Author Affiliation: School of Public Health, The University of Hong Kong, Pokfulam, Hong Kong.
Smoking kills by causing serious diseases, as confirmed by many systematic reviews and meta-analyses on specific diseases. The US Surgeon General Report does not include total mortality.1 The first meta-analysis on total mortality by Shavelle et al2 on 14 cohort studies in men showed that the relative risk (RR) in light, medium (10-15 cigarettes per day), and heavy smokers was 1.47, 2.02, and 2.38, respectively. In women, the RRs from 11 studies were similar (1.50, 2.02, and 2.66, respectively). The average age of the cohorts ranged from 30 to 65 years or older, with no significant trend of RR with age in either sex.2
In this issue of the Archives, the meta-analysis by Gellert et al3 on smoking and all-cause mortality focused on elderly people (≥60 years). Based on 17 cohort studies from 7 countries, the RR was 1.85 in current smokers. The RR for ages 60 to 69 years, 70 to 79 years, and 80 years or older was 1.94, 1.86, and 1.66, respectively.3 Note that RRs from meta-analyses could underestimate the risk, since many smokers in some populations, especially those from Asia, had not smoked as long as those in the United States and United Kingdom, and in all cohorts, current smokers at baseline are survivors.
The World Health Organization statement “Tobacco kills up to one in every two users”4 is derived from Peto,5(p938) who stated “about half of teenagers who keep smoking steadily will eventually be killed by tobacco (about a quarter in old age plus a quarter in middle age).” This is based on the translation of the RR of 2 in the British doctor cohort with 40 years of follow-up.6 This risk was underestimated, since results from the 50-year follow-up showed that if smoking started at a young age, 2 in 3 smokers will be killed by smoking.7
Relative risk is the ratio of the absolute risk of a disease in exposed people to the absolute risk of a disease in unexposed people. Translating that RR into an absolute risk in exposed people without knowing the absolute risk in unexposed people is only feasible for total mortality. This is based on the attributable fraction (AF) in the exposed (AF = [RR − 1]/RR).8 The AF in smokers of 0.5, based on an RR of 2, means that of all smokers who have died, half of the deaths are due to smoking. Because all smokers will eventually die, the denominator “all smoker deaths” can be replaced with “all current smokers,” and thus “1 in 2 smokers will be killed by smoking.”
The RRs reported in the study by Gellert et al3 can be translated into AFs of 48%, 46%, and 40% for the respective age groups. Hence, the statement that about 1 in 2 smokers will be killed by smoking is true even for elderly smokers. The smaller risk at 80 years or older is most likely due to survivor effect.
In the study by Gellert et al,3 the RR of 1.34 for former smokers vs 1.85 for current smokers means a risk reduction of 28% ([1.85-1.34]/1.85). The RR (% risk reduction) for ages 60 to 69 years, 70 to 79 years, and 80 years or older of 1.54 (21%), 1.36 (27%), and 1.27 (24%), respectively, showed substantial benefits even at very old age. Such benefits are underestimated because some former smokers might have quit because of ill health and hence would die more quickly than continuing smokers (reverse causality). Those who quit before age 40 years or well before avoid most of the excess risk.9
The absolute risk that 1 in 2 smokers will be killed by smoking should have a stronger impact than RR of total mortality or specific diseases or total number of tobacco deaths in a specific population or globally during a specific period in the past or future. To current smokers, the latter appear to be others' problems. The 2 in 3 absolute risk7 is more striking and is particularly relevant for young smokers.
Most smokers grossly underestimate their own risks. Many older smokers misbelieve that they are too old to quit or too old to benefit from quitting. Because of reverse causality and from seeing deaths of old friends who had quit recently, some misbelieve that quitting could be harmful. A simple, direct, strong, and evidence-based warning is needed. Clinically, brief interventions (1 to <20 minutes) by health care professionals, though effective,10 are not practicable in busy service settings. The absolute risk warning of 1 in 2 deaths only takes a few seconds, but evidence of its effectiveness is lacking. Randomized controlled trials of minimal (<1 minute) intervention based on the absolute risk warning are warranted.
For unknown reasons, such information and warning cannot be found in many websites to help smokers quit. Existing pictorial warnings also do not show or highlight this risk. The 1 in 2 risk has never been used as a key campaign message until recently. In 2011, Ireland was the first country to use 1 in 2 for a major social marketing campaign resulting in wide media coverage and great increases in calls to the Quitline and demand for smoking cessation services.11
For public health information and education, the great death risk of 1 in 2 to 2 in 3 should be highlighted consistently and widely. This warning could be placed on cigarette packages so that all smokers know that they are betting their lives on the toss of a coin.
For those who want to act now, one can look to the acronym “AWARD” as a guide: (1) A sk about smoking. (2) W arn your smoking patients by saying “If you continue to smoke, your chance of dying from smoking-induced diseases (such as cancer, heart diseases, stroke, and respiratory and many other serious diseases) is 50% (67% for the very young; 40% for the very old).” (3) A dvise: “If you quit now, your risk will be greatly reduced (by 25% at old age, and by much more before age 40 years).” This will take about 10 seconds. (4) R efer to a cessation clinic or hotline, and (5) D o it again until they quit. If you have helped 2 smokers quit, you have saved (at least) 1 life.
Correspondence: Dr Lam, School of Public Health, The University of Hong Kong, 21 Sassoon Rd, Pokfulam, Hong Kong (firstname.lastname@example.org).
Financial Disclosure: None reported.
Lam TH. Absolute Risk of Tobacco Deaths: One in Two Smokers Will Be Killed by SmokingComment on “Smoking and All-Cause Mortality in Older People”. Arch Intern Med. 2012;172(11):845-846. doi:10.1001/archinternmed.2012.1927