Monday, 22 July 2024

Fact checking the "Swedish Experience"

This post responds to recurring efforts by tobacco companies and their allies to promote the "Swedish Experience" as template for harm reduction based on oral nicotine delivery. 

Data from reliable sources are used fact check the industry's claims. They show that for most indicators Sweden has not out-performed Canada with respect to reducing smoking prevalence or reducing tobacco-related disease. The Swedish experience is one where twice as many people use nicotine... and smoking is not being phased out more quickly.

Pouch promoters say Sweden is a tobacco control success story

Tobacco companies promote Sweden's permissive approach to snus as a way to keep smokers happy (with nicotine), keep investors happy (with high sales and profits) and keep health regulators happy (with low levels of disease).  

From BAT Presentation, 2024

Examples of this sales pitch can be found in investor presentations (such as those earlier this spring by British American Tobacco and Philip Morris International), in messaging from allied agencies (such as Smokefree Sweden), in industry-friendly media and on social media.

These pro-pouch arguments boil down to two claims:
* Widespread snus use in Sweden has led to lower smoking rates
* Widespread snus use in Sweden has led to low rates of tobacco-caused disease 

This is not the first time that the Swedish Experience has been trotted out by industry and others to encourage a permissive approach for oral nicotine. The current reprise is being delivered in the context of concerns in the US about the rapid increase in the sales of Zyn, and among European initiatives to regulate the introduction of nicotine pouches and Canada's Health Minister's plans to restrict how these products are marketed. 

(Nor is it the first time that tobacco control agencies have pushed back against these efforts. The data provided below is an update to our 2007 analysis.)

FACT CHECK: IN RECENT DECADES, THE "SWEDISH EXPERIENCE" HAS PRODUCED WORSE OUTCOMES THAN CANADIAN TOBACCO CONTROL. 
(Although the situation in Sweden was better in the last century).

As detailed below:

Widespread snus use in Sweden has NOT led to lower smoking rates
  • Estimates of smoking rates in Sweden are generally not lower than those in Canada. 
  • Smoking rates in Sweden have fallen, but not at a faster rate than in Canada 
  • Twice as many Swedes use tobacco-industry nicotine products as do Canadians (30% vs 15%).
Widespread snus use in Sweden has NOT led to low rates of tobacco-caused disease 
  • Lung cancer death rates among men are lower in Sweden than Canada, but the situation for other major tobacco-related diseases is not better.
  • Tobacco-related deaths have fallen faster in Canada than in Sweden
  • Sweden's lower lung cancer rates are rooted in lower tobacco use in the last century (not an outcome of current snus use)
Smoking rates in Sweden are not lower than in Canada

For decades, the OECD has tracked daily smoking rates among men and women in its high-income member states. The OCED provides access to this data, and also presents it in  data visualizations.  

OECD Countries identified as having lower daily cigarette smoking rates in 2021 than Sweden (9.7%) include New Zealand (9.4%), the United States (8.8%), Canada (8.7%), Mexico (8.6%), Norway (8%), Costa Rica (7.8%) and Iceland (7%). The World Health Organizations' report on tobacco smoking rates similarly shows Sweden behind Canada (12.6% vs 12%, Table A1.2), although rates of smoking manufactured cigarettes are slightly higher in Canada (where roll-your-own tobacco is no longer common).

Smoking rates in Sweden have fallen, but not at a faster rate than in Canada 

The Global Burden of Disease group and the Institute for Health Metrics and Evaluation provide historic smoking prevalence data which has been age-standardized to allow comparisons among countries and across time periods.(The data are available here). 

Their data shows that Sweden has done well (ranking 12th among 204 countries)  - but that other countries, including Canada (ranked 9th) and Australia (ranked 7th) have done better. The country with the greatest progress was Brazil, where the rate of oral tobacco use is very low and e-cigarettes are not legal for sale.  



Twice as many Swedes use tobacco-industry nicotine products as do Canadians 

There are 4 main categories of nicotine products sold by tobacco companies:  cigarettes and cigarette tobacco, electronic cigarettes/vaping products, heated tobacco products and nicotine pouches. 

There are currently  no global databases with information on the overall use of nicotine through these products, but national surveys can be used for this purpose. 

Results from these two surveys show similar rates for past month use of cigarettes (12% in Sweden, 11% in Canada), and not very different rates for vaping (4% in Sweden to 6% in Canada). One in 5 Swedes used either snus or non-tobacco nicotine pouches (20%) compared with fewer than 1% of Canadians. Because of snus use, twice as many Swedes use nicotine as do Canadians (30% vs. 15%).

(A data sheet with the information presented below as well as for men and women can be downloaded here).

Lung cancer rates among men are lower in Sweden than in Canada, but the situation for other major tobacco-related diseases is no better.

The World Health Organization provides information on the number and rate of deaths from specific diseases, and also provides data files which make age-adjustments to this data to allow for comparison between countries. 

Using only countries for which WHO signals the data is robust enough for comparison, Sweden is in the second-best quintile for overall deaths from lung cancer and Canada is in the worst. 

For lung cancer rates among men, Sweden is in the second-best quintile for men, but in the second-worst for women. Canada is in the middle quintile for men but the worst for women. 

The 20 countries for which comparable data is available and which have lower age-standardized lung-cancer rates than Sweden for the whole population are (in decreasing order):  Jamaica,  Venezuela, Kyrgyzstan, Philippines, Brazil, Chile, Trinidad-Tobacco, Belize, Grenada, Mauritius, Colombia, Saint Lucia, Barbados, Kuwait, Panama, Mexico, Antigua-Barbuda, Nicaragua, St Vincent and the Grenadines and Costa Rica.

(A data sheet with lung cancer rates extracted from WHO's files can be downloaded here). 

For deaths from other major tobacco-related diseases (e.g. Ischaemic Heart Disease and COPD), Sweden does worse or not much better than Canada. 

The same age-standardized World Health Organization data allow for comparisons with the number of deaths from other tobacco-caused diseases in Sweden and other countries.

With respect to heart disease, Swedish men and women both experience higher death rates than in Canada. In 2019, there were an estimated 6 more deaths from heart disease in Sweden for every 100,000 people than there were in Canada. For Chronic Obstructive Pulmonary Disease, Canada's rates for men are slightly higher than those for Sweden, and those for women are slightly lower: overall Canada had one more death from COPD for 100,000 people than Sweden. Public health agencies in Sweden have also pointed out that tobacco mortality in Sweden is not the lowest in Europe.

The data shown below is available on a downloadable data sheet; Data visualizations are available on Our World in Data. 


Tobacco-related deaths have fallen faster in Canada than in Sweden


Their estimates for Canada and Sweden (downloadable here) show similar results: 1 in 5.6 deaths in Canada is due tobacco, compared with 1 in 6.6 deaths in Sweden. When comparing progress in age-standardized deaths from tobacco between 1990 and 2021, Canada was in a somewhat better condition than Sweden:  the overall rate of deaths from tobacco-caused deaths fell by 15.6% in Canada, compared with 12.3% in Sweden. 
 

Sweden's lower lung cancer rates are rooted in lower tobacco use in the 20th century - not today's nicotine pouches


Lung cancer can occur many decades after tobacco use, and current lung-cancer rates are an indicator of tobacco use in previous periods. One likely reason that Sweden has lower rates of lung cancer now than Canada does is because much less tobacco was smoked in the last century.

Reports on historic tobacco sales produced by PN Lee and colleagues show that over the past century per capita tobacco sales (including cigarettes, cigars, pipe and snus) remained constant in Sweden. By contrasts, per capita consumption in Canada rose dramatically in the middle of the past century (to more than double the rate in Sweden) before falling.

Grams of tobacco consumed per day per adult.
From P.N. Lee, International Smoking Statistics
for Sweden and Canada 


In living memory, most of the tobacco consumed in Canada has been in the form of manufactured or roll-your-own cigarettes. By contrast, snus has been the predominant product in Sweden for 30 years. 


There are many reasons other than the presence of snus that may explain why cigarette smoking never emerged with the ferocity in Sweden as it did in Canada and other European countries. Among these are differing experiences with returning soldiers (Sweden was a neutral country during World War Two), control of the tobacco market by a government-owned company,  and the absence of television advertising in Sweden. Although Sweden is not now looked on as a leader in tobacco control, during the 1970s and 1980s it pioneered several tobacco control measures, including advertising restrictions (1979) and rotating health warnings (1987).