Search

The increasing proportion of lung cancers classified as adenocarcinoma has been a topic of interest and research. The main objective of the analyses reported here is to summarize how the proportion of adenocarcinoma varies in never smokers by time, sex and region based on published evidence on the distribution of lung cancer types available from epidemiological studies. Based on 219 sex- and period-specific blocks of data drawn from 157 publications, there appears to be a clear time-related increase in the proportion of lung cancers in never smokers that are adenocarcinoma, which is evident in both sexes, and not specific to any region. It is seen whether the denominator of the proportion is made up of adenocarcinoma plus squamous cell carcinoma cases, cases of the four major types combined, or all lung cancer cases. The ratio of adenocarcinoma to squamous cell carcinoma rose continuously from 1950 to 69 to be almost 4 times higher for the data from 2000 onwards. We discuss factors that may have contributed to the observed findings, including changes in lung cancer classification. Our findings argue against the hypothesis that increases in the ratio arise from changes in cigarette design and composition.

We quantified the decline in COPD risk following quitting using the negative exponential model, as previously carried out for other smoking-related diseases. We identified 14 blocks of RRs (from 11 studies) comparing current smokers, former smokers (by time quit) and never smokers, some studies providing sex-specific blocks. Corresponding pseudo-numbers of cases and controls/at risk formed the data for model-fitting. We estimated the half-life (H, time since quit when the excess risk becomes half that for a continuing smoker) for each block, except for one where no decline with quitting was evident, and H was not estimable. For the remaining 13 blocks, goodness-of-fit to the model was generally adequate, the combined estimate of H being 13.32 (95% CI 11.86–14.96) years. There was no heterogeneity in H, overall or by various studied sources. Sensitivity analyses allowing for reverse causation or different assumed times for the final quitting period little affected the results. The model summarizes quitting data well. The estimate of 13.32years is substantially larger than recent estimates of 4.40years for ischaemic heart disease and 4.78years for stroke, and also larger than the 9.93years for lung cancer. Heterogeneity was unimportant for COPD, unlike for the other three diseases.

The excess lung cancer risk from smoking declines with time quit, but the shape of the decline has never been precisely modelled, or meta-analyzed. From a database of studies of at least 100 cases, we extracted 106 blocks of RRs (from 85 studies) comparing current smokers, former smokers (by time quit) and never smokers. Corresponding pseudo-numbers of cases and controls (or at-risk) formed the data for fitting the negative exponential model. We estimated the half-life (H, time in years when the excess risk becomes half that for a continuing smoker) for each block, investigated model fit, and studied heterogeneity in H. We also conducted sensitivity analyses allowing for reverse causation, either ignoring short-term quitters (S1) or considering them smokers (S2). Model fit was poor ignoring reverse causation, but much improved for both sensitivity analyses. Estimates of H were similar for all three analyses. For the best-fitting analysis (S1), H was 9.93 (95% CI 9.31–10.60), but varied by sex (females 7.92, males 10.71), and age (

Background: To present up-to-date meta-analyses of evidence from Japan relating smoking to major smoking-related diseases. Methods: We restricted attention to lung cancer, chronic obstructive pulmonary disease (COPD), ischemic heart disease (IHD) and stroke, considering relative risks (RRs) for current and ex-smokers relative to never smokers. Evidence by amount smoked and time quit was also considered. For IHD and stroke only, studies had to provide age-adjusted RRs, with age-specific results considered. For each disease we extended earlier published databases to include more recent studies. Meta-analyses were conducted, with random-effects RRs and tests of heterogeneity presented. Results: Of 40 studies, 26 reported results for lung cancer and 7 to 9 for each other disease. For current smoking, RRs (95%CIs) were lung cancer 3.59 (3.25-3.96), COPD 3.57 (2.72-4.70), IHD 2.21 (1.96-2.50) and stroke 1.40 (1.25-1.57). Ex-smoking RRs were lower. Data for lung cancer and IHD showed a clear tendency for RRs to rise with increasing amount smoked and decrease with increasing time quit. Dose-response data were unavailable for COPD and unclear for stroke, where the association was weaker. Conclusions: Compared to studies in other Asian and Western countries, current smoking RRs were quite similar for IHD and stroke. The comparison is not clear for COPD, where the Japanese data, mainly from cross-sectional studies, is limited. For lung cancer, the RRs are similar to those in other Asian countries, but substantially lower than in Western countries. Explanations for this are unclear, but less accurate reporting of smoking by Japanese may contribute to the difference.

Over the period 1951-1995, lung cancer rates in men aged 35-74 yr more than doubled in the United States but declined slightly in the United Kingdom. In women, rates rose about sevenfold in the United States but only about threefold in the United Kingdom. To investigate whether these very different trends in lung cancer risk could be explained by smoking habits, trends in smoking were compared in the two countries and a multistage model was used to predict lung cancer rates from detailed data on age of starting and stopping smoking, amount smoked per smoker, and sales-weighted average tar levels. In both countries, there was a similar switch to filter cigarettes, reduction in tar levels and the average age of starting to smoke, and decline in prevalence of smoking in women aged under 50 yr and in men. Although some differences were evident, most notably in older women where prevalence of smoking and consumption per adult has increased more in the United States, these trends do not appear to explain the markedly different trends in lung cancer, evident in both sexes and all age groups. The multistage analyses confirmed these tentative conclusions-the differing trends in smoking in the two countries could not explain the markedly differing trends in lung cancer. Lung cancer trends in the United Kingdom were found to be clearly more favorable than expected on the basis of smoking trends, while trends in the United States were less favorable. In sensitivity analyses, these conclusions were found not to be materially dependent on the precise methods used, including whether tar reduction was or was not assumed to be beneficial. The explanation for these findings must lie in changes over time, differing in the two countries, in aspects of smoking not considered in these analyses and/or in exposure to other risk factors. Evidence relating to a number of possible such smoking variables (including type of tobacco, curing, use of pesticides and additives, and butt length) or other risk factors (including air pollution, radon, asbestos, obesity, and marijuana) is discussed, but no clear explanation of the findings is offered. Further research is urgently needed to investigate the causes of these apparently anomalous trends in lung cancer and in smoking habits. Criticism is also presented of the views recently expressed by the authors of NCI Monograph 13 that tar reduction has been ineffective in lowering lung cancer risk and that trends in lung cancer in the United States fit in well with trends in smoking habits.

We examine in detail arguments put forward to support the claim that exposure to environmental tobacco smoke (ETS) increases the risk of lung cancer in non-smokers. Hackshaw et al. [Br Med J 1997;315:980-988] have esti mated that the risk increases by 23% (95% Cl 14-32%) per 10 cigarettes/day smoked by the husband. The esti mated increase essentially disappears if proper adjust ment is made for smoking misclassification bias, if cor rection is made for the joint effects of confounding by fruit, vegetables, dietary fat and education, if errors in published data in one study are corrected, and if results from all pertinent studies are included (and not just those which report risk by level of smoking by the husband). Taking account of all these factors and using un- weighted estimates of the association between ETS ex posure and the confounding variables (as one very large study reported results discrepant from those for numer ous smaller studies), the risk increase per 10 cigarettes/ day was found to be 2% (95% Cl -3 to +7.5%), based on data from 47 ETS/lung cancer studies. Using weighted estimates, the risk increase was 5.5% (95% Cl 0 to +11 %). Restricting attention to the 36 studies that had adjusted for age, the increase was reduced further to -2% (95% Cl -6 to+3%) using un-weighted estimates, or to + 1 % (95% Cl -4 to +6%) using weighted estimates. These estimates are not materially affected by bias due to the reference group (non-smokers married to non-smokers) having some ETS exposure from other sources. Other sources of potential upward and downward bias, not formally taken account of in the analysis, are discussed. Based on extrapolation from the known lung cancer risk in smok ers, Hackshaw et al. [Br Med J 1997;315:980-988] esti mate that environmental tobacco smoke exposure would be expected to increase the risk of lung cancer in non-smokers by 19%. Using more appropriate assump tions (for the relative exposure to smoke constituents of passive and active smokers, for the lung cancer risk in those who have ever smoked and for the dose-response model) leads to a much lower estimate of about 0.5%. Even this estimate is open to question as a threshold might exist for the effects of tobacco smoke constituents on lung cancer risk. Whether or not a true risk exists, it is clear that this is not demonstrated by the overall evi dence. The true increase in risk per 10 cigarettes/day smoked by the husband is very unlikely to be as large as 23%. It might be as much as 5%, but it could well be 1 % or less, or even zero.

There is highly significant (p < 0.001) heterogeneity between estimates from 47 studies of the lung cancer risk in lifelong non-smoking women associated with each 10 cigarettes per day smoked by the husband or with whether or not the husband smokes. Sources of this heterogeneity were investigated by relating risk esti mates to 15 different study characteristics using multiple logistic regression analysis. There was a marked tenden cy for risk estimates to be higher in the 11 studies that had not adjusted for age, and there is a strong case for removing such studies from meta-analyses as being of unacceptable quality. Studies which report dose-re sponse results for smoking by the husband also report higher risks, demonstrating that estimates of the risk per 10 cigarettes per day smoked by the husband based only on those studies will be biased upward. There is also evi dence that risk estimates tend to be lower in large stud ies, in studies published in the 1990s, in studies not requiring histological confirmation of all cases and in studies where the proportion of proxy respondents was The work was supported by the Tobacco Manufacturers Associa tion. The views expressed are those of the authors. no higher in cases than in controls, though these associa tions are not independent. Variation in risk by study char acteristics largely explains the apparently low relative risk reported in one large Chinese study, arguing against the view that it is an outlier and should be excluded from meta-analyses.

Hackshaw et al. estimated that adjustment for bias due to misclassification of smoking habits reduces the ob served relative risk of lung cancer in non-smoking wom en associated with smoking by the husband from 1.24 to 1.18. This bias estimate is little affected by using an alter native method for misclassification adjustment or by using updated data from 47 studies rather than data from the 37 studies used by Hackshaw et al. The bias is increased if strong evidence of much higher misclassifi cation rates in Asian women is taken into account and could then explain about half the observed association. Misclassification correction has not previously been at tempted for dose-response data. We describe a suggest ed approach and apply it to data relating risk to amount smoked by the husband. As shown in paper II of this series the unadjusted increase in risk per 10 cigarettes/ day smoked by the husband is 10% (95% Cl 5 to 15%), reducing to either 6% (95% Cl 1 to 11 %) or 9% (95% Cl 5 to 14%) after adjustment for confounding by fruit, vege table and fat consumption and education using respec tively unweighted or weighted means to combine evi dence on the ETS/confounder association from different studies. Further adjustment for plausible levels of mis classification similarly reduced these two estimates to, respectively, 2% (95% Cl -3 to 8%) or 5.5% (95% Cl 0 to 11 %). Difficulties in applying misclassification bias cor rections are discussed. Other sources of bias will be con sidered in later papers in this series.

The relationship of environmental tobacco smoke to lung cancer risk in lifelong non-smokers is commonly studied using marriage to a smoker as the index of exposure. As smokers tend to marry smokers, relative risk estimates will be biased if some current or former smokers are misclassified as lifelong non-smokers. This paper shows how various factors affect the magnitude of the bias and describes a method for obtaining misclassification-adjusted relative risk estimates. Application of the method to U.S. and Asian data for women suggests misclassification is an important determinant of the slight excess risk observed in non-smokers married to smokers. Reasons why our conclusions differ from those of others are discussed, as are other difficulties in interpreting the association between spousal smoking and lung cancer risk.

Background National smoking-specific lung cancer mortality rates are unavailable, and studies presenting estimates are limited, particularly by histology. This hinders interpretation. We attempted to rectify this by deriving estimates indirectly, combining data from national rates and epidemiological studies. Methods We estimated study-specific absolute mortality rates and variances by histology and smoking habit (never/ever/current/former) based on relative risk estimates derived from studies published in the 20th century, coupled with WHO mortality data for age 70–74 for the relevant country and period. Studies with populations grossly unrepresentative nationally were excluded. 70–74 was chosen based on analyses of large cohort studies presenting rates by smoking and age. Variations by sex, period and region were assessed by meta-analysis and meta-regression. Results 148 studies provided estimates (Europe 59, America 54, China 22, other Asia 13), 54 providing estimates by histology (squamous cell carcinoma, adenocarcinoma). For all smoking habits and lung cancer types, mortality rates were higher in males, the excess less evident for never smokers. Never smoker rates were clearly highest in China, and showed some increasing time trend, particularly for adenocarcinoma. Ever smoker rates were higher in parts of Europe and America than in China, with the time trend very clear, especially for adenocarcinoma. Variations by time trend and continent were clear for current smokers (rates being higher in Europe and America than Asia), but less clear for former smokers. Models involving continent and trend explained much variability, but non-linearity was sometimes seen (with rates lower in 1991–99 than 1981–90), and there was regional variation within continent (with rates in Europe often high in UK and low in Scandinavia, and higher in North than South America). Conclusions The indirect method may be questioned, because of variations in definition of smoking and lung cancer type in the epidemiological database, changes over time in diagnosis of lung cancer types, lack of national representativeness of some studies, and regional variation in smoking misclassification. However, the results seem consistent with the literature, and provide additional information on variability by time and region, including evidence of a rise in never smoker adenocarcinoma rates relative to squamous cell carcinoma rates.

OBJECTIVE--To investigate the association between bird keeping and risk of lung cancer. DESIGN--Case-control study asking detailed questions on exposure to domestic birds and other pets, smoking, and various demographic and potentially confounding variables. SETTING--District general hospital; current admissions interviewed in hospital or recent admissions interviewed at home. PATIENTS--143 patients with lung cancer, 143 controls with heart disease, and 143 controls with orthopaedic conditions individually matched for age, sex, date of admission, and current or past admission. MAIN OUTCOME MEASURES--Odds ratios for lung cancer in relation to various aspects of bird keeping, after adjustment for smoking and other relevant confounding variables. RESULTS--Risk of lung cancer was not significantly associated with household exposure to pet birds at any time or at various specific periods in life, or to keeping large numbers of birds. For specific types of birds no association was seen for living in households with budgerigars or canaries but risk was significantly associated with keeping pigeons (odds ratio 3.53, 95% confidence interval 1.56 to 7.98). This remained significant after regression analysis to account for confounding variables (3.9, 1.2 to 12.62) in both sexes and all age groups. CONCLUSION--Bird keeping may confer some risk of lung cancer but the relation is not as strong as previously reported.

We compared risk of lung cancer and chronic obstructive pulmonary disease (COPD) associated with flue-cured and blended cigarettes. Mortality and smoking data were collected for 1971-2000 by sex, age, and period for three countries with a mainly flue-cured market and four with a blended market. Epidemiological relative risk estimates for current and ex smoking were summarized. Smoking statistics and mortality were compared between flue-cured cigarette and blended cigarette countries. Unadjusted mortality rates were generally lower in blended cigarette countries early on, with the difference diminishing or reversing by the 1990s. Differences by cigarette type were rarely significant, due to variations, particularly for COPD, between countries within cigarette type. Current smoking prevalence was generally lower in blended cigarette countries in 1971-1975, with the difference reducing over time. Differences by type were never significant, with blended cigarette countries varying markedly. Ex-smoking increased over time and was lower for blended cigarette countries, generally not significantly. Consumption per smoker was somewhat lower for blended cigarette countries. Relative risk estimates for smoking, derived mainly from U.S. and UK studies, varied little by cigarette type. Conclusions based on estimated smoking-related excess mortality were similar to those based on unadjusted mortality rates. There was little indication of any difference between flue-cured and blended cigarettes on risk of lung cancer or COPD. Our approach could have detected differences of about 40% for male lung cancer, or twofold differences for females or for COPD, had they existed. Between-country differences in rates of two major diseases predominantly caused by smoking cannot materially be explained by whether the countries use flue-cured or blended cigarettes.

Trends in smoking associated respiratory diseases in England and Wales during 1941-85 have been studied, with careful attention to problems caused by changes in classification of cause of death. Three diseases were selected for analysis: lung cancer, emphysema, and chronic obstructive lung disease. During 1971-85 deaths that would previously have been certified under chronic bronchitis have increasingly tended to be classified under chronic airways obstruction. The definition of chronic obstructive lung disease that was used includes both terms to avoid the artificial decline caused by consideration of chronic bronchitis in isolation. Age specific rates for all three diseases show a pronounced cohort (period of birth) pattern, rates for men rising up to the rates for those born shortly after the turn of the century and then declining, and rates for women peaking in the cohort born 20-25 years later. For chronic obstructive lung disease, but not for lung cancer and emphysema, the cohort peak is superimposed on a sharply declining downward trend. In both sexes cohort patterns of cumulative cigarette consumption peak at a time broadly similar to those seen for the three diseases. Trends in cigarette consumption, however, cannot explain the underlying steeply declining rate of chronic obstructive lung disease. Nor can they fully explain the declining trends in lung cancer and emphysema rates in younger men and women.

Between 1985 and 2000, lung cancer rates in U.S. men and women aged 35-54 yr have declined. To investigate whether these declines can adequately be explained by changes in smoking prevalence, consumption, and duration, or if changes in tar and nicotine yields also contributed, two model-fitting approaches were used. Both approaches used individual person National Health Interview Survey data on smoking prevalence, age of starting and time of quitting, and national estimates of consumption per smoker and yields. Both approaches compared observed rates (by sex and age) relative to 1985, with those predicted after successively including various smoking variables into the model, making varying allowance for compensation for reduced yield. Approach A was simpler, based on mean smoking statistics estimated separately for current and former smokers. Approach B used the multistage model and individual smoking histories to estimate risk. Both approaches showed observed declines in risk were (except for men aged 35-39 yr) clearly greater than predicted based only on prevalence, consumption, and duration. Including yield generally improved the fit. At younger ages, models assuming substantial compensation (consistent with evidence from studies relating nicotine yield and intake) fitted well, but at age 50-54 yr in both sexes and age 45-49 yr in women, the decline was better fitted by models assuming little compensation. The conclusions were not sensitive to the precise parameter values assumed in the modeling. Interpretation is not straightforward, but the findings suggest declines in yields have contributed to the recent declines in rates in young U.S. men and women.

One possible contributor to the reported rise in the ratio of adenocarcinoma to squamous cell carcinoma of the lung may be differences in the pattern of decline in risk following quitting for the two lung cancer types. Earlier, using data from 85 studies comparing overall lung cancer risks in current smokers, quitters (by time quit) and never smokers, we fitted the negative exponential model, deriving an estimate of 9.93years for the half-life – the time when the excess risk for quitters compared to never smokers becomes half that for continuing smokers. Here we applied the same techniques to data from 16 studies providing RRs specific for lung cancer type. From the 13 studies where the half-life was estimable for each type, we derived estimates of 11.68 (95% CI 10.22–13.34) for squamous cell carcinoma and 14.45 (11.92–17.52) for adenocarcinoma. The ratio of the half-lives was estimated as 1.32 (95% CI 1.20–1.46, p

Background Smoking is a known lung cancer cause, but no detailed quantitative systematic review exists. We summarize evidence for various indices. Methods Papers published before 2000 describing epidemiological studies involving 100+ lung cancer cases were obtained from Medline and other sources. Studies were classified as principal, or subsidiary where cases overlapped with principal studies. Data were extracted on design, exposures, histological types and confounder adjustment. RRs/ORs and 95% CIs were extracted for ever, current and ex smoking of cigarettes, pipes and cigars and indices of cigarette type and dose–response. Meta-analyses and meta-regressions investigated how relationships varied by study and RR characteristics, mainly for outcomes exactly or closely equivalent to all lung cancer, squamous cell carcinoma (“squamous”) and adenocarcinoma (“adeno”). Results 287 studies (20 subsidiary) were identified. Although RR estimates were markedly heterogeneous, the meta-analyses demonstrated a relationship of smoking with lung cancer risk, clearly seen for ever smoking (random-effects RR 5.50, CI 5.07-5.96) current smoking (8.43, 7.63-9.31), ex smoking (4.30, 3.93-4.71) and pipe/cigar only smoking (2.92, 2.38-3.57). It was stronger for squamous (current smoking RR 16.91, 13.14-21.76) than adeno (4.21, 3.32-5.34), and evident in both sexes (RRs somewhat higher in males), all continents (RRs highest for North America and lowest for Asia, particularly China), and both study types (RRs higher for prospective studies). Relationships were somewhat stronger in later starting and larger studies. RR estimates were similar in cigarette only and mixed smokers, and similar in smokers of pipes/cigars only, pipes only and cigars only. Exceptionally no increase in adeno risk was seen for pipe/cigar only smokers (0.93, 0.62-1.40). RRs were unrelated to mentholation, and higher for non-filter and handrolled cigarettes. RRs increased with amount smoked, duration, earlier starting age, tar level and fraction smoked and decreased with time quit. Relationships were strongest for small and squamous cell, intermediate for large cell and weakest for adenocarcinoma. Covariate-adjustment little affected RR estimates. Conclusions The association of lung cancer with smoking is strong, evident for all lung cancer types, dose-related and insensitive to covariate-adjustment. This emphasises the causal nature of the relationship. Our results quantify the relationships more precisely than previously.