Tobacco use, body mass index and cancer

Tobacco use, body mass and cancer mortality in Mumbai Cohort StudyMangesh S.. The Mumbai Cohort Study MCS, one of the largest prospective studies in the world and based in the largest ci

Trang 1

Tobacco use, body mass and cancer mortality in Mumbai Cohort Study

Mangesh S Pednekara,* , James R He´berta,b,c, Prakash C Guptaa,c

a Healis-Sekhsaria Institute for Public Health, CBD Belapur, Navi Mumbai, India

b South Carolina Statewide Cancer Prevention & Control Program, University of South Carolina, Columbia, SC, United States

c

Department of Epidemiology and Biostatistics, Arnold School of Public Health, University of South Carolina, Columbia, SC, United States

1 Introduction

In developed countries, smoking and excess body weight (based

on measures of body mass adjusted for height), are two of the most

important modiﬁable risk factors for chronic disease incidence and

premature death[1–4] Use of tobacco in all forms is known to

increase the risk of many cancers[5–7], especially those of the

upper aerodigestive tract[1,8,9]

Both body weight and tobacco usage have strong behavioral

determinants, but neither is controlled adequately by

population-based approaches to behavior change [10–13] Although the

prevalence of smoking has declined modestly in some countries

over the past 30 years[14,15], the prevalence of both overweight

and obesity have risen steadily[16,17] It is well known that diet

and tobacco habits are related[14,18–20], though the evidences

from developing countries are often much different than what is

observed in Europe and North America[12,21–24]

Studies conducted primarily in the West have demonstrated that

overweight and obesity increase the risk of various types of cancer,

especially colorectal[25]and those at hormone-sensitive sites[26–

29] In contrast, studies from Asia, where overweight and obesity are relatively recent phenomena and prevalence is still low, results have often departed from what is observed in studies from the West[30] Furthermore, smoking of tobacco, a known strong risk factor for many cancers (but not particularly so for cancers related to obesity),

is less negatively associated with affluence and measures of socioeconomic status (SES) in developing countries than in the West[17,31–34] So, there is increased potential for confounding and effect modification of tobacco by body mass and SES The Mumbai Cohort Study (MCS), one of the largest prospective studies in the world and based in the largest city in South Asia, was designed specifically to investigate the effect of tobacco on various health outcomes, including cancer MCS has already demonstrated that use of tobacco in all forms[35]and low body mass index[36] [BMI = weight (kg)/height (m2)] independently increase the risk of cancer deaths Also the joint effect of tobacco use and BMI on all-cause mortality had been reported earlier[37] Now using data from the MCS we report the findings of the joint effect of tobacco and BMI on cancer mortality

2 Materials and methods Recruitment: The MCS was conducted in the main city of Mumbai, with mortality as the endpoint[38] A total of 148,173

A R T I C L E I N F O

Article history:

Accepted 24 September 2009

Keywords:

Cancer

Mortality

Body mass index

Tobacco

Cohort study

A B S T R A C T

Background: Tobacco use and body mass are major risk factors for many cancers Despite this, very little

is known about their combined effect on cancer mortality These relationships are virtually unexplored

in populations having patterns of both tobacco use and body habitus atypical of those typically enrolled

in epidemiologic studies Methods: A prospective cohort study of 148,173 men and women aged 35 years was conducted in Mumbai, India Subjects were recruited during 1991–1997, measured for a variety of risk factors, including tobacco use and anthropometry, and then followed for approximately 5–

6 years Results: During 774,129 person-years of follow up, a total of 796 cancer deaths were observed Tobacco use, especially smoking in men, was associated with particularly high risk of death in extreme categories of body mass At highest risk were underweight smoking males [hazard ratio (HR) = 9.45, 5.87, and 5.75 for those smokers who were extremely thin (<16.0 kg/m2), very thin (16.0 to <17.0 kg/m2), or thin (17.0 to <18.5), respectively] Signiﬁcant effects of underweight among never and smokeless tobacco users disappeared with exclusion of individuals with 2 years of follow up Extremely thin (<16.0 kg/m2) women smokeless tobacco users had an elevation in risk, HR = 2.95, that actually appeared to increase (to 3.21) with exclusion of individuals who were diagnosed within 2 years of follow

up Conclusion: Tobacco use and undernutrition are known to be serious problems in developing countries The current study underlines the strikingly elevated risk of cancer when they occur together

* Corresponding author Tel.: +91 22 2757 5487; fax: +91 22 2786 3750.

E-mail address: pednekarmangesh@healis.org (M.S Pednekar).

Contents lists available atScienceDirect

Cancer Epidemiology The International Journal of Cancer Epidemiology, Detection, and Prevention

j o u r n a l h o m e p a g e : w w w c a n c e r e p i d e m i o l o g y n e t

Trang 2

individuals 35 years of age were recruited during 1991–1997.

House-to-house interviews were conducted in a face-to-face

manner using a structured questionnaire Electoral rolls (ofﬁcial

municipal government records), organized by area with a polling

station of 1000–1500 individuals as the smallest geographical unit,

were used as the sampling frame These electoral rolls provided

name, age, sex, and address of all the individuals aged 18 years

Face-to-face interviews were conducted by trained ﬁeld

investi-gators using hand-held computers (electronic diaries) in local

languages (e.g., Marathi) but the responses were recorded in

English The study satisﬁed all the criteria with regard to the ethical

treatment of human subjects, especially those formulated by the

Indian Council of Medical Research Details regarding the

recruitment procedures have been described earlier[12,30,35–38]

Data Sources: The baseline survey included the following two

components: (1) anthropometry to measure weight (using a

bathroom scale that was calibrated to the nearest kilogram) and

height (using a measuring tape that was calibrated to the nearest

centimeter); and (2) interviewer administration of a structured

questionnaire For this study, data regarding age, sex, education

(proxy for SES), religion, mother tongue, height, weight, and details

regarding tobacco use were abstracted from the baseline survey[38]

Follow up: An active house-to-house follow up was conducted

on average 5.5 years after the initial survey The ﬁeld investigators

were instructed to revisit each person If the person was alive and

available, a face-to-face re-interview was conducted If the person

was reported to have died, the date and place of death were

recorded with extra questioning and follow up as necessary

Permanent migration from the study area was considered as

withdrawal from the study, and the date of migration was noted

The re-interviews were conducted during 1997–2003 Follow up

results have been described earlier[35–37]

Cause of death: Deaths recorded during the MCS follow up were

linked with the dataset obtained from the Bombay Municipal

Corporation (BMC) death registers In Mumbai, almost all the

deaths are registered and medically certiﬁed [35] For matched

deaths, the underlying cause of death was derived from the cause

information copied from the corporation death registers and then

coded according to the International Classification of Diseases—10 guidelines Cause-specific analyses were performed for cancer deaths (ICD-10 codes C00–97) Of the total matched deaths 1685 were randomly selected, an independent field check was performed; matching was found to be nearly 100% accurate[35] Statistical analysis: Methodologic details regarding the estima-tion of person-years of follow up, anthropometric measurements, and information collected from the structured questionnaire have been described earlier [12,35–39] Multivariate analysis was performed by using the Cox proportional hazards regression modeling [40] The response variable, cancer death, was a dichotomous (‘‘yes’’ or ‘‘no’’) and the time to event (or censoring) was continuous BMI categories were defined as follows (all units kg/m2): extremely thin (BMI < 16.0); very thin (BMI 16.0 to

<17.0); thin (BMI 17.0 to <18.5); normal (BMI 18.5 to <25.0); overweight (BMI 25.0 to <30.0); and obese (BMI 30.0) Details regarding the BMI distribution in the Mumbai cohort have been published earlier[12,39] Respondents were broadly classified as having never used tobacco, used smokeless tobacco only, being a smoker (which also may include smokeless tobacco use) Being consistent with our earlier definition of the reference category, which was based on observed lowest death rate category [37]; never tobacco users having BMI 25 to <30 kg/m2was used as the referent throughout this analysis as well Age, education, religion, mother tongue, tobacco use, and BMI were fit as independent variables in the final Cox proportional hazards model Adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated for the joint effect of tobacco use and BMI on cancer mortality, stratified by gender Analyses also were conducted to study the multiplicative effect of tobacco use and BMI on cancer mortality Expected HRs were calculated by multiplying individual HRs for the various categories of tobacco users by BMI categories

3 Results

Of the total 148,173 (59,515 women and 88,658 men) cohort members, 7265 were not traceable (Fig 1); the most common reason for non-traceability was the demolition of their residential

Trang 3

building (6452 persons) for the purpose of redevelopment Among

the remaining 140,908 individuals, 25,777 subjects, although alive,

had migrated outside the study area; thus, 115,131 subjects were

re-contacted A total of 13,261 deaths were reported For 260

deaths, date of expiry was found to precede the date of

recruitment; hence, these subjects were excluded Detailed

investigation of a sample of these deaths revealed that the death

had occurred very close to the date of recruitment of these

subjects Of the total 9259 matched and coded deaths 796 cancer

deaths were available for these analyses Re-interviews were

conducted for 90,282 individuals; the remaining 11,588

indivi-duals, although assessed to be alive and traceable, were not

available despite multiple visits In the latter group, the last date of

attempted follow up was regarded as the withdrawal date

In Table 1, we show the distributions of important variables,

including BMI and tobacco use, as well as important covariates that

could affect cancer-related outcomes Because they could function

as confounders and effect modiﬁers of hypothesized effects of BMI

and tobacco use in relation to cancer endpoints, they were

controlled in the Cox regression analyses

When compared with male never tobacco users having BMI 25

to <30 kg/m2 (Table 2), an elevated risk of cancer death was observed for smokers across all BMI categories In general, the thinnest individuals experienced the highest levels of risk In all instances, except in the obese, exclusion of individuals who had died in the ﬁrst 2 years of follow up yielded results that were still signiﬁcant at the nominal a= 0.05 In all but very thin (16.0 to

<17.0 kg/m2) smokeless tobacco users, there was an apparent reduction in risk, with the largest such reduction among the extremely thin (<16.0 kg/m2) In never-using men, risks at the two thinnest levels of BMI were signiﬁcantly greater than 1.0 However, exclusion of individuals who had died in the ﬁrst 2 years of follow

up yielded non-signiﬁcant results Similarly, elevations of risk in extremely thin and normal weight male smokeless tobacco users disappeared when individuals with 2 years or less of follow up were excluded; though the reduction in normal-weight individuals was small

Only about half as many women died of cancer compared to men Very few women reported smoking, hence data on women smokers were excluded from the analysis The highest risk group,

Table 1

Summary of descriptive data.

Overall (n = 59,515) Deaths (n = 244) Overall (n = 88,658) Deaths (n = 552) Continuous variables Mean (SEM) Mean (SEM) Mean (SEM) Mean (SEM) Age (years) 48.13 (0.046) 56.96 (0.736) 52.68 (0.037) 62.24 (0.445) BMI (kg/m 2

Categorical variables % (95% CI) % (95% CI) % (95% CI) % (95% CI)

Age

35–39 24.84 (24.49–25.18) 5.74 (2.82–8.66) 9.45 (9.26–9.64) 1.81 (0.70–2.92) 40–44 16.90 (16.60–17.20) 7.38 (4.10–10.66) 7.94 (7.77–8.12) 2.17 (0.96–3.39) 45–49 15.35 (15.06–15.65) 9.02 (5.42–12.61) 26.74 (26.45–27.03) 8.70 (6.35–11.05) 50–54 13.30 (13.02–13.57) 17.62 (12.84–22.40) 16.15 (15.91–16.39) 9.60 (7.14–12.06) 55–59 9.87 (9.63–10.11) 12.70 (8.53–16.88) 12.15 (11.94–12.37) 12.50 (9.74–15.26) 60–64 9.20 (8.97–9.43) 18.85 (13.94–23.76) 11.06 (10.85–11.27) 18.66 (15.41–21.91) 65–69 5.09 (4.91–5.27) 12.70 (8.53–16.88) 7.60 (7.43–7.78) 18.84 (15.58–22.10)

70 and up 5.46 (5.28–5.64) 15.98 (11.39–20.58) 8.91 (8.72–9.10) 27.72 (23.98–31.45) BMI categories

Extremely thin 5.75 (5.56–5.93) 11.89 (7.82–15.95) 4.30 (4.16–4.43) 11.05 (8.44–13.67) Very thin 4.32 (4.16–4.49) 4.92 (2.20–7.63) 3.95 (3.82–4.08) 7.07 (4.93–9.20) Thin 9.12 (8.88–9.35) 10.66 (6.78–14.53) 9.16 (8.97–9.35) 12.32 (9.58–15.06) Normal 51.36 (50.96–51.76) 50.41 (44.14–56.68) 62.39 (62.07–62.71) 54.71 (50.56–58.86) Overweight 22.54 (22.21–22.88) 16.39 (11.75–21.04) 17.54 (17.29–17.79) 12.68 (9.91–15.46) Obese 6.91 (6.71–7.12) 5.74 (2.82–8.66) 2.67 (2.56–2.77) 2.17 (0.96–3.39) Education

Illiterate 45.30 (44.90–45.70) 56.15 (49.92–62.37) 17.02 (16.77–17.27) 14.49 (11.56–17.43) Primary school 35.03 (34.65–35.42) 33.20 (27.29–39.11) 37.84 (37.52–38.16) 48.01 (43.84–52.18) Middle school 13.77 (13.49–14.05) 7.38 (4.10–10.66) 29.41 (29.11–29.71) 24.46 (20.87–28.04) Secondary school 4.26 (4.10–4.42) 2.05 (0.27–3.83) 9.37 (9.18–9.56) 7.61 (5.40–9.82) College 1.64 (1.53–1.74) 1.23 ( 0.15–2.61) 6.36 (6.20–6.52) 5.43 (3.54–7.33) Religion

Hindu 82.02 (81.72–82.33) 81.56 (76.69–86.42) 76.84 (76.56–77.11) 73.37 (69.68–77.06) Muslim 6.65 (6.45–6.85) 5.74 (2.82–8.66) 16.04 (15.80–16.29) 14.13 (11.22–17.04) Buddhist 7.61 (7.40–7.82) 8.20 (4.75–11.64) 4.20 (4.07–4.34) 6.52 (4.46–8.58) Christian 3.15 (3.01–3.29) 4.10 (1.61–6.59) 2.49 (2.39–2.59) 5.43 (3.54–7.33) Others 0.56 (0.50–0.62) 0.41 ( 0.39–1.21) 0.42 (0.38–0.47) 0.54 ( 0.07–1.16) Mother tongue

Marathi 75.29 (74.94–75.64) 81.56 (76.69–86.42) 55.50 (55.18–55.83) 60.69 (56.61–64.76) Hindi 6.91 (6.71–7.12) 2.05 (0.27–3.83) 16.91 (16.66–17.15) 4.53 (2.79–6.26) Gujarathi 5.53 (5.35–5.72) 3.69 (1.32–6.05) 9.61 (9.42–9.81) 17.39 (14.23–20.55) Urdu 3.75 (3.60–3.90) 4.51 (1.90–7.11) 8.87 (8.68–9.06) 8.51 (6.19–10.84) Sindian 8.03 (7.81–8.25) 7.38 (4.10–10.66) 8.91 (8.72–9.10) 8.33 (6.03–10.64) Others 0.49 (0.43–0.54) 0.82 ( 0.31–1.95) 0.20 (0.17–0.23) 0.54 ( 0.07–1.16) Tobacco use

Non-user 40.26 (39.86–40.65) 26.64 (21.09–32.19) 30.10 (29.79–30.40) 18.12 (14.90–21.33) Smokeless tobacco 59.30 (58.90–59.69) 72.54 (66.94–78.14) 38.48 (38.16–38.81) 30.43 (26.60–34.27) Smoker 0.45 (0.39–0.50) 0.82 ( 0.31–1.95) 31.42 (31.11–31.73) 51.45 (47.28–55.62)

Trang 4

and the only one with a statistically signiﬁcant risk of increase in

comparison to never-using women, were extremely thin

smoke-less tobacco users Exclusion of such women who died within 2

years of enrolment actually increased the point estimate of the RR

The observed joint effects of smokeless tobacco use and

thinness (thin, very thin, extremely thin) in terms of HRs were

higher than the expected HRs (estimated separately by assuming

additive and multiplicative interactions, Table 2); indicating

synergistic interaction in women In contrast, the observed joint

effects were lower than the expected, indicating antagonistic

interaction in men Also, the observed joint effects of smoking and

thinness among men were between expected effects estimated by

using additive and multiplicative interactions

4 Discussion

Worldwide, there are two major risk factors underlying the

major causes of death, tobacco use and body habitus[16,22,41–

43] Their effects are now increasing rapidly, with high and

increasing prevalence rates of smoking and other forms of tobacco

use in many parts of the world[15,32] Simultaneously, there is a

virtual epidemic of overweight and obesity and chronic

under-nutrition in some of the same, as well as other, parts of the world

[44–46] If the current patterns persist, there will be about 1 billion

deaths from tobacco in the 21st century, compared with only about

0.1 billion (100 million) during the entire 20th century Of the total

55.9 million annual deaths worldwide, tobacco and BMI are

responsible for approximately 20% [22] Diseases, especially

chronic diseases, and injuries are almost always caused by

multiple risk factors [47,48] Estimating the joint effects of

multiple distal and proximal risks is particularly important

because many factors act through complicated pathways involving numerous feedback mechanisms[22] Additionally they often act

in combination with other intermediate factors that interact with one another and with either tobacco or body habitus or both[49– 51] Considerable, but not conclusive, evidence has been provided regarding the effect of smoking on BMI-mortality association[52] Numerous prospective epidemiologic studies have evaluated the relationship between cancer mortality and relative weight, commonly expressed as BMI[25,52,53] Inconsistencies evident in the literature may be due to the confounding effects of tobacco use and the fact that many cancers are slowly fatal conditions associated with wasting and cachexia [54,55] Thus, a low BMI may be reﬂective of either a healthy lifestyle, including well regulated energy intake and physical activity expenditure, or may

be secondary to tobacco use (especially smoking), illness, or both Understanding the effects of BMI on mortality requires considera-tion of tobacco use and the potential effects of weight loss owing to pre-existing illnesses[56] Most studies have either adjusted for, or stratiﬁed by, BMI in assessing the mortality association according

to tobacco use However, none of the large prospective studies have assessed the joint effect of tobacco use and BMI on cancer mortality This is a ﬁrst such an attempt from a large cohort study from India

In India, public health attention traditionally has focused on problems of undernutrition, though now there is evidence of a double burden of undernutrition as well as overnutrition, as was observed in this study[39]and in other places[57,58] Using MCS

we had reported low BMI as an independent risk factor for all cause

as well as cancer mortality[36] Tobacco is a public health concern worldwide However, patterns of tobacco use in India are quite different from those observed in the other developed countries In

Table 2

Adjusted relative risks a

(RRs) and associated 95% conﬁdence intervals (CIs) by sex, categories of body mass index [BMI = weight (kg)/height (m 2

)] and various tobacco habits, Mumbai Cohort Study.

Tobacco habits Relative risk of cancer death

BMI (kg/m 2 ) category Extremely

thin (<16.0)

Very thin (16.0 to <17.0)

Thin (17.0 to <18.5)

Normal (18.5 to <25.0)

Overweight (25.0 to <30.0)

Obese (30)

Men

Never users

Observed RR 4.35 (1.99–9.49) 3.07 (1.30–7.26) 1.55 (0.68–3.54) 1.11 (0.63–1.97) Reference 1.29 (0.37–4.47) Excluding ﬁrst 2 year deaths b 2.04 (0.66–6.34) 1.48 (0.42–5.27) 1.77 (0.72–4.35) 0.96 (0.50–1.85) 1.08 (0.24–4.83) Smokeless tobacco users c

Observed RR 2.43 (1.09–5.45) 2.28 (0.99–5.25) 1.89 (0.93–3.80) 1.78 (1.03–3.08) 1.50 (0.78–2.89) 1.76 (0.58–5.31) Expected RR

Excluding ﬁrst 2 year deaths b

1.85 (0.69–4.96) 0.96 (0.27–3.41) 1.39 (0.60–3.25) 1.59 (0.85–2.95) 1.22 (0.57–2.62) 2.18 (0.70–6.77) Smoker d

Excluding ﬁrst 2 year deaths b

4.85 (2.28–10.34) 6.18 (2.96–12.92) 4.38 (2.19–8.79) 2.97 (1.62–5.44) 3.04 (1.54–6.01) 1.49 (0.33–6.66) Women

Never users

Observed RR 0.88 (0.21–3.81) 1.15 (0.27–4.97) 1.32 (0.49–3.55) 1.01 (0.57–1.77) Reference 0.60 (0.20–1.75) Excluding ﬁrst 2 year deaths b

0.71 (0.09–5.50) 0.93 (0.12–7.17) 0.42 (0.05–3.24) 1.17 (0.58–2.34) 0.69 (0.20–2.47) Smokeless tobacco users c

Excluding ﬁrst 2 year deaths b 3.21 (1.49–6.95) 2.04 (0.78–5.30) 1.83 (0.81–4.13) 1.79 (0.94–3.42) 1.12 (0.52–2.41) 1.76 (0.68–4.53) a

Adjusted for age, education, religion, and mother tongue; note that all relative risk signiﬁcantly different from the referent are denoted by the use of bold font b

Excluding deaths occurring in the ﬁrst 2 years to reduce effect of weight loss or smoking cessation due to symptoms of disease.

c

Includes all types of smokeless tobacco products.

d

May include bidi plus cigarette smokers and may include smokers plus mixed (smoking and smokeless) users.

Trang 5

India, tobacco is used in a wide variety of forms[59] Use of all

forms of tobacco is associated with higher all cause mortality in the

Indian population [60,61] In additional to corroborating these

national ﬁndings, MCS also highlights an association of tobacco use

with cancer mortality[35] Previously, we reported that tobacco

use (smoking and smokeless) is also associated with low BMI[12]

The high prevalence of tobacco use[38,62]and its association with

low BMI[12]raises important questions about its impact on public

health in India, a country which has a high prevalence of low BMI

among adults[39,57,58]

In developed countries, smoking and excess body weight are

two of the most important risk factors implicated in chronic

diseases and premature death[52,63] Smoking is associated with

lower body mass; quitting smoking has been associated with

signiﬁcant amount of weight gain[17] In fact, smoking cessation

was estimated to be responsible for about one quarter of the

increase in prevalence of overweight among men in the USA during

the 1980s [64] In Mumbai, increased all-cause mortality was

observed not only among underweight and obese smokers, but also

among underweight and obese never tobacco users[37], and is

true for cancer mortality (Table 2) The population of India is not

immune to obesity [39,57,58,65–67] Additionally, patterns of

tobacco use[59,62]are very different than those observed in other

developed countries Speciﬁcally, smoking which is very common

among men is mainly in the form of bidi and followed by cigarette

smoking [59,62] Although smoking in women is very rare,

chewing of tobacco is quite common[38,59,62]

In India, nearly half of all rural adults and a quarter of urban

adults have a low BMI (i.e., <18.5 kg/m2)[57,58,68,69] Although

chronic energy deﬁciency due to inadequate diet may be the main

factor placing the population at risk of low BMI, factors other than

diet may play a signiﬁcant role in explaining the low BMI within

this population These factors may act directly (by affecting

appetite or other aspects of physiology) or indirectly (by

decreasing the purchasing power for food) We had reported that

all forms of tobacco use are associated with low BMI independent

of (i.e., after accounting for) age, education, mother tongue, and

religion in this same population[12] In this study, the risk of

cancer death among men increases progressively from never

tobacco users having a BMI 18.5–25.0 kg/m2to smokers having a

BMI <16.0 kg/m2, when compared with the reference group A

similar increasing pattern of risk was observed for overweight/

obese men; the risk of death increases from smokeless tobacco

users having a BMI 25.0 to <30 kg/m2to smokers having a BMI

30.0 kg/m2

In the Mumbai Cohort Study, we had reported that education

(generally used as a proxy for SES) was associated with both

tobacco use [38] and BMI [39] Additionally, low BMI was

associated with lower educational attainment while higher

education was associated with high-BMI, a pattern that was

observed for both men and women[39] Higher cancer mortality

was observed among both extremely thin and obese men and

women in Mumbai One possible explanation might be the

relationship between occupation and education In countries in

transition less-educated people tend to be in labor-intensive

occupations and people with higher education generally lead more

sedentary lifestyles This is in contrast to economically advanced

countries where lower education may be associated with higher

unemployment or low-paid jobs that are not necessarily

labor-intensive Thus, the increased risk of death among obese men

across different tobacco habits may be an indication of the

upcoming obesity epidemic in India The protective effect[36]that

we observed in overweight (but not obese) individuals may be an

artifact of where India stands in the demographic transition; i.e.,

these individuals may have been in normal weight categories just a

few years earlier[39]and this risk estimate is residual to the effect

of being normal weight for a long time during early and middle adulthood This ﬁnding raises important questions about the magnitude of the adverse joint impact of tobacco use and BMI on public health

In this population, infectious agents and pollution are the other environment factors that may play a role in the interaction between tobacco and body habitus Tobacco use[70–72]and poor nutrition [73] impair the immune system Hence, tobacco users are more susceptible to infectious agents For example, smoking is associated with higher relative risk of TB mortality and prevalence of active TB

in several studies in rural and urban India[61] In corroboration of this, increased mortality risk from TB was observed in MCS[74] Besides the direct physiological effect, tobacco use among the economically disadvantaged is known to reduce the resources available to purchase food, clothing, health, and education, all factors that contribute to poor nutritional status[75] This helps to explain why changes in the relationship between smoking and BMI vary with the secular shift toward afﬂuence[76]

The current study not only underscores the importance of joint effect of tobacco use and BMI on cancer mortality, but also demonstrates that the joint effect of smokeless tobacco use and thinness is antagonistic in men and synergistic in women Also, the effect is most prominent for smoker men having BMI <16.0 or having BMI 30.0

Limitations: The study has several limitations The cause of death information was obtained from local death registries that may be associated with imprecision On the other hand, the Mumbai Registry is one of the oldest and the most efﬁcient systems

of mortality ascertainment and therefore should be most reliable data from this country[35] Also, we excluded polling stations that served the upper-middle class and upper-class housing complexes because of security issues (i.e., they were essentially ‘‘gated communities’’) It should be noted, however, that they constituted

a small proportion when the cohort was formed during early 1990s We also excluded the very lowest SES individuals, consisting of footpath dwellers because they are not generally included in the electoral rolls; hence they would be very difﬁcult to follow up[12,30,35–39] Therefore, the cohort may not be truly representative of Mumbai population Most other limitations of this study, including loss to follow up due to high migration and its impact on tobacco and mortality association[35,74], limitations in BMI[12,36,39](e.g., very few obese individuals and few who have been overweight or obese through most of their adulthood) have been discussed in earlier publications

Conclusions: This study, in a large well designed and conducted cohort representing all but the very highest and lowest economic strata of Indian society, shows that all forms of tobacco use[35] and BMI[36](a proxy for nutritional status) have an independent,

as well as joint impact (Table 2) on cancer mortality Tobacco and undernutrition are known to be serious problems in India As the prevalence of obesity is low, obesity is thus far associated with modest excess mortality However, obesity appear to be increasing [39]and with it we can expect increases in rates of chronic diseases [65], including cancer[25,77] It appears that the joint impact is stronger than their individual effects; thus, tobacco control and improvements in nutritional status will be necessary to reduce cancer mortality in India

Data access: Every author had access to data used in the paper Financial support: This work was supported by funding from the International Agency for Research on Cancer, Lyon, France (Collaborative Research Agreement DEP/89/12), the Clinical Trial Service Unit of the University of Oxford, Oxford, United Kingdom, and the World Health Organization, Geneva, Switzerland Dr He´bert was supported by a USIA Fulbright Senior Research Fellowship for the 2008–9 academic year through the US Educational Foundation in India

Trang 6

Conﬂict of interest statement

None declared

Authors contributions: MS Pednekar who developed the concept

for this paper, supervised the ﬁeld work, was responsible for the

data management, data analyses and the statistical procedures

and tests, interpreted the results, conducted the literature search,

wrote the original draft of the paper, and interacted with

co-authors in subsequent drafts of the paper PC Gupta

conceptua-lized the study, evolved ﬁeld instruments and procedures,

directed the ﬁeld work, data management and data analyses,

interpreted the results and contributed towards writing of the

paper at all stages JR Hebert was responsible for helping in data

analysis, interpretation of results, and writing of the paper at all

stages

Acknowledgments

The authors appreciate the collaborative efforts of the

Inter-national Agency for Research on Cancer, Lyon, France, The Clinical

Trial Service Unit of the University of Oxford, Oxford, United

Kingdom, and the World Health Organization, Geneva,

Switzer-land The authors also are grateful for the co-operation of the

Municipal Corporation of Greater Mumbai (BMC) in providing

access to information on cause of death

References

[1] Day TA, Chi A, Neville B, Hebert JR Prevention of head and neck cancer Curr

Oncol Rep 2005;7(2):145–53.

[2] Freedman LS, Carroll RJ, Wax Y Estimating the relation between dietary intake

obtained from a food frequency questionnaire and true average intake Am J

Epidemiol 1991;134:310–20.

[3] U.S Department of Health and Human Services The Health Beneﬁts of

Smoking Cessation: A Report of the Surgeon General Atlanta, GA: Public

Health Service, Centers for Disease Control, Center for Chronic Disease

Pre-vention and Health Promotion, Ofﬁce on Smoking and Health Report No.: Publ.

No (CDC); 2004.

[4] Ward E, Jemal A, Cokkinides V, Singh GK, Cardinez C, Ghafoor A, et al Cancer

disparities by race/ethnicity and socioeconomic status CA A Cancer J Clin

2004;54(2):78–93.

[5] Sommer F, Klotz T, Schmitz-Drager BJ Lifestyle issues and genitourinary

tumours World J Urol 2004;21(6):402–13.

[6] Slattery ML, Samowtiz W, Ma K, Murtaugh M, Sweeney C, Levin TR, et al CYP1A1,

cigarette smoking, and colon and rectal cancer Am J Epidemiol 2004;160(9):842–

52.

[7] Twombly R Tobacco use a leading global cancer risk J Natl Cancer Inst

2003;95(1):11–2.

[8] Alberg AJ, Samet JM Epidemiology of lung cancer Chest 2003;123(Suppl.

1):21S–49S.

[9] Gupta PC, Hebert JR, Bhonsle RB, Murti PR, Mehta H, Mehta FS Inﬂuence of

dietary factors on oral precancerous lesions in a population-based case–

control study in Kerala, India Cancer 1999;85(9):1885–93.

[10] Demark-Wahnefried W, Aziz NM, Rowland JH, Pinto BM Riding the crest of the

teachable moment: promoting long-term health after the diagnosis of cancer J

Clin Oncol 2005;23(24):5814–30.

[11] Dansinger ML, Gleason JA, Grifﬁth JL, Selker HP, Schaefer EJ Comparison of the

atkins, ornish, weight watchers, and zone diets for weight loss and heart

disease risk reduction: a randomized trial JAMA 2005;293(1):43–53.

[12] Pednekar MS, Gupta PC, Shukla HC, Hebert JR Association between tobacco

use and body mass index in urban Indian population: implications for public

health in India BMC Public Health 2006;6:70.

[13] Desai PB Cancer control efforts in the Indian subcontinent Jpn J Clin Oncol

2002;(Suppl 32):S13–6.

[14] Hyman DJ, Pavlik VN, Taylor WC, Goodrick GK, Moye L Simultaneous vs

sequential counseling for multiple behavior change Arch Intern Med

2007;167(11):1152–8.

[15] Jha P, Chaloupka FJ, Corrao M, Jacob B Reducing the burden of smoking

world-wide: effectiveness of interventions and their coverage Drug Alcohol Rev

2006;25(6):597–609.

[16] Mathers CD, Loncar D Projections of global mortality and burden of disease

from 2002 to 2030 PLoS Med/Public Libr Sci 2006;3(11):e442.

[17] Schoenborn CA, Adams PF, Barnes PM, Vickerie JL, Schiller JS Health behaviors

of adults: United States, 1999–2001 Vital Health Stat 2004;10(219):1–79.

[18] Le Marchand L, Wilkens LR, Kolonel LN, Hankin JH, Lyu LC Associations of

sedentary lifestyle, obesity, smoking, alcohol use, and diabetes with the risk of

colorectal cancer Cancer Res 1997;57(21):4787–94.

[19] Emmons KM, Thompson B, Feng Z, Hebert JR, Heimendinger J, Linnan L Dietary intake and exposure to environmental tobacco smoke in a worksite popula-tion Eur J Clin Nutr 1995;49:336–45.

[20] Castro FG, Newcomb MD, McCreary C, Baezconde-Garbanati L Cigarette smo-kers do more than just smoke cigarettes Health Psychol 1989;8(1):107–29 [21] Pal SK, Mittal B Improving cancer care in India: prospects and challenges Asian Pac J Cancer Prev 2004;5(2):226–8.

[22] Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJL Comparative risk assessment collaborating G Selected major risk factors and global and regio-nal burden of disease Lancet 2002;360(9343):1347–60 [see comment] [23] Parsons AC, Shraim M, Inglis J, Aveyard P, Hajek P Interventions for preventing weight gain after smoking cessation Cochrane Database Syst Rev 2009;(1): CD006219.

[24] Wack JT, Rodin J Smoking and its effects on body weight and the systems of caloric regulation Am J Clin Nutr 1982;35(2):366–80.

[25] Moghaddam AA, Woodward M, Huxley R Obesity and risk of colorectal cancer: a meta-analysis of 31 studies with 70,000 events Cancer Epidemiol Biomark Prev 2007;16(12):2533–47.

[26] Dal Maso L, Zucchetto A, Talamini R, Serraino D, Stocco CF, Vercelli M, et al Prospective analysis of case–control studies on environmental factors and health study G Effect of obesity and other lifestyle factors on mortality in women with breast cancer Int J Cancer 2008;123(9):2188–94.

[27] Whiteman MK, Hillis SD, Curtis KM, McDonald JA, Wingo PA, Marchbanks PA Body mass and mortality after breast cancer diagnosis Cancer Epidemiol Biomarkers Prev 2005;14(8):2009–14.

[28] Loi S, Milne RL, Friedlander ML, McCredie MR, Giles GG, Hopper JL, et al Obesity and outcomes in premenopausal and postmenopausal breast cancer Cancer Epidemiol Biomarkers Prev 2005;14(7):1686–91.

[29] Makinen T, Karhunen P, Aro J, Lahtela J, Maattanen L, Auvinen A Assessment of causes of death in a prostate cancer screening trial Int J Cancer 2008;122(2):413–7.

[30] Pednekar MS, Hakama M, Hebert JR, Gupta PC Association of body mass index with all-cause and cause-speciﬁc mortality: ﬁndings from a prospective cohort study in Mumbai (Bombay), India Int J Epidemiol 2008;37(3):524–35 [31] Ma GX, Tan Y, Toubbeh JI, Edwards RL, Shive SE, Siu P, et al Asian tobacco education and cancer awareness research special population network A model for reducing Asian American cancer health disparities Cancer 2006;107(Suppl 8):1995–2005.

[32] Anderson P Global use of alcohol, drugs and tobacco Drug Alcohol Rev 2006;25(6):489–502.

[33] Sorensen G, Gupta PC, Pednekar MS Social disparities in tobacco use in Mumbai, India: the roles of occupation, education, and gender Am J Public Health 2005;95(6):1003–8.

[34] Stellman SD, Resnicow K Tobacco smoking cancer and social class, vol 138 IARC Scientiﬁc Publications; 1997 p 229–50.

[35] Gupta PC, Pednekar MS, Parkin DM, Sankarnarayanan R Tobacco associated mortality in Mumbai (Bombay), India: results of the Bombay Cohort Study Int

J Epidemiol 2005;34(6):1395–402.

[36] Pednekar MS, Hakama M, Hebert JR, Gupta PC Association of body mass index with all-cause and cause-speciﬁc mortality: ﬁndings from a prospective cohort study in Mumbai (Bombay), India Int J Epidemiol 2008;37(June 3):524–35.

[37] Pednekar MS, Gupta PC, Hebert JR, Hakama M Joint effect of tobacco use and body mass on all-cause mortality in Mumbai, India: results from a population-based cohort study Am J Epidemiol 2008;167:330–40.

[38] Gupta PC Survey of sociodemographic characteristics of tobacco use among 99,598 individuals in Bombay, India using handheld computers Tob Control 1996;5(2):114–20.

[39] Shukla HC, Gupta PC, Mehta HC, Hebert JR Descriptive epidemiology of body mass index of an urban adult population in western India J Epidemiol Community Health 2002;56(11):876–80.

[40] Hosmer DW, Lemeshow S Applied logistic regression New York City: John Wiley & Sons, 1989.

[41] Beaglehole R, Yach D Globalisation and the prevention and control of non-communicable disease: the neglected chronic diseases of adults Lancet 2003;362(9387):903–8 [see comment].

[42] Boyle P, Ariyaratne MAY, Barrington R, Bartelink H, Bartsch G, Berns A, et al Tobacco: deadly in any form or disguise Lancet 2006;367(9524):1710–2 [43] Lopez AD, Mathers CD Measuring the global burden of disease and epide-miological transitions: 2002–2030 Ann Trop Med Parasitol 2006;100(5– 6):481–99.

[44] de Onis M, Blossner M, Borghi E, Frongillo EA, Morris R Estimates of global prevalence of childhood underweight in 1990 and 2015 JAMA 2004;291(21): 2600–6.

[45] Sorensen TI The changing lifestyle in the world Body weight and what else? Diabetes Care 2000;23(Suppl 2):B1–4.

[46] Uauy R, Lock K Commentary: the importance of addressing the rise of over-weight and obesity—progress or lack of action during the last ﬁfty years? Int J Epidemiol 2006;35(1):18–20 [comment].

[47] Gopalan C Rising incidence of obesity, coronary heart disease and diabetes in the Indian urban middle class Possible role of genetic and environmental factors World Rev Nutr Diet 2001;90:127–43.

[48] Popkin BM The nutrition transition and obesity in the developing world J Nutr 2001;131:871S–3S.

[49] Hung RJ, Boffetta P, Brennan P, Malaveille C, Hautefeuille A, Donato F, et al GST, NAT, SULT1A1, CYP1B1 genetic polymorphisms, interactions with

Trang 7

environmental exposures and bladder cancer risk in a high-risk population Int

J Cancer 2004;110(4):598–604.

[50] Jeffcoat R Obesity—a perspective based on the biochemical interrelationship

of lipids and carbohydrates Med Hypotheses 2007;68(5):1159–71.

[51] Wogan GN, Hecht SS, Felton JS, Conney AH, Loeb LA Environmental and

chemical carcinogenesis Semin Cancer Biol 2004;14(6):473–86.

[52] Freedman DM, Ron E, Ballard-Barbash R, Doody MM, Linet MS Body mass

index and all-cause mortality in a nationwide US cohort Int J Obes (Lond)

2006;30(5):822–9.

[53] Warburton DE, Nicol CW, Bredin SS Health beneﬁts of physical activity: the

evidence CMAJ 2006;174(6):801–9.

[54] Ryan JL, Carroll JK, Ryan EP, Mustian KM, Fiscella K, Morrow GR Mechanisms

of cancer-related fatigue Oncologist 2007;12(Suppl 1):22–34.

[55] Douglas RG, Shaw JHF Metabolic effects of cancer Br J Surg 1990;77:246–54.

[56] Willett WC, Dietz WHC, Coldietz GA Guidelines for healthy weight N Engl J

Med 1999;341:427–34.

[57] Shetty PS Nutrition transition in India Public Health Nutr 2002;5(1A):175–

82.

[58] Vijayaraghavan K, Rao DH Diet & nutrition situation in rural India Indian J

Med Res 1998;108:243–53.

[59] Bhonsle RN, Murti PR, Gupta PC Tobacco habits in India In: Gupta PC, Hamner

JE, Murti PR, eds Proceedings of an International Symposium on Control of

Tobacco-Related Cancers and Other Diseases: Proceedings of an International

Symposium in Bombay, India, 15–19 January 1990 Oxford: Oxford University

Press, 1992: 25–46.

[60] Gupta PC, Bhonsel RB, Mehta FS, Pindborg JJ Mortality experience in relation

to tobacco and smoking habits from a 10-year follow-up study in Ernakulam

District Int J Epidemiol 1984;13:184–7.

[61] Gajalakshmi V, Peto R, Kanaka TS, Jha P Smoking and mortality from

tuber-culosis and other diseases in India: retrospective study of 43000 adult male

deaths and 35000 controls Lancet 2003;362(9383):507–15.

[62] Reddy KS, Gupta PC Report on tobacco control in India New Delhi: Ministry of

Health and Family Welfare, Government of India, 2004.

[63] The Surgeon General’s 1990 Report on The Health Beneﬁts of Smoking

Cessation Executive Summary MMWR Recomm Rep 1990;39 (RR-12):

i–xv,1–12.

[64] Flegal KM, Troiano RP, Pamuk ER, Kuczmarski RJ, Campbell SM The inﬂuence

of smoking cessation on the prevalence of overweight in the United States N Eng J Med 1995;333(18):1165–70.

[65] Misra A, Misra R, Wijesuriya M, Banerjee D The metabolic syndrome in South Asians: continuing escalation & possible solutions Indian J Med Res 2007;125(3): 345–54.

[66] Marita AR, Sarkar JA, Rane S Type 2 diabetes in non-obese Indian subjects is associated with reduced leptin levels: study from Mumbai, Western India Mol Cell Biochem 2005;275(1–2):143–51.

[67] Ghosh A Anthropometric, central obesity, metabolic and blood pressure variables in dyslipidaemic and non-dyslipidaemic adult Bengalee Hindu men of Calcutta, India Nutr Metab Cardiovasc Dis 2004;14(3):170–2 [68] Grifﬁths PL, Bentley ME The nutrition transition is underway in India J Nutr 2001;131:2629–700.

[69] India Nutrition Proﬁle New Delhi: Ministry of Human Resource Development, Government of India, 1998.

[70] Bamia C, Trichopoulou A, Lenas D, Trichopoulos D Tobacco smoking in relation

to body fat mass and distribution in a general population sample Int J Obes Rel Metabol Disord J Int Assoc Study Obes 2004;28(8):1091–6.

[71] Corwin EJ, Klein LC, Rickelman K Predictors of fatigue in healthy young adults: moderating effects of cigarette smoking and gender Biol Res Nurs 2002;3(4): 222–33.

[72] Moszczynski P, Zabinski Z, Moszczynski Jr P, Rutowski J, Slowinski S, Tabarowski

Z Immunological ﬁndings in cigarette smokers Toxicol Lett 2001;118(3):121–7 [73] Keusch GT The history of nutrition: malnutrition, infection and immunity J Nutr 2003;133(1):336S–40S.

[74] Pednekar MS, Gupta PC Prospective study of smoking and tuberculosis in India Prev Med 2007;44(6):496–8.

[75] Efroymson D, Ahmed S, Townsend J, Alam SM, Dey AR, Saha R, et al Hungry for tobacco: an analysis of the economic impact of tobacco consumption on the poor in Bangladesh Tob Control 2001;10(3):212–7 [see comment] [76] Marti B, Tuomilehto J, Korhonen HJ, Kartovaara L, Vartiainen E, Pietinen P, et al Smoking and leanness: evidence for change in Finland BMJ 1989;298(6683): 1287–90.

[77] Bray GA The underlying basis for obesity: relationship to cancer J Nutr 2002;132(Suppl 11):3451S–5.

Định dạng
Số trang	7
Dung lượng	202,85 KB