Body mass index (BMI), change in BMI Body mass index (BMI), change in BMI and cancers of the and cancers of the endometrium endometrium , , upper upper - - aerodigestive aerodigestive tract (UADT), and lung tract (UADT), and lung S. Lani Park S. Lani Park October 9, 2009 October 9, 2009
78
Embed
Weight change and Upper-aerodigestive tract (UADT) Cancers › epi › faculty › zhang › courses › epi295... · leanness. Rationale for leanness, UADT cancer association Temporal
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Body mass index (BMI), change in BMI Body mass index (BMI), change in BMI and cancers of the and cancers of the endometriumendometrium, ,
upperupper--aerodigestiveaerodigestive tract (UADT), and lungtract (UADT), and lung
S. Lani ParkS. Lani ParkOctober 9, 2009October 9, 2009
Presentation OutlinePresentation Outline
Overview of body weight and cancerOverview of body weight and cancerEndometrial cancerEndometrial cancerUADT cancerUADT cancerLung and UADT cancerLung and UADT cancer
With the increasing obesity epidemicWith the increasing obesity epidemic……
RenehanRenehan, 2008, 2008
Obesity and Cancer Obesity and Cancer biomechanismbiomechanism
Calle EE, Kaaks R. Nature, 2004
van van KruijsdijkKruijsdijk RCM C, CEBP, 2009RCM C, CEBP, 2009
Potential pathways Potential pathways linking obesity with linking obesity with cancercancer
Markers for adiposityMarkers for adiposity
Weight: adulthood weight gain (age 18 to Weight: adulthood weight gain (age 18 to 55) usually reflects an increase in body fat.55) usually reflects an increase in body fat.BMI (kg/mBMI (kg/m22): highly correlated with both ): highly correlated with both absolute body fat and percent body fat. absolute body fat and percent body fat. –– WHO categories: WHO categories:
Normal: 18.5Normal: 18.5--24.9 kg/m24.9 kg/m22
Moderate: 25Moderate: 25--29.9 kg/m29.9 kg/m22
Obese: Obese: ≥≥ 30 kg/m30 kg/m22
WaistWaist--HipHip--Ratio: measure for central Ratio: measure for central adiposityadiposity
Things to consider when Things to consider when studying BMI in cancerstudying BMI in cancer
Reverse causalityReverse causalityConfounding variablesConfounding variablesEthnic differencesEthnic differencesSelf report validity, information biasSelf report validity, information bias
Validity of selfValidity of self--reported height and weightreported height and weight
Hu FB, Obesity Epidemiology
Statistical ModelsStatistical Models Hu FB, Obesity Epidemiology
Factors affecting body weightFactors affecting body weight
Energy Balance Energy Balance The balance between energy taken in, The balance between energy taken in, generally by food and drink, and energy generally by food and drink, and energy expended. expended. Energy expenditure is influenced by genetics, Energy expenditure is influenced by genetics, body size and amount of muscle, and by body size and amount of muscle, and by physical activity. physical activity. While calories are probably the most critical While calories are probably the most critical element in maintaining your energy balance, element in maintaining your energy balance, other factors in your diet such as how much other factors in your diet such as how much fiber or calcium you eat may influence your fiber or calcium you eat may influence your energy expenditure and how much muscle energy expenditure and how much muscle and fat you have. and fat you have.
Vaughn P, www.cancer.gov
Example of SemiExample of Semi--quantitative FFQquantitative FFQDIETARY HABITS [Example, will need to adapt to each country] How often did you consume the following foods and beverages one year ago?
UnitUnit Food item Food item How many times per day, week, month, How many times per day, week, month, year? (mark one column only)year? (mark one column only)
Different measures of body Different measures of body weight change and cancerweight change and cancer
Weight differenceWeight differenceBMI differenceBMI difference% change (weight or BMI)% change (weight or BMI)–– [(measure at baseline minus measure at age [(measure at baseline minus measure at age
21) / measure at age 21]21) / measure at age 21]××100 100 % Average annual BMI change% Average annual BMI change–– % BMI change/(time between BMI measures)% BMI change/(time between BMI measures)
Multiethnic Cohort StudyMultiethnic Cohort StudyCohort established to study diet and chronic disease endpoints iCohort established to study diet and chronic disease endpoints in the US n the US among a multiethnic populationamong a multiethnic populationConsists of 215,251 adult men and women living in Hawaii and in Consists of 215,251 adult men and women living in Hawaii and in California.California.–– African Americans (16.3%) African Americans (16.3%) –– Latino (22.0%)Latino (22.0%)–– Japanese American (26.4%)Japanese American (26.4%)–– Native Hawaiian (6.5%) Native Hawaiian (6.5%) –– White (22.9%) White (22.9%) –– Other (5.8%).Other (5.8%).
Recruited in 1993Recruited in 1993--1996, Ages 451996, Ages 45--7575Males: 96,810 and Females: 118,441Males: 96,810 and Females: 118,441
Epidemiology of Endometrial CancerEpidemiology of Endometrial CancerThe fourth most common The fourth most common cancer among US cancer among US femalesfemales42,160 estimated new 42,160 estimated new cases for 2009.cases for 2009.Average 5 year survival Average 5 year survival rate is 83%.rate is 83%.White women have a White women have a higher incidence of higher incidence of endometrial cancer endometrial cancer compared to any other compared to any other ethnic group in the US.ethnic group in the US.However, AfricanHowever, African--American women have a American women have a mortality rate 79% greater mortality rate 79% greater than that of Caucasian than that of Caucasian women.women.
Risk factorsRisk factorsEstrogenEstrogen--only therapy in only therapy in postpost--menopausal womenmenopausal womenObesityObesityLate age at menopauseLate age at menopauseEarly age of menarcheEarly age of menarcheLimited epidemiological Limited epidemiological evidence for diabetes and evidence for diabetes and hypertension. hypertension.
Protective factorsProtective factorsUse of oral Use of oral contraceptivescontraceptivesIncreasing parityIncreasing paritySmokingSmoking
FollowFollow--up and Exclusion criteriaup and Exclusion criteriaFollowFollow--up:up:
FollowFollow--up time began at completion of baseline question, up time began at completion of baseline question, and continued until:and continued until:–– Diagnosis of endometrial cancerDiagnosis of endometrial cancer–– Death, or Death, or –– End of followEnd of follow--up (December 31, 2004)up (December 31, 2004)
Incident endometrial cancer cases were identified by Incident endometrial cancer cases were identified by record linkage to the Hawaii Tumor Registry, the Cancer record linkage to the Hawaii Tumor Registry, the Cancer Surveillance Program for Los Angeles County, and the Surveillance Program for Los Angeles County, and the California State cancer registry.California State cancer registry.
Exclusion criteria: Exclusion criteria: –– Had a hysterectomyHad a hysterectomy–– Previous history cancer other than Previous history cancer other than nonmelanomanonmelanoma skin cancerskin cancer–– missing variables of wmissing variables of weight at baseline or age 21, height, age at eight at baseline or age 21, height, age at
menarche, age at menopause, parity, oral contraceptive use, menarche, age at menopause, parity, oral contraceptive use, postmenopausal hormone therapy, smoking status, physical postmenopausal hormone therapy, smoking status, physical activity, and educationactivity, and education
Epidemiological data and Epidemiological data and Statistical AnalysisStatistical Analysis
Measures of height and weight were self reported Measures of height and weight were self reported on baseline questionnaireon baseline questionnaireCox proportional hazard models, age in days is Cox proportional hazard models, age in days is the underlying time variablethe underlying time variableIn preliminary analysis, we also adjusted for In preliminary analysis, we also adjusted for calories and physical activity and observed no calories and physical activity and observed no change in our findings. change in our findings. Final adjusted variables: education, body mass Final adjusted variables: education, body mass index, age at menarche, age at menopause, index, age at menarche, age at menopause, duration and type of hormone therapy, duration duration and type of hormone therapy, duration and oral contraceptive use, parity, smoking and oral contraceptive use, parity, smoking history, diabetes, hypertension and BMI at age 21.history, diabetes, hypertension and BMI at age 21.SAS v 9.1 (Cary, NC)SAS v 9.1 (Cary, NC)STATA v10 (College Station, TX)STATA v10 (College Station, TX)
Measures in adulthood more predictive of
endometrial cancer weight
Weight is a confounder
Similar findings as weight
Ethnic differences between BMI Ethnic differences between BMI and body fatand body fat
Hu FB, Obesity Epidemiology
Example: Physical activity and Example: Physical activity and endometrial cancerendometrial cancer
No. Cases RR1 (95% CI) RR2 (95% CI) RR3 (95% CI)
Physical activity (METs)3
Quartile 1: <1.4015 134 1.00 1.00 1.00Quartile 2: 1.4015 to
1 Age-adjusted RR. 2 RRs were adjusted for age, ethnicity, education, age at menarche, menopausal status, age at menopause, duration and type of hormone therapy, oral contraceptive use, parity, smoking history, diabetes, and hypertension. 3 Additionally adjusted for BMI at baseline (quartiles).
BMI change and endometrial cancer
BMI at
Age 21 BMI at baseline
BMI change from age 21 to baseline
Endometrial Cancer
??
Confounding variables
r= 0.74
BMI change in Endometrial cancerBody weightchange (%)
P trend <0.001 <0.001 0.044 0.0321RRs were adjusted for age, ethnicity, education, age at menarche, age at menopause, duration and type of hormone therapy, oral contraceptive use, parity, smoking history, diabetes, and hypertension.2 Model 1 and adjusting for BMI at age 21 (quartiles)3 Model 1 and adjusting for BMI at baseline (WHO categories)3 Model 1 and adjusting for BMI at age 21 (quartiles) and BMI at baseline (WHO categories)
Conclusions for BMI change and endometrial cancer
Limitations of study– Self reported weight
differential misclassification as a result of self-perceived weight– Cannot account for weight fluctuations– Selection bias from differing response rates by ethnic groups– No measures of WHR
Strengths of study– Prospective study design– Ability to control for a variety of confounding variables in a
multiethnic population.Conclusion– Adult BMI gain is associated with increased risk in endometrial
cancer– Risk may differ between ethnic groups, particular among
Japanese Americans– Postmenopausal hormone use and parity may modify the
endometrial cancer risk associated with adult BMI gain
The association between change in body The association between change in body mass index and upper mass index and upper aerodigestiveaerodigestive tract tract
cancers in the ARCAGE project: multicenter cancers in the ARCAGE project: multicenter casecase--control studycontrol study
Sungshim Lani Park1, 2, Yuan-Chin Amy Lee1, 2, Manuela Marron2, Antonio Agudo3, Wolfgang Ahrens4, Luigi Barzan5, Vladimir Bencko6, Simone Benhamou7, 8, Christine Bouchardy9, Cristina Canova10, Xavier Castellsague3, David I Conway11, 12, Claire M Healy13, Ivana Holcátová6, Kristina Kjaerheim14, Pagona Lagiou15, Raymond J Lowry16, Tatiana V Macfarlane17, Gary J Macfarlane17, Bernard E McCartan13, Patricia A McKinney11, 18, Franco Merletti19, Hermann Pohlabeln4, Lorenzo Richiardi19, Lorenzo Simonato10, Linda Sneddon16, Renato Talamini20, Dimitrios Trichopoulos21, Ariana Znaor22, Paul Brennan2, Mia Hashibe2
Manuscript in preparation/submissionManuscript in preparation/submission
Epidemiology of Upper Epidemiology of Upper AerodigestiveAerodigestive tract tract (UADT) Cancers(UADT) Cancers
Sites: Oral cavity, pharynx, Sites: Oral cavity, pharynx, hypopharynxhypopharynx larynx, larynx, esophagusesophagusNewly incident cases in the U.S.:Newly incident cases in the U.S.:–– Oral cavity and pharynx: 25,240 for males 10,480 for Oral cavity and pharynx: 25,240 for males 10,480 for
femalesfemales–– Larynx: 9,920 for males; 2370 for femalesLarynx: 9,920 for males; 2370 for females–– EsophagusEsophagus (both SCC and Adenocarcinoma)12,940 for (both SCC and Adenocarcinoma)12,940 for
males 3,530 for femalesmales 3,530 for femalesIn the U.S. 5 year survival is approximately 50%.In the U.S. 5 year survival is approximately 50%.EU estimated 180,000 new cases per year, causes EU estimated 180,000 new cases per year, causes 110,000 deaths (110,000 deaths (FerlayFerlay, 2004), 2004)Major risk factors include tobacco smoking and alcohol Major risk factors include tobacco smoking and alcohol drinking, however at least 25% of cases are attributed to drinking, however at least 25% of cases are attributed to other factors, such as low SES, poor nutrition, HPV, and other factors, such as low SES, poor nutrition, HPV, and leanness. leanness.
Prior Literature for UADT cancersPrior Literature for UADT cancersWeight measures from study entryAuthor Year Site Gender Measure OR (95% CI)Kabat 1994 Oral and pharyngeal Men Lowest quartile 3.09 (2.12, 4.5)
Women Lowest quartile 1.84 (0.92, 3.69)Franceschi 2001 Oral cavity and pharynx Men <22.7 kg/m2 3.78 (2.52, 5.67)
Women <22.7 kg/m2 1.99 (0.89, 4.44)Gaudet 2008 HNC Both < 18.5 kg/m2 2.11 (1.70, 2.61)Gallus 2003 Laryngeal Women >26 kg/m2 0.4 (0.2, 1.1)
Earlier Weight measuresAuthor Year Site Gender Time Measure OR (95% CI)Garavello 2006 Laryngeal Both Age 30 Lowest
quartile1.13 (0.71, 1.81)
Franceschi 2001 Oral cavity and pharynx
Men Age 30 <21.2 kg/m2 2.15 (1.34, 3.45)
Women Age 30 <21.2 kg/m2 1.21 (0.55, 2.64)Gaudet 2008 HNC Both Age 20 to 30 < 18.5 kg/m2 0.90 (0.68, 1.19)
Limitations of prior literatureLimitations of prior literature
Does not explain whether observed effect Does not explain whether observed effect is due to leanness as a result of weight is due to leanness as a result of weight change between these time points.change between these time points.Particularly since smoking and alcohol Particularly since smoking and alcohol drinking can affect body weight.drinking can affect body weight.
AlcoholAlcohol--related cancer and genetic related cancer and genetic susceptibility in Europe (ARCAGE) studysusceptibility in Europe (ARCAGE) study
Initialized by IARCInitialized by IARCStudy designed to investigate and clarify Study designed to investigate and clarify the role of smoking, drinking, and genetic the role of smoking, drinking, and genetic factors of UADT cancers.factors of UADT cancers.14 14 centerscenters and 10 European countriesand 10 European countriesRecruitment between 2002Recruitment between 2002--20052005
Cases and ControlsCases and Controls2,304 newly diagnosed cases (2,109 eligible 2,304 newly diagnosed cases (2,109 eligible squamoussquamouscell carcinoma (SCC) and 195 eligible noncell carcinoma (SCC) and 195 eligible non--SCC cases)SCC cases)Case eligibility: UADT cancer according to ICDCase eligibility: UADT cancer according to ICD--O3.O3.All ARCAGE centers used hospitalAll ARCAGE centers used hospital--based controls, based controls, except for UK center.except for UK center.2,227 controls, frequency matched to cases by sex, age 2,227 controls, frequency matched to cases by sex, age (5(5--year interval), and referral (or residence) area.year interval), and referral (or residence) area.Control eligibility include recent disease diagnosis, not Control eligibility include recent disease diagnosis, not related to alcohol, tobacco, or dietary practices, and related to alcohol, tobacco, or dietary practices, and short hospital stay.short hospital stay.
Epidemiological dataEpidemiological dataEpidemiological data collection was performed by Epidemiological data collection was performed by identical interviewers using a lifestyle questionnaireidentical interviewers using a lifestyle questionnaire3 measures of weight: time of recruitment, 2 years 3 measures of weight: time of recruitment, 2 years prior to recruitment, 30 years of age.prior to recruitment, 30 years of age.–– Weight at study entry (interviewer measured)Weight at study entry (interviewer measured)
Not available at UK sitesNot available at UK sites
–– Weights at 2 years prior and age 30 (self reported)Weights at 2 years prior and age 30 (self reported)Not available for Paris siteNot available for Paris site
% BMI change = % BMI change = (BMI at 2 years prior to study entry (BMI at 2 years prior to study entry --BMI at age BMI at age 30)/BMI at age 30 30)/BMI at age 30
Statistical AnalysesStatistical AnalysesExcluded: Excluded: –– 195 non195 non--SCC cases, due to limitation in sample sizeSCC cases, due to limitation in sample size–– 63 SCC cases and 54 controls for missing data.63 SCC cases and 54 controls for missing data.
Adjustment variables: Adjustment variables: centercenter, education, sex, age, fruit , education, sex, age, fruit and vegetable intake, tobacco status/frequency, and and vegetable intake, tobacco status/frequency, and alcohol frequency.alcohol frequency.Performed using SAS v9.1 (Cary, NC)Performed using SAS v9.1 (Cary, NC)Unconditional logistic regression model.Unconditional logistic regression model.
Baseline characteristics for ARCAGESCC UADT Mean distribution of BMI 2 years ago Mean % BMI change (age 30 to 2 years prior)Ca Co Ca Co Ca Co
Total (N) 2048 2173 25.1 26.3 4.2 9.0Center % % mean (SD) mean (SD) mean (SD) mean (SD)
*Adjusted for center, education, sex, age, fruit and vegetable intake, tobacco status/frequency, and alcohol frequency.** UK and Dublin centers not included in the model† Paris not included in the model.
Associations between BMI, BMI change and UADT cancers, stratified by subsite
Oral-oropharyngeal Hypopharynx and larynx EsophagealCa Co OR* 95% CI Ca Co OR* 95% CI Ca Co OR* 95% CI
Associations between BMI change and UADT cancers, stratified by risk factors, con’t.
Limitations, Strengths and ConclusionsLimitations– Self-reported measures of weight– No WHR– Use of semi-food frequency questionnaire limited the
possibility to account for energy balanceStrengths– large sample size– BMI measures at different time points– the potential to adjust for a variety of smoking and
drinking measures. Conclusions– BMI gain is inversely associated with UADT cancers – This may be due to alterations in smoking and/or
drinking behaviors– BMI change may also serve as a potential indicator of
early tumor development among a smoking and alcohol drinking population.
The association between change The association between change in body mass index and lung and in body mass index and lung and upper aeroupper aero--digestive tract cancersdigestive tract cancers
Sungshim Lani Park1, Yuan-Chin Amy Lee1, Yan Cui1, Mia Hashibe2, Sander Greenland1,3, Hal Morgenstern5, Donald P. Tashkin6, Jenny T. Mao6, Wendy Cozen6, Thomas M. Mack6, Zuo-Feng Zhang1, 4
Manuscript in preparationManuscript in preparation
Epidemiology of Lung CancerEpidemiology of Lung CancerHighest incidence secondary to prostate and breast Highest incidence secondary to prostate and breast cancercancer–– 116,090 estimated new male cases for 2009116,090 estimated new male cases for 2009–– 103,350 estimated new female cases for 2009103,350 estimated new female cases for 2009
The highest mortality rate The highest mortality rate –– Estimated deaths for males: 88,900 Estimated deaths for males: 88,900 –– Estimated deaths for females: 70,490 Estimated deaths for females: 70,490
Poorest survivalPoorest survival–– FiveFive--year survival rate of 15% from 1996year survival rate of 15% from 1996--2004.2004.11
1 Jemal A, et al., ACS Annual report, Cancer Statistics 2009.
Risk factors for lung and UADT cancersRisk factors for lung and UADT cancersLungLung cancerscancers
Tobacco smokingTobacco smokingGenderGenderPassive smokePassive smokeRadon gasRadon gasIndoor air pollutionIndoor air pollutionPoor dietPoor diet(Alcohol drinking) (Alcohol drinking) (HPV)(HPV)
Prior studies investigating weight Prior studies investigating weight change and lung cancerschange and lung cancers
Author year journalstudy design study
weight measure
difference RR LL UL
Kabat 2007 epidemiology cohort
Canadian National Breast Screening Study
1 year prior
18 years women 50 lbs+ 0.5 0.22 1.12
Kabat 2008 AJE cohort WHI baseline 18 years
women never smokers 30 lbs+ 0.76 0.45 1.3
Kabat 2008 AJE cohort WHI baseline 18 years
women former smokers 30 lbs+ 0.76 0.56 1.03
Kabat 2008 AJE cohort WHI baseline 18 years
women current 30 lbs+ 0.63 0.43 0.93
Los Angeles Lung and UADT cancer studyLos Angeles Lung and UADT cancer studyPopulation Based CasePopulation Based Case--Control StudyControl Study–– 611 Cases and 1040 controls611 Cases and 1040 controls–– Newly diagnosed cases obtained using the rapid ascertainment Newly diagnosed cases obtained using the rapid ascertainment
system of the Cancer Surveillance Program for L.A. County, system of the Cancer Surveillance Program for L.A. County, administered by the Keck School of Medicine and Norris administered by the Keck School of Medicine and Norris Comprehensive Cancer Center (USC).Comprehensive Cancer Center (USC).
–– Controls were identified using a formal algorithm providing a liControls were identified using a formal algorithm providing a list st of households within the neighborhood of each individual case. of households within the neighborhood of each individual case.
–– Controls were matched to cases by age and gender.Controls were matched to cases by age and gender.Study CriteriaStudy Criteria–– Subjects were residents of LA County at time of recruitment Subjects were residents of LA County at time of recruitment
(1999(1999--2004).2004).–– During the study period were 18During the study period were 18--65 years of age.65 years of age.–– Subjects were able to speak either English or Spanish.Subjects were able to speak either English or Spanish.
Epidemiological data and statistical analysisEpidemiological data and statistical analysisEpidemiological dataEpidemiological data
Epidemiological data collection was performed by Epidemiological data collection was performed by identical interviewers using a lifestyle questionnaireidentical interviewers using a lifestyle questionnaire2 measures of weight: 1 year prior to recruitment 2 measures of weight: 1 year prior to recruitment and weight at age 21. and weight at age 21. % BMI change = (BMI at 2 years prior to study entry % BMI change = (BMI at 2 years prior to study entry -- BMI at age 30)/ BMI at age 30BMI at age 30)/ BMI at age 30
Statistical analysisStatistical analysisAnalyzed using unconditional logistic regression.Analyzed using unconditional logistic regression.Matching was broken allowing us to include cases Matching was broken allowing us to include cases without match controls.without match controls.SAS v9.1SAS v9.1Adjustment variables: gender, smoking status, packAdjustment variables: gender, smoking status, pack--years, education, race, calories, and age (and years, education, race, calories, and age (and drinking for UADT cancers)drinking for UADT cancers)
Table I. Baseline characteristicsTable I. Baseline characteristicsVariableVariable Lung cancerLung cancer SCC UADT cancerSCC UADT cancer ControlsControls
Table 2. Association between anthropometric measures and lung cancer, stratified by subsite
Adjustment variables: smoking status packAdjustment variables: smoking status pack years education race calories and ageyears education race calories and age
Association between BMI change and lung cancer, stratified by riAssociation between BMI change and lung cancer, stratified by risk factorssk factors
<−5% (weight loss) −5% to +5% 5% to <25% ≥25% PtrendBMI at age 21Tertile 1
Adjustment variables: smoking status, packAdjustment variables: smoking status, pack--years, education, race, calories, and ageyears, education, race, calories, and age
Association between BMI change and lung cancer, stratified by riAssociation between BMI change and lung cancer, stratified by risk factors, sk factors, concon’’tt..<−5% (weight loss) −5% to +5% 5% to <25% ≥25% Ptrend
Adjustment variables: smoking status, packAdjustment variables: smoking status, pack--years, education, race, calories, and ageyears, education, race, calories, and age
Association between BMI, BMI change and UADT cancersAssociation between BMI, BMI change and UADT cancersVariables Ca/Co Crude OR (95% CI) Adjust OR1(95% CI)BMI at age 21
Adjustment variables: gender, smoking status, packAdjustment variables: gender, smoking status, pack--years, education, drinking, race, calories, and ageyears, education, drinking, race, calories, and age
Association between BMI, BMI change and UADT cancer stratified bAssociation between BMI, BMI change and UADT cancer stratified by gendery genderMale Females
Variables Ca/Co COR (95% CI) Adj OR1 (95% CI) Ca/Co COR (95% CI) Adj OR1 (95% CI)BMI at age 21
Adjustment variables: gender smoking status packAdjustment variables: gender smoking status pack years education drinking race calories and ageyears education drinking race calories and age
Adjustment variables: gender, smoking status, packAdjustment variables: gender, smoking status, pack--years, education, drinking, race, calories, and ageyears, education, drinking, race, calories, and age
Discussion for associations between BMI Discussion for associations between BMI change and lung and UADT cancerschange and lung and UADT cancers
BMI gain is inversely associated with UADT and lung BMI gain is inversely associated with UADT and lung cancers cancers Potential reasons includePotential reasons include–– Residual confoundingResidual confounding
However, adjusted for other smoking and drinking variablesHowever, adjusted for other smoking and drinking variables–– Reverse causalityReverse causality–– Biological mechanismBiological mechanism——Smoking/drinking, weight loss antiSmoking/drinking, weight loss anti--
oxidative pathwayoxidative pathway88--OHdG levels slightly higher in smokers than nonsmokersOHdG levels slightly higher in smokers than nonsmokers88--OHdG levels increase per unit decrease in BMI (OHdG levels increase per unit decrease in BMI (MizoueMizoue, , 2007)2007)Alcohol intake can induce free radical species in mice Alcohol intake can induce free radical species in mice leukocytesleukocytes
–– LowLow--BMIBMI--highhigh--waist circumference paradoxwaist circumference paradoxObservation that smokers and possibly alcohol drinkers, tend Observation that smokers and possibly alcohol drinkers, tend to have lower BMI but greater central adiposity.to have lower BMI but greater central adiposity.
Limitations, Strengths and ConclusionsLimitations– Self-reported measures of weight– No WHR– Measure at 1 year prior to interview may not be early
enoughStrengths– Large population-based sample size– the potential to adjust for a variety of smoking and
drinking measures. Conclusions– BMI gain is inversely associated with lung and UADT
cancers – This may be due to alterations in smoking and/or
drinking behaviors– BMI change may also serve as a potential indicator of
early tumor development among a smoking and/or alcohol drinking population.
Molecular approachMeasure inflammatory markersMeasure IGF1 and IGFBP3 levelsGenotype sometimes correlates to phenotype (Cheng I, 2007; Al-ZahraniA, 2006)
Calle EE, Kaaks R. Nature, 2004
In SummaryIn SummaryThe study of anthropometric measures The study of anthropometric measures is necessary in cancer.is necessary in cancer.These measures are fairly valid, costThese measures are fairly valid, cost--effective means to measure adiposity.effective means to measure adiposity.As of now weight maintenance is an As of now weight maintenance is an effective means of prevention against effective means of prevention against obesity related cancerobesity related cancer
Future directionsInvestigating genetic and biologic relationships Investigating genetic and biologic relationships between obesity and cancer would be beneficial.between obesity and cancer would be beneficial.For questionnaire design, it would be ideal to have more than 1 measure of adult weightOne could also look into changes in adult height Measures of WHR would assist in understanding Measures of WHR would assist in understanding the observed associations between BMI change the observed associations between BMI change and cancer.and cancer.