The 2004 Progress Report on Tobacco Control

Tracking Key Indicators

Effective tobacco control requires partners to work together towards common goals. It requires integration, coordination and complementarity of a diverse array of strategies. The federal / provincial / territorial governments and tobacco control NGOs play a critical leadership role in achieving national action. This array of strategies, including research, policy and programme components, developed and coordinated at the local, provincial / territorial, national and international levels, is needed to successfully reduce tobacco consumption. When each jurisdiction and organization with an interest in tobacco control is integrated within an overall action plan, then Canada can be seen to be implementing an effective, comprehensive strategy.

Smoking prevalence in Canada

According to the latest results from the Canadian Tobacco Use Monitoring Survey (CTUMS), for data collected between February and December 2003, over 5 million people, representing roughly 21% of the population aged 15 years and older, were current smokers. Of the general population aged 15 years and older, 17% reported smoking daily.

Sources: 1986, Labour Force Survey Supplement; 1991, General Social Survey; 1994, Survey on Smoking in Canada; 1996/97, National Population Health Survey; 1999-2003, Canadian Tobacco Use Monitoring Survey (Annual).

Approximately 23% of men were current smokers, higher than the
proportion of women (18%).

Sources: 1986, Labour Force Survey Supplement; 1991, General Social Survey; 1994, Survey on Smoking in Canada; 1996/97, National Population Health Survey; 1999-2003, Canadian Tobacco Use Monitoring Survey (Annual).

The decline in smoking among youth aged 15-19 years continued in 2003 to 18%, with 12% reporting daily smoking and 7% occasional smoking. This is a decrease from 22% in 2002 and is a ten percentage point improvement from 28% in 1999, when CTUMS was first conducted. Slightly more teen girls reported smoking than boys (20% vs. 17%). However, among daily smokers, boys smoke slightly more cigarettes per day (13.0) than girls (11.7).

Sources: 1986, Labour Force Survey Supplement; 1991, General Social Survey; 1994, Survey on Smoking in Canada; 1996/97, National Population Health Survey; 1999-2003, Canadian Tobacco Use Monitoring Survey (Annual).

The prevalence of smoking among young adults aged 20-24 years was reported at 30% for 2003 (21% daily, 9% occasionally), also slightly decreased from the 2002 rate (31%). There is a slight difference in the daily smoking rates between males (23%) and females (19%) aged 20-24 years. Overall, young adults continue to report the highest rate of smoking.

All provinces have witnessed a drop in their smoking rates compared to 1999. In 2003, British Columbia continued to report the lowest rate of current smokers (16%) and Québec had the highest (25%). Québec reported the highest average number of cigarettes consumed per day by daily smokers (16.8), while Alberta reported the lowest average (14.6).

Sources: 1991, General Social Survey; 2003, Canadian Tobacco Use Monitoring Survey (Annual).

Source: 2003, Canadian Tobacco Use Monitoring Survey (Annual).

Tobacco Consumption in Canada

Not only are fewer Canadians smoking, but they are also smoking fewer cigarettes on a daily basis. In 1985, daily smokers consumed an average of 20.6 cigarettes per day. Since then, the number of cigarettes smoked has been gradually declining to the current level of 15.9 cigarettes per day reported for 2003. Men continued to smoke more cigarettes than women: 17.3 cigarettes per day for males as compared to 14.0 for females. The majority of Canadian smokers (58%) reported consuming some type of 'light' or 'mild' cigarette, compared to 42% who smoke a 'regular' type of cigarette.

Sources: 1985 & 1990, Health Promotion Survey; 1996/97, National Population Health Survey; 1999-2003, Canadian Tobacco Use Monitoring Survey (Annual). * Provincial data only

Sources: 1985 & 1990, Health Promotion Survey; 1996/97, National Population Health Survey; 1999-2003, Canadian Tobacco Use
Monitoring Survey (Annual). * Provincial data only

Tobacco Use, Health Determinants & Health Disparities

The determinants of health approach provides a useful framework for examining tobacco use. Determinants of health (factors that interact in complex ways to impact on people's health) include: income and social status; gender; culture; education; and biological and genetic traits. In consideration of continuing disparities in health trends in Canada between males and females, tobacco control experts are investigating gender (the culturally specific set of characteristics that identifies the social behaviour of men and women and the relationship between them) and sex (the biological differences between men and women) factors in tobacco use.

Trends in Smoking Prevalence

Historically, the rate of smoking has been higher among men than it has been among women. Over the last few decades, however, the decline in smoking prevalence has been more pronounced among men. From 1965 to 2003, the rate for men has dropped from 61% to 23%, as compared with 38% to 18% for women during the same period.

The difference in the rate of decline is likely the result of two phenomena: (1) men generally adopted cigarette smoking before women did, and in turn quit smoking earlier relative to women; and (2) a potential peak in female smokers was likely suppressed during the era when more intense tobacco control activities were implemented (e.g. the decline in female smokers was less because the rise in female smokers was less).

Health Consequences of Smoking

While tobacco use has similar health consequences for men and women (i.e., lung and other cancers, cardiovascular disease, stroke, chronic bronchitis and emphysema), some of its effects are sex specific. Men who smoke are at risk for erectile dysfunction and reduced fertility, while women are at risk for increased cardiovascular disease while using oral contraceptives, reduced fertility, cervical cancer, early menopause and bone fractures. Moreover, smoking during pregnancy can result in premature birth, malformation of the fetus, low birth-weight and stillbirth.

Smoking-related Deaths

The number of smoking-related deaths in Canada increased between 1989 and 1998; this increase was steeper among women. During this time period smoking-related deaths rose from 38,357 to 47,581-representing an increase of 9,224-with females accounting for 6,531of the increase.

This sex difference in smoking-related deaths reflects the smoking behaviour of the population two to three decades earlier. The effect of the drop in tobacco consumption among men, beginning in the mid-1960s, is reflected by a levelling off in the mid-1980s and then a continuous decline in male lung cancer rates. (Lung cancer was the leading cause of smoking-related deaths between 1989 and 1998; cardiovascular disease is now the major cause of smoking related death.) In contrast, women's smoking rates peaked in the late 1970s and declined slightly over the past three decades. Female lung cancer mortality rates more than quadrupled between 1969 and 1998, and can be expected to rise for the next few years before decreasing.

The significant decline in the rate of smoking over the past 40 years will soon be reflected in the rate of smoking-related deaths. While the male smoking-related death rate has already peaked (1988-1989), the rate for females will probably continue to rise, but much more slowly. In addition, it is expected that the rate for lung cancers in females will peak at a level lower than that for males. One of the more likely factors behind this trend is the implementation of tobacco control measures during the 1970s, 1980s, and 1990s.

Tobacco Sales vs. Prevalence Trends

In 2003, in order to determine the impact of changes in price on tobacco consumption and smoking prevalence, Health Canada collected and analysed relevant data (including monthly sales, reported annual prevalence rates, tobacco price indices, surveillance data, etc.) from across Canada for the period between January 1999 and December 2002.

Price Elasticity

Between January 1999 and December 2002 the average price of tobacco products rose by 75% nationally. Provincially, the increase in price ranged from 35% in Newfoundland to 85% in Ontario. During the same period, annual wholesale sales of cigarettes fell by 16% nationally, with a range between 7% (B.C.) and 27% (Quebec).

The point price elasticity of demand (the ratio of percent change in sales divided by the percent change in price, taken between two time periods) at the national level was -0.22, with a range between -0.16 (Ontario) and -0.32 (Quebec). (It must be borne in mind that the point price elasticity of participation is affected by those who quit because of price and those who quit for other reasons [this will be discussed later]).

Prevalence

Analysis of prevalence rates was problematic. The Canadian Tobacco Use Monitoring Survey (CTUMS) provided timely, reliable and continual data on tobacco use and related issues between 1999 and 2002, on both semi-annual and annual bases. Analysis of the annual data for the population aged 15 and older indicated that the differences in prevalence rates between 1999 and 2002 were statistically significant at the national level (25% to 21%).

The point price elasticity of participation for each group was -0.21 (Canada), -0.25 (N.B.), and -0.29 (Sask). (Again, the point price elasticity of participation is affected by those who quit because of price, those who quit for other reasons, and those who did not take up smoking because of increased price.)

With respect to those aged 15-19, nationally, prevalence rates fell from 28% in 1999 to 22% in 2002. This is a decline of 21% (six percentage points) and is statistically significant. With respect to youth aged 15-19, the point price elasticity of participation was -0.27. Since provincial prevalence rates are highly variable, analysis of provincial point price elasticities of participation was not performed.

Sales

What the data imply is that, at the national level, for every 10% increase in price, sales declined by 2.2%. The decline in sales results from a combination of reductions in the number of cigarettes smoked by remaining smokers and reductions in the number of smokers. The data imply that, for a 10% increase in price, there will be a 2.1% decrease in prevalence for smokers aged 15 and older. This does not mean that the change in prevalence is 2.2 percentage points (25% to 23%); it means that the difference between the starting and ending prevalences is 2.2% (25% to 24.5%).

Cessation

With respect to quitting, CTUMS indicated that, in 2002, there were approximately 700,000 fewer smokers than in 1999. This number must be interpreted with caution, since smokers and former smokers continue to change their smoking habits on a regular basis.

The CTUMS data provided some insight into motivations for quitting. When asked, "What would it take to quit smoking?", 34% of smokers responded "more willpower". This is contrasted with the 4% who responded "more effective / affordable stop-smoking programmes / cigarette substitutes" and the 6% who responded "increased cost / difficulty affording cigarettes".

This was confirmed in a recent survey of 300 recent quitters, only 10% of whom spontaneously suggested that "cost" was their reason for quitting. With respect to youth aged 15-19, less than 1% responded "increased cost / difficulty affording cigarettes". What this means is that the point price elasticity of participation may be a combination of a price component, approximately 10%, and other causes, 90%.

Analysis Issues

There are other analyses that can be performed, including those for sensitivity, but there are a number of serious data gaps which result from the information not being collected or not being statistically significant. For example, if a province implements a subsidy programme for nicotine replacement therapies, there is currently no national survey that will collect the information necessary to evaluate this initiative. Indeed, CTUMS indicates that only two provinces have statistically significant reductions in adult prevalence rates. As a result, at present, even provincial strategies cannot be evaluated using CTUMS.

To alleviate many problems inherent in evaluation around the issue of price elasticity of demand it is suggested that, wherever and whenever possible, initiatives be evaluated at the lowest level of implementation. For example, if a province implements a strategy targeted towards youth aged 15-19, then the strategy should be evaluated against its objectives and within the target group.

Broader Issues

With respect to the impact of various strategies and policies (taxes, mass media campaigns, cessation campaigns, regulations, etc.), evaluation and analysis continues to be difficult. In some cases, such as with respect to tax policy, change is instantaneous, given that smokers react as soon as the price goes up or down. The metric used to measure this strategy is sales, and a minimum of three months is required for an impact to be shown.

In other cases, for example with respect to cessation or mass media campaigns, the metrics are numbers of smokers and impacts of the campaigns. Often we use campaign recall or intent to quit as measures of success. While these metrics can be measured within weeks of a campaign's conclusion, they are not wholly accurate measures of actual behaviour. More accurate measures of behaviour are based upon phenomena such as smoking prevalence and numbers of attempts to quit. Moreover, these metrics are usually taken long after cessation or mass media campaigns have concluded (one or more years), when it is difficult to determine whether campaigns have a direct or indirect effect.

Confounding all of this is the fact that interventions occur in groups. Taxes are sometimes raised concurrently with the introduction of a regulation and a new cessation campaign. The net result may be a decline in smokers, generally years after the interventions, when it is next to impossible to attribute which intervention or combination of interventions contributed in what way to the decline.