Racial and Ethnic Differences in Health Behaviors Among Cancer Survivors

Racial and Ethnic Differences in Health Behaviors Among Cancer Survivors Pratibha Nayak, PhD, Raheem J. Paxton, PhD, Holly Holmes, MD, Hoang Thanh Nguyen, PhD, Linda S. Elting, DrPH Introduction: Previous studies of health behaviors of adult cancer survivors have not adequately examined racial and ethnic differences because of small sample sizes. A national data set was used to examine differences in health behaviors between cancer survivors and controls and between racial and ethnic groups among survivors. Methods: The study analyzed 2009 Behavioral Risk Factor Surveillance System survey data in 2012–2014. Descriptive statistics were used to examine differences in health behaviors between cancer survivors and controls aged 20–64 years. Multivariable analysis was conducted to examine associations between race/ethnicity (white, African American, Hispanic, Asian, or Native American) and health behaviors (BMI, fruit and vegetable consumption, physical activity, and smoking status) while adjusting for demographic and medical characteristics. Significance was set at po0.01. Results: Compared with controls (n¼245,283), cancer survivors (n¼17,158) had higher prevalence rates for overweight/obese status (67% vs 65%); not meeting physical activity recommendations (53% vs 49%); and current smoking status (22% vs 20%). In the multivariable model, diet and smoking behavior differed across cancer status. African American (AOR¼1.95) and Hispanic (AOR¼2.06) survivors were more likely to have higher BMI than white survivors. African American survivors (AOR¼1.6) were less likely to meet physical activity guidelines. Native American (AOR¼3.08) and multiracial (AOR¼1.74) survivors were more likely to be current smokers than non-Hispanic white survivors. Conclusions: This study suggests that racial and ethnic differences exist in the adoption of recommended health behaviors; future research should identify factors to reduce these differences. (Am J Prev Med 2015;48(6):729–736) & 2015 American Journal of Preventive Medicine Introduction C ancer survivors (CS) face an increased risk of developing comorbid conditions such as cardiovascular disease, diabetes, and second cancers, leading to premature mortality and morbidity compared with age- and sex-matched controls.1,2 This burden is higher for Hispanic and African American survivors than for non-Hispanic white survivors.3 Engaging in recommended health behaviors (e.g., recommended diet and From the Department of General Internal Medicine (Nayak, Holmes), Department of Health Services Research (Nguyen, Elting); University of Texas MD Anderson Cancer Center; Health Promotion and Behavioral Sciences (Nayak), University of Texas Health Science Center, Houston; and the Department of Behavioral and Community Health (Paxton), University of North Texas Health Science Center, Fort Worth, Texas Address correspondence to: Pratibha Nayak, PhD, Department of General Internal Medicine, Unit 1465, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston, TX 77030. E-mail: pnayak@mdanderson.org. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2014.12.015 & 2015 American Journal of Preventive Medicine physical activity [PA]) may prevent adverse cancer sequelae.4–6 However, limited data exist on engagement of CS in these behaviors.7,8 This study used data from the 2009 Behavioral Risk Factor Surveillance System (BRFSS) to examine racial and ethnic differences that may exist for these health behaviors. The findings of this study suggest opportunities for improving adherence to health behaviors and reducing racial and ethnic disparities among CS and the general population. Methods Study Sample The BRFSS9 is an annual, random-digit-dial telephone survey that estimates the health behaviors of non-institutionalized U.S. residents aged 20–64 years.10 Respondents reporting a history of cancer were selected as CS (n¼17,158), and respondents with no history of cancer served as controls (n¼245,283). Those diagnosed with cancer o1 year prior and those diagnosed with nonmelanoma skin cancer were excluded.  Published by Elsevier Inc. Am J Prev Med 2015;48(6):729–736 729 730 Nayak et al / Am J Prev Med 2015;48(6):729–736 Table 1. Characteristics of Cancer Survivors and Controls Aged 20–64 Years Cancer survivors Controls Characteristic Sample size Population estimate (%) Sample size Population estimate (%) Total 17,158 7,449,459 245,283 158,169,193 Age (years) o0.0001 20–44 2,791 1,992,723 (26.75) 93,251 92,408,596 (58.42) 45–64 14,367 5,456,735 (73.25) 152,032 65,760,597 (41.58) Sex Male Female o0.0001 4,648 2,524,226 (33.88) 97,270 79,321,919 (50.15) 12,510 4,925,232 (66.12) 148,013 78,847,274 (49.85) Education 0.3609 rHigh school 10,881 4,657,218 (62.52) 151,030 97,830,440 (61.85) 4High school 6,277 2,792,240 (37.48) 94,253 60,338,753 (38.15) Race/ethnicity White o0.0001 14,447 5,865,448 (78.76) 192,082 108,138,785 (68.38) 1,060 623,956 (8.38) 20,944 16,288,405 (10.30) Hispanic 700 538,691 (7.23) 17,508 22,658,948 (14.33) Asian 114 85,484 (1.15) 4,762 5,513,280 (3.49) Native American 314 94,516 (1.27) 3,815 1,653,068 (1.04) Multiracial 520 239,045 (3.21) 6,117 3,879,572 (2.45) African American Comorbidity count o0.0001 0 4,143 2,138,196 (28.70) 104,619 82,304,792 (52.04) 1–2 9,193 3,867,043 (51.91) 110,821 62,781,734 (39.69) 42 3,822 1,444,220 (19.39) 29,841 13,081,665 (8.27) BMI o0.0001 Normal 5,465 2,376,861 (32.87) 79,099 52,946,699 (34.73) Overweight 5,736 2,462,764 (34.06) 84,411 55,214,481 (36.22) Obese 5,339 2,390,553 (33.06) 72,035 44,291,121 (29.05) Fruit and vegetable intake Meets recommendations Does not meet recommendations o0.0001 4,464 1,986,995 (26.67) 56,708 36,394,961 (23.02) 12,691 5,462,068 (73.33) 188,441 121,695,767 (76.98) Physical activity o0.0001 Meets physical activity guidelines 7,811 3,430,802 (47.27) 119,023 78,980,713 (51.02) Does not meet physical activity guidelines 8,832 3,827,286 (52.73) 120,007 75,810,366 (48.98) 0.0047 Smoking status Former/never smoker Current smoker p-value 13,277 5,782,713 (78.03) 196,513 125,653,921 (79.75) 3,836 1,628,508 (21.97) 47,956 31,907,286 (20.25) Note: Boldface indicates statistical significance (po0.01). www.ajpmonline.org June 2015 Table 2. Prevalence of Health Behaviors Among Cancer Survivors by Race and Ethnicity White Health behavior Sample size Population estimate (%) African American Sample size Population estimate (%) Hispanic Sample size Population estimate (%) Asian Sample size Population estimate (%) Native American Sample size Population estimate (%) Multiracial Sample size Population estimate (%) BMI p-value o0.0001 4,799 1,994,889 (35) 194 121,568 (20) 185 114,046 (22) 59 54,287 (64) 94 31,691 (34) 134 60,380 (26) Overweight 4,844 1,983,387 (35) 344 197,120 (33) 243 164,754 (31) 38 19,547 (23) 88 25,364 (28) 176 70,275 (31) Obese 4,286 1,710,775 (30) 482 286,849 (47) 241 248,331 (47) 15 11,487 (13) 125 34,686 (38) 190 98,426 (43) Fruit and vegetable intake 0.6698 Meets recommendations 3,745 1,531,706 (26) 239 165,406 (27) 199 157,472 (29) 33 27,133 (32) 80 27,083 (29) 166 76,923 (32) Does not meet recommendations 10,699 4,333,346 (74) 821 458,550 (73) 501 381,219 (71) 81 58,351 (68) 234 67,434 (71) 354 162,123 (68) Nayak et al / Am J Prev Med 2015;48(6):729–736 Normal 0.0002 Physical activity Meets physical activity guidelines 6,767 2,806,469 (49) 318 204,581 (34) 303 228,632 (43) 46 31,168 (39) 126 36,022 (39) 248 121,611 (52) Does not meet physical activity guidelines 7,272 2,913,829 (51) 698 397,123 (66) 376 300,811 (57) 64 47,888 (61) 174 56,150 (61) 248 111,485 (48) 731 (continued on next page) Nayak et al / Am J Prev Med 2015;48(6):729–736 o0.0001 181 48,599 (52) 88,407 (37) BMI was used to determine whether participants were overweight (25–29.9) or obese (Z30).11 Consuming more than five servings of fruits and vegetables (F&V) per day was considered meeting F&V intake recommendations.12–14 Those who engaged in vigorous PA for at least 20 minutes on at least 3 days per week or moderate PA for at least 30 minutes on at least 5 days per week were considered to have met PA guidelines.14–16 Smoking status was grouped into two categories: current smokers and former or never smokers.14 The comorbidity score was a summative score calculated from the presence of diabetes, hypertension, arthritis, hyperlipidemia, and coronary vascular disease14 and was divided into three categories: 0, 1–2, or 42.17 Self-reported demographic information was included for age, sex, education, and race and ethnicity. The Strengthening the Reporting of Observational Studies in Epidemiology guidelines18 were used in reporting this cross-sectional study. Statistical Analysis 143 338 45,343 (48) All analyses were performed in SAS, version 9.3, with significance set at po0.01 a priori. The data were summarized within each group using descriptive statistics that included chi-square tests (i.e., PROC SURVEYFREQ) and ANOVA. Multivariable logistic regression (i.e., PROC SURVEYLOGISTIC) models were used to examine differences in health behaviors between racial and ethnic groups, and another model was used among CS while adjusting for age, sex, educational attainment, time since diagnosis, comorbidity score, and BMI (except in the model where BMI was the outcome of interest). Analyses were conducted in 2012–2014. 14 84,802 (16) 5,086 (6) 100 Results A greater number of CS compared to controls were aged between 45 and 64 years, were female, identified as nonHispanic white, and reported having one or more comorbid condition (Table 1). Compared to controls, CS had slightly higher rates of meeting F&V recommendations, but lower rates for meeting PA, weight status, and smoking recommendations (all po0.01). Note: Boldface indicates statistical significance (po0.01). 145 106,763 (17) 3,127 Current smoker 1,294,851 (22) 226 550 827 4,556,935 (78) 11,290 Former/ never smoker Smoking status Sample size 514,928 (83) Sample size Sample size Population estimate (%) Health behavior 432,286 (84) Population estimate (%) Sample size 80,398 (94) Population estimate (%) 169 Sample size Population estimate (%) Sample size 150,505 (63) Population estimate (%) Measures Participant Characteristics Population estimate (%) Asian Hispanic African American White Table 2. Prevalence of Health Behaviors Among Cancer Survivors by Race and Ethnicity (continued) Native American Multiracial p-value 732 Differences Between Racial and Ethnic Groups in Health Behaviors The prevalence of health behaviors among CS by racial and ethnic groups is shown in Table 2. Rates for being overweight or obese were higher in African Americans and Hispanics. Among all CS, 73% did not meet F&V intake recommendations, with no racial and ethnic differences. About 66% of African Americans did not meet PA guidelines, followed by Asians (61%); Native Americans (61%); and Hispanics (57%). Multiracial survivors had the highest prevalence rates for meeting PA guidelines (52%), and Native Americans (52%) had the highest smoking rates. Racial and ethnic differences were significant across all four health behaviors (Table 3, Model 1). Compared to www.ajpmonline.org Nayak et al / Am J Prev Med 2015;48(6):729–736 733 Table 3. Multivariable Logistic Regression Models for Health Behaviors Among Cancer Survivors and Controls Overweight/ obese BMIa (AOR [99% CI]) Characteristic Model 1 Did not meet fruit and vegetable intake recommendationsb (AOR [99% CI]) Did not meet physical activity guidelinesb (AOR [99% CI]) Current smokerb (AOR [99% CI]) Multivariable logistic regression models examining adherence to health behaviors (BMI, diet, physical activity, and smoking) among cancer survivors and participants with no history of cancer Cancer diagnosis No ref ref ref ref 0.94 (0.86, 1.02) 0.88 (0.81, 0.97) 1.05 (0.97, 1.14) 1.13 (1.02, 1.24)** ref ref ref ref African American 1.64 (1.48, 1.82)*** 095 (0.86, 1.05) 1.39 (1.28, 1.51)*** 0.92 (0.83, 1.01) Hispanic 1.51 (1.37, 1.66)*** 0.89 (0.81, 0.98)** 1.26 (1.16, 1.37)*** 0.60 (0.53, 0.67)*** Asian 0.45 (0.39, 0.53)*** 0.85 (0.71, 1.02) 1.89 (1.61, 2.23)*** 0.61 (0.42, 0.89)*** Native American 1.29 (1.01, 1.64)** 0.85 (0.68, 1.07) 0.91 (0.75, 1.10) 1.92 (1.56, 2.34)*** Multiracial 0.95 (0.81, 1.11) 0.73 (0.62, 0.86)*** 0.89 (0.77, 1.03) 1.36 (1.14, 1.62)*** Yes * Race/ethnicity White Model 2 Multivariable logistic regressions comparing the health behaviors (BMI, diet, physical activity and smoking) among cancer survivors across racial groups Age (years) 20–44 ref ref ref ref 45–64 0.89 (0.71, 1.12) 0.82 (0.65, 1.04) 1.10 (0.88, 1.37) 0.47 (0.37, 0.60)*** ref ref ref ref 1.97 (1.61, 2.41)*** 1.75 (1.44, 2.14)*** 0.85 (0.71, 1.03) 0.83 (0.65, 1.05) rHigh school ref ref ref ref 4High school 0.70 (0.59, 0.84)*** 0.51 (0.43, 0.61)*** 0.76 (0.64, 0.89)*** 0.28 (0.22, 0.36)*** 1–2 ref ref ref ref 2–5 0.95 (0.71, 1.26) 0.96 (0.72, 1.29) 0.88 (0.67, 1.15) 0.92 (0.67, 1.27) 5–10 0.99 (0.76, 1.29) 1.04 (0.78, 1.37) 1.08 (0.84, 1.41) 1.37 (1.00, 1.87) 410 0.87 (0.69, 1.10) 1.05 (0.82, 1.36) 0.97 (0.77, 1.22) 1.44 (1.10, 1.88)*** Sex Female Male Education Time since diagnosis (continued on next page) June 2015 Nayak et al / Am J Prev Med 2015;48(6):729–736 734 Table 3. Multivariable Logistic Regression Models for Health Behaviors Among Cancer Survivors and Controls (continued) Characteristic Overweight/ obese BMIa (AOR [99% CI]) Did not meet fruit and vegetable intake recommendationsb (AOR [99% CI]) Did not meet physical activity guidelinesb (AOR [99% CI]) Current smokerb (AOR [99% CI]) ref ref ref ref Comorbidity count 0 1–2 2.01 (1.65, 2.44) *** 0.98 (0.79, 1.21) 1.26 (1.03, 1.53) 1.48 (1.14, 1.92)** 42 5.37 (4.09, 7.05)*** 0.98 (0.75, 1.29) 1.79 (1.38, 2.31)*** 2.11 (1.56, 2.86)*** ref ref ref BMI Normal ** NA — 1.35 (1.11, 1.63) *** 0.73 (0.58, 0.92)** Overweight — 1.17 (0.96, 1.44) Obese — 1.33 (1.06, 1.67)** 2.01 (1.62, 2.49)*** 0.55 (0.43, 0.72)*** ref ref ref Race/ethnicity White ref ** ** 0.87 (0.59, 1.28) 1.60 (1.08, 2.38) 0.68 (0.46, 1.00) 2.06 (1.30, 3.26)*** 0.76 (0.50, 1.15) 1.14 (0.76, 1.72) 0.62 (0.37, 1.04) Asian 0.37 (0.12, 1.12) 1.08 (0.30, 3.82) 1.99 (0.58, 6.77) 0.29 (0.09, 0.91)** Native American 0.93 (0.51, 1.70) 0.81 (0.42, 1.56) 1.36 (0.75, 2.46) 3.08 (1.62, 5.87)*** Multiracial 1.31 (0.70, 2.43) 0.60 (0.34, 1.05) 0.77 (0.46, 1.31) 1.74 (1.08, 2.82)** African American 1.95 (1.24, 3.07) Hispanic Note: Boldface indicates statistical significance. a Adjusted for age, sex, education, race/ethnicity, and number of comorbidities. b Adjusted for age, sex, education, race/ethnicity, number of comorbidities, and BMI. n po0.01; nnpo0.001; nnn po0.0001. controls, CS were more likely to meet F&V recommendations but were less likely to meet smoking recommendations. No differences were observed for being overweight or obese and for meeting PA guidelines. Among CS, African Americans and Hispanics were more likely to report higher BMI than non-Hispanic whites (all po0.01, Table 3). African Americans were also less likely to follow PA guidelines than non-Hispanic whites (po0.01). Native American and multiracial survivors were more likely to be current smokers than non-Hispanic whites. Asians were less likely to be current smokers than non-Hispanic whites. Health behaviors differed by demographic and medical characteristics (Table 3). Attaining more than a high school education was positively associated with normal BMI, meeting F&V intake recommendations, meeting PA guidelines, and non-current smoking status. Obese survivors were less likely to follow PA guidelines than overweight survivors. Importantly, the time since diagnosis was not associated with BMI, F&V intake, or PA. However, those who survived beyond 10 years after their cancer diagnosis were more likely to be current smokers. Discussion This study indicates that racial and ethnic groups vary in their adherence to health behavior recommendations, and unhealthy behaviors are not limited to a racial or ethnic group. Although racial and ethnic minorities were less likely than non-Hispanic whites to engage in some preventive health behaviors, the reverse was true for other health behaviors. Thus, racial and ethnic minorities are heterogeneous with respect to these associations.8,19,20 CS and individuals in the general population indicated similar behavioral patterns for BMI and PA guidelines but varied in F&V and smoking behaviors. Perhaps having cancer encourages individuals to improve certain behaviors, but long-term abstinence from smoking presents a more difficult challenge. African American and Hispanic survivors were more likely to be overweight or obese than non-Hispanic white survivors, which is consistent with previous research.7,21,22 Obesity significantly impacts physical function, which may influence future risks for disease www.ajpmonline.org Nayak et al / Am J Prev Med 2015;48(6):729–736 23–25 recurrence or premature death ; therefore, African Americans and Hispanics may be at greater risk for adverse outcomes than non-Hispanic whites.26,27 In addition, fewer African Americans than nonHispanic whites followed PA guidelines, consistent with previous studies.7,8 There is a need to identify the correlates of these behaviors to understand how best to intervene. Such interventions may improve functional status, reduce high BMI, and improve cancer-specific outcomes.4,6 Smoking rates were highest among Native Americans, which is consistent with the findings of other studies from the general poulation28,29 and concurs with a prior study showing low smoking-cessation rates among Native Americans.30 More research is needed to examine why racial and ethnic differences exist in smoking behaviors of CS and what types of tailored interventions will be most effective for these populations. Limitations Some limitations were noted. These data are crosssectional; therefore, they cannot be used to infer causal relationships. All data are self-reported, with potential recall and reporting biases. Income was not included in the multivariable model, owing primarily to a large number of missing observations (9%); however, education served as a suitable proxy. Conclusions An important strength of this study was that these data are representative of the U.S. population and provide population estimates of racial and ethnic differences in the adoption of recommended health behaviors. This study suggests that surviving cancer does not eliminate the significant racial and ethnic differences in the adoption of preventive behaviors that exist in the general population. Future research should focus on gaining a better understanding of these differences in the adoption of health behaviors among racial and ethnic minorities as well as among CS to reduce smoking. This research was supported by funds from the University Cancer Foundation and the Duncan Family Institute for Cancer Prevention and Risk Assessment via the Cancer Survivorship Research Seed Money Grants at The University of Texas MD Anderson Cancer Center, and by funds from the Cancer Prevention and Research Institute of Texas through the CERCIT grant (RP101207 P04 02—L. Elting, principal investigator [PI]) to the University of Texas Medical Branch at Galveston. This work was presented as a poster at the 6th Biennial Cancer Survivorship Research Conference, June 2012, Arlington, Virginia. June 2015 735 RJP is supported by a grant from the National Cancer Institute (5K01CA158000). HMH is supported by a grant from NIH (K23 AG038476). LSE is supported by the Cancer Prevention and Research Institute of Texas through the Comparative Effectiveness Research on Cancer in Texas grant (RP101207 P04 02—L. Elting, PI) to the University of Texas Medical Branch at Galveston. The University of Texas MD Anderson Cancer Center is supported in part by the NIH through Cancer Center Support Grant P30CA016672. No other financial disclosures were reported by the authors of this paper. References 1. Patnaik JL, Byers T, Diguiseppi C, Dabelea D, Denberg TD. Cardiovascular disease competes with breast cancer as the leading cause of death for older females diagnosed with breast cancer: a retrospective cohort study. Breast Cancer Res. 2011;13(3):R64. http://dx.doi.org/ 10.1186/bcr2901. 2. Travis LB, Hill D, Dores GM, et al. Cumulative absolute breast cancer risk for young women treated for Hodgkin lymphoma. J Natl Cancer Inst. 2005;97(19):1428–1437. http://dx.doi.org/10.1093/jnci/dji290. 3. Mao JJ, Armstrong K, Bowman MA, Xie SX, Kadakia R, Farrar JT. Symptom burden among cancer survivors: impact of age and comorbidity. J Am Board Fam Med. 2007;20(5):434–443. http://dx.doi.org/ 10.3122/jabfm.2007.05.060225. 4. Morey MC, Snyder DC, Sloane R, et al. Effects of home-based diet and exercise on functional outcomes among older, overweight long-term cancer survivors: RENEW: a randomized controlled trial. JAMA. 2009;301(18):1883–1891. http://dx.doi.org/10.1001/jama.2009.643. 5. Pekmezi DW, Demark-Wahnefried W. Updated evidence in support of diet and exercise interventions in cancer survivors. Acta Oncol. 2011;50(2): 167–178. http://dx.doi.org/10.3109/0284186X.2010.529822. 6. Blanchard CM, Courneya KS, Stein K. Cancer survivors’ adherence to lifestyle behavior recommendations and associations with healthrelated quality of life: results from the American Cancer Society’s SCS–II. J Clin Oncol. 2008;26(13):2198–2204. http://dx.doi.org/ 10.1200/JCO.2007.14.6217. 7. Schootman M, Deshpande AD, Pruitt SL, Aft R, Jeffe DB. National estimates of racial disparities in health status and behavioral risk factors among long-term cancer survivors and non-cancer controls. Cancer Causes Control. 2010;21(9):1387–1395. http://dx.doi.org/10. 1007/s10552-010-9566-x. 8. Smith AW, Alfano CM, Reeve BB, et al. Race/ethnicity, physical activity, and quality of life in breast cancer survivors. Cancer Epidemiol Biomarkers Prev. 2009;18(2):656–663. http://dx.doi.org/10.1158/ 1055-9965.EPI-08-0352. 9. CDC. Behavioral Risk Factor Surveillance System Survey Data. Atlanta, GA: USDHHS, CDC; 2009. 10. CDC. Behavioral Risk Factor Surveillance System Survey. Atlanta, GA: USDHHS, CDC; 2009. 11. WHO. Obesity: Preventing and managing the global epidemic. Report on a WHO consultation on obesity; June 3-5, 1997. Geneva, Switzerland; 1998. 12. Rock CL, Doyle C, Mark-Wahnefried W, et al. Nutrition and physical activity guidelines for cancer survivors. CA Cancer J Clin. 2012;62(4): 243–274. http://dx.doi.org/10.3322/caac.21142. 13. USDHHS. Healthy People 2010. Understanding and Improving Health and Objectives for Improving Health. 2nd edition, Washington, DC: Government Printing Office, 2000; 2011. 14. CDC. Behavioral Risk Factor Surveillance System Survey Questionnaire. Atlanta, GA: USDHHS, CDC; 2009. 736 Nayak et al / Am J Prev Med 2015;48(6):729–736 15. Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116(9):1081–1093. http://dx.doi.org/10.1161/CIRCULATIONA HA.107.185649. 16. Yore MM, Ham SA, Ainsworth BE, et al. Reliability and validity of the instrument used in BRFSS to assess physical activity. Med Sci Sports Exerc. 2007;39(8):1267–1274. http://dx.doi.org/10.1249/mss.0b013e31 80618bbe. 17. Eberth JM, Prarelkar P, Nguyen H, Sun C, Irvin-Vidrine J, Elting LS. The human and economic burden of cervical cancer in Texas. Tex Public Health J. 2013;65(2):51–55. 18. von EE, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453–1457. http://dx.doi.org/ 10.1016/S0140-6736(07)61602-X. 19. Ollberding NJ, Maskarinec G, Wilkens LR, Henderson BE, Kolonel LN. Comparison of modifiable health behaviours between persons with and without cancer: the Multiethnic Cohort. Public Health Nutr. 2011;14 (10):1796–1804. http://dx.doi.org/10.1017/S136898001000340X. 20. Rogers LQ, Courneya KS, Paragi-Gururaja R, Markwell SJ, Imeokparia R. Lifestyle behaviors, obesity, and perceived health among men with and without a diagnosis of prostate cancer: a population-based, crosssectional study. BMC Public Health. 2008;8:23. http://dx.doi.org/ 10.1186/1471-2458-8-23. 21. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 19992004. JAMA. 2006;295(13):1549–1555. http://dx.doi.org/10.1001/ jama.295.13.1549. 22. Paxton RJ, Phillips KL, Jones LA, et al. Associations among physical activity, body mass index, and health-related quality of life by race/ 23. 24. 25. 26. 27. 28. 29. 30. ethnicity in a diverse sample of breast cancer survivors. Cancer. 2012;118(15):4024–4031. http://dx.doi.org/10.1002/cncr.27389. Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ. Overweight, obesity, and mortality from cancer in a prospectively studied cohort of U.S. adults. N Engl J Med. 2003;348(17):1625–1638. http://dx.doi.org/ 10.1056/NEJMoa021423. Chang SH, Pollack LM, Colditz GA. Obesity, mortality, and life years lost associated with breast cancer in nonsmoking U.S. Women, National Health Interview Survey, 1997-2000. Prev Chronic Dis. 2013;10:E186. http://dx.doi.org/10.5888/pcd10.130112. Schmitz KH, Neuhouser ML, gurs-Collins T, et al. Impact of obesity on cancer survivorship and the potential relevance of race and ethnicity. J Natl Cancer Inst. 2013;105(18):1344–1354. http://dx.doi.org/10. 1093/jnci/djt223. Pinto BM, Trunzo JJ. Health behaviors during and after a cancer diagnosis. Cancer. 2005;104(11)(suppl):2614–2623. http://dx.doi.org/ 10.1002/cncr.21248. Ward E, Jemal A, Cokkinides V, et al. Cancer disparities by race/ ethnicity and socioeconomic status. CA Cancer J Clin. 2004;54(2): 78–93. http://dx.doi.org/10.3322/canjclin.54.2.78. Pleis JR, Lethbridge-Cejku M. Summary health statistics for U.S. adults: National Health Interview Survey, 2005. Vital Health Stat 10. 2006; (232):1–153. Steele CB, Cardinez CJ, Richardson LC, Tom-Orme L, Shaw KM. Surveillance for health behaviors of American Indians and Alaska Natives —findings from the behavioral risk factor surveillance system, 2000-2006. Cancer. 2008;113(5)(suppl):1131–1141. http://dx.doi.org/10.1002/cncr.23 727. Gohdes D, Harwell TS, Cummings S, Moore KR, Smilie JG, Helgerson SD. Smoking cessation and prevention: an urgent public health priority for American Indians in the Northern Plains. Public Health Rep. 2002;117(3): 281–290. http://dx.doi.org/10.1016/S0033-3549(04)50162-0. www.ajpmonline.org