect(p=0. 19). Smokers and ex- smokers reported smoking for a mean of 20 and 18 years respectively (P=0.74). Smokers consumed on an average 16 cigarettes per day and spent Rs. 12 per day, or 8% of the total reported household income, on the habit. Ex-smokers reported smoking on the average 25 cigarettes per day, and spending on average Rs.37 (US $ 1.1) daily, or 25% of the total household income, on the habit. In their lifetimes smokers hadspentapproximately Rs. 113,840 (US $3250) and ex-smokers Rs.163,590 (US $ 4670) on the habit (p value, <0.00 1). People who had never smoked were less educated and had lower incomes than non-smokers. Indeed, persons with 10 ormore years of of schooling were 3 times more likely to have never smoked (3 3%) as compared to people with less than 10 years of schooling (12%) (Odds ratio [OR]=3.6, 95% confidence interval 1 .2, 11.4). Current smokers were more than twice as likely as non.smokers to have monthly household incomes less than Rs.5000 (73% versus 53% OR= 2.4. 95% confidence interval 0.8. 7.3. p-.07) Most of the men were well aware that smoking affected health arid the majontv recognized that it caused cancer. heart disease and affected the health of others living in their house (Table II).
Thers is a high prevalence of smoking among males in Azani Basti. a poor conuminity where about 30% of the children under 5 years of age are malnourished. About half the men interviewed reported to be either current or past smokers. Most smokers in our sample knew that smoking would advcrsclv affect their health, yet they continued with the habit. This is consistent with other studies show lug that knowledge of ill of smoking alone are not sufficient to restrain the habit5. Because the questionnaires were anony mous and we did not ascertain malnutrition in children under five yars in this study. we do not known the proportion of malnourished children in households with and without smokers. but there is evidence pointing to a relationship between smoking and malnourishment in children. Smoking is common in urban squatter settlements in Karachi. Parental smoking and its relationship with malnutrition in children under live is not well documented or publicized even though there is evidence that it has a contribution6,7. We propose that primary health care programs consider smoking pievention and cessation as community based interventions.
We are grateful to participants of the Regional Training Program who conducted the interviews and the Aga Khan Foundation who funded the program.
1. Mackay, J.L. The fight against to becco in developing countries. Tubercle Lung. Dis., 1994;75:8-24.
2. Ball, K. Pakistan .Attempts to control damage by tobacco smoking. LANCET., 1983;2:1413.
3. Pakistan Health Education Survey, 1991-92. 1993:115-122.
4. Dean, A.D., Dean, J.A and Burton, J.H. Epi Info, (Computer program) 1990, (Abstract).
5. Derman, U., Demir, G and Akan, P. Is awareness of its risk enough to stop people from smoking? J Cancer Edue., 1995;10:68-70.
6. Cohen, N Smoking, health and survival: Prospects in Bangladesh LANCET., 1981;1:1090-92.
7. Kebede, A. and Larson, C. The health consequences of intrauterine growht retardation in southwestern Ethiopia Trop. Doct., 1994;24:64-69.