Nicotine & Tobacco Research, Volume 16, Number 8 (August 2014) 1104–1111
Original Investigation
Differential Patterns of Tobacco Use Among Black Men and Women in Cape Town: The Cardiovascular Risk in Black South Africans Study Nasheeta Peer PhD1, Carl Lombard PhD2, Krisela Steyn MD3, Naomi Levitt MD3,4 1Non-communicable Diseases Research Unit, Medical Research Council, Durban, South Africa; 2Biostatistics Unit, Medical Research Council, Cape Town, South Africa; 3Department of Medicine, Chronic Disease Initiative for Africa, University of Cape Town, Cape Town, South Africa; 4Department of Medicine, Division of Endocrinology and Diabetes, University of Cape Town, Cape Town, South Africa
Corresponding Author: Nasheeta Peer, PhD, Non-communicable Diseases Research Unit, Medical Research Council, 491 Ridge Road, Overport, Durban 4001, South Africa. Telephone: 27 31 203 4816; Fax: 27 31 203 4836; E-mail: [email protected]
Abstract Introduction: To examine the prevalence and determinants of tobacco use in the 25–74-year-old urban Black population of Cape Town and to examine the changes between 1990 and 2008–2009 in the 25–64-year-old sample. Methods: In 2008–2009 (n = 1,099), a representative cross-sectional sample was randomly selected from the same townships sampled in 1990 (n = 986). Sociodemographic characteristics, tobacco use by the World Health Organization (WHO) STEPwise questionnaire, and psychosocial stress, including sense of coherence (SOC), locus of control, and adverse life events, were determined. Survey logistic regression analysis assessed the determinants of smoking ≥1 cigarette/day. Results: There were 392 men and 707 women. Age-standardized prevalence of smoking ≥1 cigarette/day was 48.5% (95% confidence interval [CI] = 43.0–54.0) in men and 7.8% (95% CI = 5.8–10.5) in women (p < .001). Prevalence in men was lower in 2008–2009 (51.0%, 95% CI = 45.2–56.7) compared with 1990 (59.7%, 95% CI = 53.8–65.4) but unchanged in women (2008/09: 8.0%, 95% CI = 5.9–10.7; 1990: 8.4%, 95% CI = 6.0–11.8). In the logistic model for men, smoking was associated with younger age (p = .005) and being poor (p = .024). In women, spending more than half their lives in the city (p < .001), being poor (p = .002), and coping poorly with stress (defined by lower SOC; OR = 1.04, 95% CI = 1.01–1.08; p = .035) were associated with smoking. Increasing number of adverse events, which replaced SOC in the same models, was significant for women (OR = 1.10, 95% CI = 1.01–1.21; p = .047) but not for men. Education level, employment status, and housing quality were not relevant for men or women. Conclusions: The high smoking prevalence in men and unchanged rate in women require additional interventions to curtail this behavior.
Introduction The colonization of the Cape in the 17th century was soon accompanied by the use of tobacco for trading with local tribes (British American Tobacco South Africa, 2010). This practice spread rapidly, and by the following century, Cape Town had become an economic and cultural hub of tobacco trade. In 1881, the foundation for organized cigarette commerce was established and likely contributed to the prevalent tobacco use in South Africa today. Worldwide, cigarette smoking has spread on a massive scale following the appearance of mass-manufactured cigarettes (Jha & Chaloupka, 1999). Industrialization, urbanization, and other processes of modernization including globalization have enabled the rapid uptake of cigarette smoking (Deland, Lewis,
& Taylor, 2000; Jha & Chaloupka, 1999), with the tobacco industry driving the development of the tobacco pandemic in the 20th century (Deland et al., 2000). The globalization of tobacco use has imposed a huge and growing public health burden worldwide with the threat to global health greater today than ever before (Pampel, 2008). In South Africa, cigarette smoking contributes to a large burden of preventable disease. Smoking accounted for 8.0%– 9.0% of mortality in ≥15-year-old adults in 2000, the third commonest cause, after unsafe sex/sexually transmitted disease and high blood pressure, among 17 risk factors evaluated (Groenewald et al., 2007). Although lung cancer had the largest attributable fraction due to smoking, cardiovascular diseases (CVDs) contributed the principal proportion of deaths caused by smoking. Recent data also showed that smoking attributable
doi:10.1093/ntr/ntu042 Advance Access publication April 1, 2014 © The Author 2014. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: [email protected].
1104
Downloaded from http://ntr.oxfordjournals.org/ at Mount Allison University on June 19, 2015
Received October 10, 2013; accepted February 19, 2014
Nicotine & Tobacco Research analysis of the pooled data, based on the assets that defined wealth, was used to develop an asset index (Filmer & Pritchett, 2001). Categories of relative wealth were created by dividing the range into tertiles with the lowest tertile representing the poorest participants. Tobacco use was determined by the World Health Organization (WHO) STEP-wise surveillance questionnaire (Bonita, deCourten, Dwyer, Jamrozik, & Winkelmann, 2002). Participants who smoked tobacco products, irrespective of the quantity smoked, were classified as “smoke daily or occasionally.” Smoking ≥1 cigarette/day characterized participants who smoked daily. The use of smokeless tobacco included snuff and chewed tobacco irrespective of tobacco products smoked. Three instruments were used to examine psychosocial stress. Antonovsky’s sense of coherence (SOC) scale, previously validated in South Africa (Botha, Du Plessis, Van Rooyen, & Wissing, 2002), consisted of 13 items measuring comprehensibility (cognitive), manageability (instrumental/behavioral), and meaningfulness (motivational; Eriksson & Lindstrom, 2005). A low SOC inferred a poor ability to cope with stressors (Agardh et al., 2003). The locus of control (LOC), which was measured using a set of six questions, determined an individual’s perceived sense of control over his/her environment and life. A low score construed poor perceived control and a high score good perceived control (Rosengren et al., 2004). The Brugha Life Events questionnaire comprised 12 questions relating to negative life events such as illness, death, financial, or marital difficulties, among other adversities, and their impact (Brugha, Bebbington, Tennant, & Hurry, 1985). Statistical Analysis
Materials and Methods Study Population and Sampling Procedure In 2008/09, a random sample of 25–74-year-old men and women in the predominantly Black townships of Langa, Guguletu, Crossroads, Nyanga, and Khayelitsha in Cape Town participated in this cross-sectional study. These areas were selected to ensure comparability with a 1990 study, the methodology of which has been described elsewhere (Steyn et al., 1991). The sampling procedure for this study included a threestage cluster sampling described in detail previously (Peer et al., 2012). The pre-specified age and gender quotas included disproportionate sampling across age groups to ensure at least 50 men and women in each gender category. Individuals on tuberculosis or antiretroviral therapy or those who had received cancer treatment within the previous year were excluded because these conditions or their treatments may predispose to CVD risk factors, particularly diabetes and dyslipidemia, which were also assessed in this study. Data Collection and Classification Structured questionnaires were used to collect sociodemographic data such as education level, employment status, housing type, and assets defining wealth. Ownership of consumer items (durable goods), dwelling characteristics (wall and flooring materials), and the source of drinking water and toilet facilities were among the assets recorded. A principal component
Data analyses were done using Stata 11. Descriptive statistics, including crude prevalence, were calculated using the weights based on the sample design and adjusted for the realized sample. The age-standardized smoking prevalence was calculated using the WHO World Population as the standard (Ahmad et al., 2001). In the univariate analyses, the categorical data (tobacco use and sociodemographic variables) are presented as percentages and 95% confidence intervals (CI), and the continuous data (psychosocial scores) as mean values and SD. Survey-based odds ratio (OR) and 95% CI for the associations of the sociodemographic variables with smoking ≥1 cigarette/ day were calculated (unadjusted). Survey multiple logistic regression analysis determined the independent associations of the sociodemographic and psychosocial variables with smoking ≥1 cigarette/day (includes three smokers who smoked piped tobacco but not cigarettes). The psychosocial measures of SOC, LOC, and adverse life events were modeled independently as continuous variables. Genderspecific models were conducted because of the differences in associations. The models with SOC for men and women are presented in Table 3. The data for LOC and adverse life events when they independently replaced SOC in the same models are presented in the same table and changes in the direction or significance of the other variables are noted. To assess the relationship between these variables and smoking intensity, that is, the number of units of tobacco smoked/ day, survey-based zero-inflated negative binomial models of the total number of cigarettes and pipes smoked daily were also conducted. These models have two components: (a) the risk
1105
Downloaded from http://ntr.oxfordjournals.org/ at Mount Allison University on June 19, 2015
mortality was significant in 35–74-year-old Black men (7.7%), while the burden was much lower in their female counterparts (2.0%; Sitas et al., 2013). One of the most effective population-based strategies for CVD and other noncommunicable disease prevention is a comprehensive tobacco control programme. South Africa has been a frontrunner in this regard with the introduction of comprehensive legislation in the form of the Tobacco Products Control Act of 1993 (Government Gazette, 1993; Maredza, Hofman, & Tollman, 2011). Although it cannot definitely be attributed to the legislation, there has been a decrease and stabilization in the number of smoking-related CVD deaths, consistent with a declining smoking trend (Maredza et al., 2011). Nonetheless, the prevalence of tobacco use remains high, particularly among men, with South Africa having among the highest rates in Africa (Peer, Bradshaw, Laubscher, & Steyn, 2009). This underscores the need for on-going surveillance to determine the effectiveness of current tobacco control initiatives and to identify vulnerable groups in need of specific smoking prevention and cessation programmes. On a national level, the most recent surveillance data on tobacco use were obtained in 2010 (Ayo-Yusuf & Olutola, 2013), but on a regional level, these have not been examined in the urban Black population of Cape Town in almost two decades. The Cardiovascular Risk in Black South Africans study aimed to determine the prevalence of tobacco use in urban Black men and women in Cape Town in 2008/09 and to compare these findings with a similar study conducted in 1990. Additionally, this study examined the determinants (sociodemographic and psychosocial stress) of smoking ≥1 cigarette/ day in 2008/09.
Smoking prevalence in Cape Town
Results The realized study sample consisted of 1,099 participants with 392 men and 707 women and represented 64% and 108% of the planned sample, respectively. Seventeen study participants
Table 1. Associations of Sociodemographic Factors With Smoking ≥1 Cigarette/Day Among Men and Women in 2008–2009 Men
Women 95% CI
Smoking ≥1 cigarette/day Age (years) 25–34 35–44 45–54 55–64 65–74 Education (years) ≤7 >7 % of life spent in the city
Original Investigation
Differential Patterns of Tobacco Use Among Black Men and Women in Cape Town: The Cardiovascular Risk in Black South Africans Study Nasheeta Peer PhD1, Carl Lombard PhD2, Krisela Steyn MD3, Naomi Levitt MD3,4 1Non-communicable Diseases Research Unit, Medical Research Council, Durban, South Africa; 2Biostatistics Unit, Medical Research Council, Cape Town, South Africa; 3Department of Medicine, Chronic Disease Initiative for Africa, University of Cape Town, Cape Town, South Africa; 4Department of Medicine, Division of Endocrinology and Diabetes, University of Cape Town, Cape Town, South Africa
Corresponding Author: Nasheeta Peer, PhD, Non-communicable Diseases Research Unit, Medical Research Council, 491 Ridge Road, Overport, Durban 4001, South Africa. Telephone: 27 31 203 4816; Fax: 27 31 203 4836; E-mail: [email protected]
Abstract Introduction: To examine the prevalence and determinants of tobacco use in the 25–74-year-old urban Black population of Cape Town and to examine the changes between 1990 and 2008–2009 in the 25–64-year-old sample. Methods: In 2008–2009 (n = 1,099), a representative cross-sectional sample was randomly selected from the same townships sampled in 1990 (n = 986). Sociodemographic characteristics, tobacco use by the World Health Organization (WHO) STEPwise questionnaire, and psychosocial stress, including sense of coherence (SOC), locus of control, and adverse life events, were determined. Survey logistic regression analysis assessed the determinants of smoking ≥1 cigarette/day. Results: There were 392 men and 707 women. Age-standardized prevalence of smoking ≥1 cigarette/day was 48.5% (95% confidence interval [CI] = 43.0–54.0) in men and 7.8% (95% CI = 5.8–10.5) in women (p < .001). Prevalence in men was lower in 2008–2009 (51.0%, 95% CI = 45.2–56.7) compared with 1990 (59.7%, 95% CI = 53.8–65.4) but unchanged in women (2008/09: 8.0%, 95% CI = 5.9–10.7; 1990: 8.4%, 95% CI = 6.0–11.8). In the logistic model for men, smoking was associated with younger age (p = .005) and being poor (p = .024). In women, spending more than half their lives in the city (p < .001), being poor (p = .002), and coping poorly with stress (defined by lower SOC; OR = 1.04, 95% CI = 1.01–1.08; p = .035) were associated with smoking. Increasing number of adverse events, which replaced SOC in the same models, was significant for women (OR = 1.10, 95% CI = 1.01–1.21; p = .047) but not for men. Education level, employment status, and housing quality were not relevant for men or women. Conclusions: The high smoking prevalence in men and unchanged rate in women require additional interventions to curtail this behavior.
Introduction The colonization of the Cape in the 17th century was soon accompanied by the use of tobacco for trading with local tribes (British American Tobacco South Africa, 2010). This practice spread rapidly, and by the following century, Cape Town had become an economic and cultural hub of tobacco trade. In 1881, the foundation for organized cigarette commerce was established and likely contributed to the prevalent tobacco use in South Africa today. Worldwide, cigarette smoking has spread on a massive scale following the appearance of mass-manufactured cigarettes (Jha & Chaloupka, 1999). Industrialization, urbanization, and other processes of modernization including globalization have enabled the rapid uptake of cigarette smoking (Deland, Lewis,
& Taylor, 2000; Jha & Chaloupka, 1999), with the tobacco industry driving the development of the tobacco pandemic in the 20th century (Deland et al., 2000). The globalization of tobacco use has imposed a huge and growing public health burden worldwide with the threat to global health greater today than ever before (Pampel, 2008). In South Africa, cigarette smoking contributes to a large burden of preventable disease. Smoking accounted for 8.0%– 9.0% of mortality in ≥15-year-old adults in 2000, the third commonest cause, after unsafe sex/sexually transmitted disease and high blood pressure, among 17 risk factors evaluated (Groenewald et al., 2007). Although lung cancer had the largest attributable fraction due to smoking, cardiovascular diseases (CVDs) contributed the principal proportion of deaths caused by smoking. Recent data also showed that smoking attributable
doi:10.1093/ntr/ntu042 Advance Access publication April 1, 2014 © The Author 2014. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: [email protected].
1104
Downloaded from http://ntr.oxfordjournals.org/ at Mount Allison University on June 19, 2015
Received October 10, 2013; accepted February 19, 2014
Nicotine & Tobacco Research analysis of the pooled data, based on the assets that defined wealth, was used to develop an asset index (Filmer & Pritchett, 2001). Categories of relative wealth were created by dividing the range into tertiles with the lowest tertile representing the poorest participants. Tobacco use was determined by the World Health Organization (WHO) STEP-wise surveillance questionnaire (Bonita, deCourten, Dwyer, Jamrozik, & Winkelmann, 2002). Participants who smoked tobacco products, irrespective of the quantity smoked, were classified as “smoke daily or occasionally.” Smoking ≥1 cigarette/day characterized participants who smoked daily. The use of smokeless tobacco included snuff and chewed tobacco irrespective of tobacco products smoked. Three instruments were used to examine psychosocial stress. Antonovsky’s sense of coherence (SOC) scale, previously validated in South Africa (Botha, Du Plessis, Van Rooyen, & Wissing, 2002), consisted of 13 items measuring comprehensibility (cognitive), manageability (instrumental/behavioral), and meaningfulness (motivational; Eriksson & Lindstrom, 2005). A low SOC inferred a poor ability to cope with stressors (Agardh et al., 2003). The locus of control (LOC), which was measured using a set of six questions, determined an individual’s perceived sense of control over his/her environment and life. A low score construed poor perceived control and a high score good perceived control (Rosengren et al., 2004). The Brugha Life Events questionnaire comprised 12 questions relating to negative life events such as illness, death, financial, or marital difficulties, among other adversities, and their impact (Brugha, Bebbington, Tennant, & Hurry, 1985). Statistical Analysis
Materials and Methods Study Population and Sampling Procedure In 2008/09, a random sample of 25–74-year-old men and women in the predominantly Black townships of Langa, Guguletu, Crossroads, Nyanga, and Khayelitsha in Cape Town participated in this cross-sectional study. These areas were selected to ensure comparability with a 1990 study, the methodology of which has been described elsewhere (Steyn et al., 1991). The sampling procedure for this study included a threestage cluster sampling described in detail previously (Peer et al., 2012). The pre-specified age and gender quotas included disproportionate sampling across age groups to ensure at least 50 men and women in each gender category. Individuals on tuberculosis or antiretroviral therapy or those who had received cancer treatment within the previous year were excluded because these conditions or their treatments may predispose to CVD risk factors, particularly diabetes and dyslipidemia, which were also assessed in this study. Data Collection and Classification Structured questionnaires were used to collect sociodemographic data such as education level, employment status, housing type, and assets defining wealth. Ownership of consumer items (durable goods), dwelling characteristics (wall and flooring materials), and the source of drinking water and toilet facilities were among the assets recorded. A principal component
Data analyses were done using Stata 11. Descriptive statistics, including crude prevalence, were calculated using the weights based on the sample design and adjusted for the realized sample. The age-standardized smoking prevalence was calculated using the WHO World Population as the standard (Ahmad et al., 2001). In the univariate analyses, the categorical data (tobacco use and sociodemographic variables) are presented as percentages and 95% confidence intervals (CI), and the continuous data (psychosocial scores) as mean values and SD. Survey-based odds ratio (OR) and 95% CI for the associations of the sociodemographic variables with smoking ≥1 cigarette/ day were calculated (unadjusted). Survey multiple logistic regression analysis determined the independent associations of the sociodemographic and psychosocial variables with smoking ≥1 cigarette/day (includes three smokers who smoked piped tobacco but not cigarettes). The psychosocial measures of SOC, LOC, and adverse life events were modeled independently as continuous variables. Genderspecific models were conducted because of the differences in associations. The models with SOC for men and women are presented in Table 3. The data for LOC and adverse life events when they independently replaced SOC in the same models are presented in the same table and changes in the direction or significance of the other variables are noted. To assess the relationship between these variables and smoking intensity, that is, the number of units of tobacco smoked/ day, survey-based zero-inflated negative binomial models of the total number of cigarettes and pipes smoked daily were also conducted. These models have two components: (a) the risk
1105
Downloaded from http://ntr.oxfordjournals.org/ at Mount Allison University on June 19, 2015
mortality was significant in 35–74-year-old Black men (7.7%), while the burden was much lower in their female counterparts (2.0%; Sitas et al., 2013). One of the most effective population-based strategies for CVD and other noncommunicable disease prevention is a comprehensive tobacco control programme. South Africa has been a frontrunner in this regard with the introduction of comprehensive legislation in the form of the Tobacco Products Control Act of 1993 (Government Gazette, 1993; Maredza, Hofman, & Tollman, 2011). Although it cannot definitely be attributed to the legislation, there has been a decrease and stabilization in the number of smoking-related CVD deaths, consistent with a declining smoking trend (Maredza et al., 2011). Nonetheless, the prevalence of tobacco use remains high, particularly among men, with South Africa having among the highest rates in Africa (Peer, Bradshaw, Laubscher, & Steyn, 2009). This underscores the need for on-going surveillance to determine the effectiveness of current tobacco control initiatives and to identify vulnerable groups in need of specific smoking prevention and cessation programmes. On a national level, the most recent surveillance data on tobacco use were obtained in 2010 (Ayo-Yusuf & Olutola, 2013), but on a regional level, these have not been examined in the urban Black population of Cape Town in almost two decades. The Cardiovascular Risk in Black South Africans study aimed to determine the prevalence of tobacco use in urban Black men and women in Cape Town in 2008/09 and to compare these findings with a similar study conducted in 1990. Additionally, this study examined the determinants (sociodemographic and psychosocial stress) of smoking ≥1 cigarette/ day in 2008/09.
Smoking prevalence in Cape Town
Results The realized study sample consisted of 1,099 participants with 392 men and 707 women and represented 64% and 108% of the planned sample, respectively. Seventeen study participants
Table 1. Associations of Sociodemographic Factors With Smoking ≥1 Cigarette/Day Among Men and Women in 2008–2009 Men
Women 95% CI
Smoking ≥1 cigarette/day Age (years) 25–34 35–44 45–54 55–64 65–74 Education (years) ≤7 >7 % of life spent in the city
