証拠の質 4.5/5
8次元のレビュー評価。基準は 品質ルーブリック 。各次元は1〜5で評価。
- D1 出典への根拠
- 5/5
- D2 出典の権威性
- 5/5
- D3 算術
- 4/5
- D4 不確実性
- 4/5
- D5 範囲
- 5/5
- D6 文章
- 4/5
- D7 認識の誠実性
- 4/5
- D8 注意事項の完全性
- 5/5
● あなたの要因 — このリスクをクリック ▾ して表示
- あなたの要因
認知リスク
喫煙は、ほぼすべての成人がこの活動を「危険」だと正しく位置づける稀な恐怖の一つです——警告ラベル、数十年にわたる禁煙キャンペーン、そして病棟で目に見える結果が、質的なレベルでその役割を果たしてきました。典型的な読者が通常内面化しないのは、具体的な数値の大きさです。生涯にわたる日常的な喫煙者の約半分が喫煙に起因する疾患で死亡し、平均余命の損失は約10年です。ほとんどの人は、推測を求められると、50%をはるかに下回るどこかに着地します。「そう、それは悪い」と「コイン投げだ」の間のギャップが、この項目の興味深い部分です。このページは一般人口の統計ではありません——非喫煙者の多数派と平均したものではなく、実際に定期的に喫煙する人の生涯帰属死亡率です。
概算: ほとんどの成人は喫煙が危険だと知っているが、生涯リスクを2分の1をはるかに下回ると推測する
出典: 編集部の直感(調査に基づかない)
実際のリスク
生涯にわたる日常的な喫煙者の最大で半分がタバコに起因する疾患で死亡する
禁煙しない生涯にわたる日常的な喫煙者
計算を表示
参照サブグループ:成人期初期(10代半ばから20代半ば)に定期的に喫煙を始め、およそ1日1箱で喫煙を続け、禁煙しない成人。約50%という見出しは、2つの収束するエビデンスの流れに由来します。(1) WHOのタバコファクトシートは、「タバコは禁煙しない使用者の最大で半分を殺す」と明言しています。(2) 喫煙と死亡に関する米国最大の前向きコホート研究であるJhaら NEJM 2013年は、25-79歳の現在の喫煙者において全死因の死亡率が非喫煙者の約3倍であり、平均して10年超の平均余命が失われることを見出しました——これは生涯喫煙者の約半分が喫煙に起因する原因で死亡することとのみ算術的に整合する数字です。(3) 英国医師研究の50年追跡であるDollら2004年は、生涯にわたる紙巻きタバコ喫煙者と非喫煙者の間の平均10年の平均余命差を報告し、「成人期初期からの長期にわたる喫煙は年齢別死亡率を3倍にした」ことを示しました。見出しの数字は0.5(≈ 2分の1)で、不確実性の幅は0.4-0.6であり、英国医師コホート(1900-1930年生まれ、より重いフィルターなしタバコを喫煙)と現代の米国喫煙者(強度は軽いが平均して開始が早く期間が長い)の間の時代、強度、コホートの違いを反映しています。スコープはsubgroup_lifetimeと宣言されています。これは一般人口の生涯リスクではなく生涯喫煙者1人あたりの確率であり、他のLikelierページの世界/米国成人の生涯数字と直接比較できないためです。
注意事項: この項目は具体的に、成人期初期から老年期まで定期的に喫煙する人の生涯帰属死亡率であり、一般人口の平均ではありません。喫煙者と非喫煙者を平均した他のLikelie…
この項目は具体的に、成人期初期から老年期まで定期的に喫煙する人の生涯帰属死亡率であり、一般人口の平均ではありません。喫煙者と非喫煙者を平均した他のLikelierページの人口スコープの生涯数値(がん、心疾患、脳卒中)とは直接比較できません。喫煙は、それら他の多くの項目の背後にある最大の単一の可変リスク入力です。肺がん死亡の約80%、虚血性心疾患死亡の意味のある割合、そして脳卒中、COPD、膀胱がん、食道がん、いくつかの頭頸部がんの帰属負担の大半を占めます。その意味で、このページはサイト上の他の多くの項目の背後にあるメタ項目です。具体的な約50%という数字は決定論的ではありません——個人の結果は、強度(1日あたりの本数)、期間(喫煙年数)、開始年齢、禁煙年齢、および遺伝的・環境的修飾因子の長いリストに依存します。見出しはハザードの規模に対する較正の基準点であり、個人の予測ではありません。どの年齢で禁煙しても結果は有意に改善します。Jha NEJM 2013年によると、40歳前の禁煙は平均して失われた平均余命の約90%を回復します。
リスクはどう変わるか
見出しの数値は大きく異なる状況の平均です。シナリオや文脈による確率の違いは以下の通りです:
2.0分の1 · 50%
Headline subgroup. Based on WHO fact-sheet language and Doll/Jha hazard ratios.
2.9分の1 · 35%
Lower exposure dose; per-person mortality reduced but still dominant risk factor
20分の1 · 5.0%
Jha NEJM 2013: cessation before 40 reduces excess mortality by ~90%
50分の1 · 2.0%
Doll 2004: cessation at 30 avoided almost all of the excess hazard
—
No smoking-attributable mortality — this entry is about excess attributable risk only
バーの長さと濃淡は、これらのシナリオを他のリスクとではなく、互いに比較して順位付けしたものです。正確な確率は各項目の横に表示されています。
関連するリスク
似たテーマの他のリスク — 関連する不安を探るために。
このリスクが関わる決断
このリスク次第で変わる選択。そのトレードオフをどう見るか。
比較対象を選択
The headline number for a lifelong regular smoker is roughly 1 in 2. Not one in ten, not one in five — a coin flip. The World Health Organization’s tobacco fact sheet puts it as bluntly as any public-health institution ever puts anything: “tobacco kills up to half of its users who don’t quit”. Doll’s 50-year follow-up of the British Doctors cohort — the longest prospective smoking-mortality study in existence — found that men born around 1920 who smoked cigarettes from early adulthood had their age-specific mortality rates tripled, and died on average about ten years younger than lifelong non-smokers. Jha et al. in the New England Journal of Medicine in 2013 replicated the finding in a US population of roughly 200,000 adults: all-cause mortality among current smokers in the 25-79 age band was about three times the never-smoker rate, and life expectancy was shortened by more than a decade. The CDC attributes about 480,000 US deaths a year to smoking and secondhand smoke — “nearly one in five” deaths in the country. It is the single largest single-factor mortality multiplier attached to any voluntary choice the Likelier catalogue covers.
The life-expectancy arithmetic is what makes this entry unusual. Most health fears on the site have lifetime mortality figures in the single-digit-percent range; regular smoking sits an order of magnitude above them, compressed into one behavioural variable. The unusual part is that the reversibility is also huge. Jha found that cessation before age 40 reduced the excess mortality associated with continued smoking by about 90%. Doll found that cessation at age 30 “avoided almost all” of the excess hazard, cessation at 50 halved it, and even cessation at 60 bought back roughly three years of life expectancy. Very few risks in the Likelier catalogue are that steeply dose-dependent on a single ongoing choice, and almost none are that steeply reversible. This is why the entry is tagged calibrated rather than underrated: most smokers know it’s bad, and most of them could still recover the majority of the hazard by quitting.
Where the headline doesn’t apply: almost every other Likelier page on a major chronic-disease cause of death. This entry is the meta-entry behind several of them. Smoking is responsible for roughly 80% of lung cancer deaths, a material share of ischaemic heart disease mortality, and most of the attributable burden for stroke, COPD, bladder cancer, oesophageal cancer, and several head-and-neck cancers. When other Likelier entries apply a “current heavy smoker ≈ 2-3x” multiplier, the arithmetic being compressed into that multiplier is the same arithmetic on this page, unpacked into a single input. The ~50% figure is also sensitive to era and cohort: it is drawn from populations who smoked heavier unfiltered cigarettes, while today’s median smoker tends to smoke lighter cigarettes but start younger and smoke for more years on average. Modern light cigarettes are not meaningfully safer per cigarette; the difference is mainly in how many cigarettes the average modern smoker actually smokes. The subgroup definition matters: “lifelong regular smoker” is the population this page measures. Former smokers, occasional smokers, and never-smokers belong on very different parts of the distribution, which is what the regional_breakdown rows are for.
関連する豆知識
根拠台帳
以下の各数値は各出典が報告した内容であり、引用した原文の抜粋と算出方法を記載しています。リンクをクリックして直接確認できます。
-
[1] World Health Organization — Tobacco — fact sheet
Tobacco — fact sheetSee all 2 Likelier entries citing this source →
- 統計値
Tobacco kills more than 7 million people per year, including 1.6 million non-smokers from secondhand smoke; kills up to half of its users who don't quit- 抜粋
“"Tobacco kills more than 7 million people each year, including an estimated 1.6 million non-smokers who are exposed to second-hand smoke. [...] Tobacco kills up to half of its users who don’t quit." ”
- 出典データ
- 2025-07-31
- アクセス日
- 2026-04-11 · アーカイブ版
- 計算過程
- The WHO "up to half of its users who don’t quit" formulation is the single most widely cited institutional statement of the headline figure. It is the shorthand for the mortality hazard ratios reported in the Doll and Jha cohort studies and is the direct source for the 0.5 point estimate. Paired with the 7 million-deaths-per-year aggregate: across roughly 1.1 billion current smokers worldwide, 7 million deaths/year implies an annual smoking-attributable death rate of ~6.4 per 1,000 smokers, compounded over a 50-year regular-smoking career to roughly 1 − (1 − 0.0064)^50 ≈ 0.28 as a floor, which rises to ~0.5 once the hazard ratio concentration in the second half of life is accounted for (smoking-attributable mortality is dominated by ages 55-80).
- 独立性
- WHO draws on IHME Global Burden of Disease estimates for the 7-million headline and on the Doll / Jha cohort studies (cited separately below) for the "up to half" formulation. Treat WHO as the authoritative institutional endorsement of figures that ultimately trace back to the same underlying cohort literature, not as a fully independent line of evidence.
-
[2] New England Journal of Medicine (Jha, Ramasundarahettige, Landsman, Rostron, Thun, Anderson, McAfee, Peto) — 21st-Century Hazards of Smoking and Benefits of Cessation in the United States
21st-Century Hazards of Smoking and Benefits of Cessation in the United States- 統計値
Among current smokers aged 25-79, all-cause mortality was ~3x the never-smoker rate; life expectancy shortened by more than 10 years; cessation before age 40 reduces the excess mortality by ~90%- 抜粋
“"For participants who were 25 to 79 years of age, the rate of death from any cause among current smokers was about three times that among those who had never smoked. [...] Life expectancy was shortened by more than 10 years among the current smokers, as compared with those who had never smoked. [...] Adults who had quit smoking at 25 to 34, 35 to 44, or 45 to 54 years of age gained about 10, 9, and 6 years of life, respectively, as compared with those who continued to smoke." ”
- 出典データ
- 2013-01-24
- アクセス日
- 2026-04-11 · アーカイブ版
- 計算過程
- Jha et al. followed ~200,000 US adults via the National Health Interview Survey linked to the National Death Index. The 3x all-cause hazard ratio for current smokers is the primary quantitative basis for the "half die from it" shorthand: if baseline never-smoker all-cause mortality accounts for essentially all never-smoker deaths, then a 3x hazard implies roughly two-thirds of a smoker’s deaths are "excess" and smoking-attributable, which — combined with a premature death rate dominated by ages 55-80 — works out to roughly half of a lifelong-smoker cohort dying from tobacco. The "quit before 40 reduces the risk by about 90%" finding is the basis for the regional_breakdown rows on quitting and for the "quit decades ago" personal-factor multiplier.
- 独立性
- Jha et al. use US NHIS/NDI data and are methodologically independent of the Doll British Doctors cohort and of WHO/IHME modeled estimates. This is the strongest single independent cross-check on the ~50% figure.
-
[3] BMJ (Doll, Peto, Boreham, Sutherland) — Mortality in relation to smoking: 50 years' observations on male British doctors
Mortality in relation to smoking: 50 years' observations on male British doctors- 統計値
Among men born 1900-1930, prolonged cigarette smoking from early adult life tripled age-specific mortality rates; 10-year average life-expectancy loss vs non-smokers; cessation at 30, 40, 50, 60 gained ~10, 9, 6, 3 years- 抜粋
“"Men born in 1900-1930 who smoked only cigarettes and continued smoking died on average about 10 years younger than lifelong non-smokers. [...] Among the men born around 1920, prolonged cigarette smoking from early adult life tripled age specific mortality rates, but cessation at age 50 halved the hazard, and cessation at age 30 avoided almost all of it. [...] Cessation at age 60, 50, 40, or 30 years gained, respectively, about 3, 6, 9, or 10 years of life expectancy." ”
- 出典データ
- 2004-06-26
- アクセス日
- 2026-04-11 · アーカイブ版
- 計算過程
- Doll’s 50-year follow-up of the British Doctors cohort is the longest prospective smoking-mortality study in existence and the original source of the "half of smokers killed by their habit" finding. The tripled hazard ratio in the 1920-born subcohort and the 10-year life-expectancy gap are the empirical anchors for the ~50% point estimate. The British Doctors cohort smoked heavier unfiltered cigarettes than modern smokers, which is one reason the Jha US cohort finds a slightly lower hazard ratio (3x all-cause vs Doll’s 3x age-specific in the 1920-born subgroup) but a very similar life-expectancy-loss figure.
- 独立性
- The Doll cohort is independent from Jha’s NHIS/NDI cohort — different population, different era, different follow-up methodology — and provides the longest-duration anchor for the headline figure.
-
[4] US Centers for Disease Control and Prevention — About Smoking and Tobacco Use
About Smoking and Tobacco UseSee all 2 Likelier entries citing this source →
- 統計値
Smoking and secondhand smoke exposure cause >480,000 US deaths per year (~1 in 5 US deaths); >16 million Americans live with a smoking-caused disease- 抜粋
“"Smoking and secondhand smoke exposure cause more than 480,000 deaths each year in the United States. This is nearly one in five deaths. More than 16 million Americans live with a disease caused by smoking. [...] Secondhand smoke exposure contributes to over 40,000 deaths among nonsmoking adults and 400 deaths in infants each year." ”
- 出典データ
- 2024-05-15
- アクセス日
- 2026-04-11 · アーカイブ版
- 計算過程
- CDC’s ~480,000 annual US smoking-attributable deaths figure is the standard domestic headline. Across ~28 million US adult current smokers plus ~51 million former smokers who retain elevated residual risk, that implies an annual smoking-attributable mortality rate on the order of 6 per 1,000 for the combined current-plus-former population, consistent with the ~6.4 per 1,000 global per-smoker figure derived from the WHO 7-million aggregate. Used as the domestic anchor and as the basis for the "smoking causes nearly 1 in 5 US deaths" plain-English framing in the body text.
- 独立性
- CDC smoking-attributable mortality estimates use the SAMMEC (Smoking- Attributable Mortality, Morbidity, and Economic Costs) model, which draws on Cancer Prevention Study II hazard ratios — overlapping but not identical to the cohorts used by Jha and Doll. Treat as partially dependent institutional verification of the ~50% figure rather than a fully independent estimate.
-
[5] US Centers for Disease Control and Prevention — Lung Cancer Risk Factors
Lung Cancer Risk FactorsSee all 2 Likelier entries citing this source →
- 統計値
In the United States, cigarette smoking is linked to about 80% of lung cancer deaths- 抜粋
“"In the United States, cigarette smoking is linked to about 80% of lung cancer deaths." ”
- 出典データ
- 2026-05-18
- アクセス日
- 2026-07-03 · アーカイブ版
- 計算過程
- Used as the direct source for the body-text claim that smoking is responsible for roughly 80% of US lung cancer deaths, cited alongside the WHO/Jha/Doll/CDC mortality figures as evidence that this entry is the "meta-entry" behind the site's other tobacco-attributable disease pages (lung cancer, COPD, several other cancers).
- 独立性
- A separate CDC page from the mortality figures above; both are CDC institutional statements but draw on different underlying surveillance systems (cancer registries vs SAMMEC mortality modeling).








