Clinician Counseling and Weight Loss Attempts Among Survivors of Obesity-Associated Cancers Authors: Ryan Howard, Archana Radhakrishnan, Dina H. Griauzde, Jennifer J. Griggs, Lesly A. Dossett Published: 2026 ABSTRACT This cross-sectional study evaluates whether clinician counseling is associated with weight loss attempts among survivors of obesity-associated cancers and whether this differs from the general population with obesity. Introduction At least 13 cancers are causally linked to excess body weight.1 Of the 18 million US cancer survivors, an estimated two-thirds have overweight or obesity, which across cancer types is associated with cancer recurrence, second primary cancers, and worse cancer-specific and overall survival.2 Despite this, obesity is infrequently addressed as part of survivorship care.3 Whether clinician counseling is associated with weight loss attempts among survivors of obesity-associated cancers and whether this differs from the general population with obesity is unknown. Methods We pooled 3 cycles (January 2013 to December 2014, January 2015 to December 2016, and January 2017 to March 2020) of the National Health and Nutrition Examination Survey (NHANES) to identify adults (aged ≥20 years) with obesity (body mass index [BMI] ≥30; calculated as weight in kilograms divided by height in meters squared) who reported a prior diagnosis of an obesity-associated cancer as defined by the International Agency for Research on Cancer (IARC) (eTable in Supplement 1). We focused on IARC obesity-associated cancers because obesity is linked to their cause, strengthening the rationale for weight management in survivorship. We also identified a comparator of adults with obesity without a history of an IARC-defined obesity-associated cancer. The primary outcome was attempted weight loss in the past 12 months (“During the past 12 months, have you tried to lose weight?”). The exposure was clinician counseling (“During the past 12 months have you ever been told by a doctor or health professional to control your weight or lose weight?”). Adjusted probabilities of attempted weight loss were estimated using survey-weighted logistic regression models that included variables previously associated with clinician weight counseling: age, sex, race and ethnicity, insurance, education, BMI, diabetes status, hypertension, hypercholesterolemia, and smoking status. The primary analysis tested whether the association between counseling and attempted weight loss differed between survivors of obesity-associated cancers and the comparator group. Analyses were performed from April to May 2026 in Stata version 18 (StataCorp) with a 2-sided α = .05 indicating significance. The University of Michigan institutional review board deemed this analysis exempt from review and the need for informed consent because it used deidentified publicly available data. We followed Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cross-sectional studies. Results We identified 315 adults (median [IQR] age, 66.0 [56.0-74.0] years; 253 [80.3%] female) with obesity and a history of obesity-associated cancer (weighted population, 3.7 million) and 7439 comparator adults (median [IQR] age, 50.0 [36.0-63.0] years; 4101 [55.1%] female) with obesity (weighted population, 92.6 million). Survivors of obesity-associated cancers were more often Medicare-insured and had more cardiometabolic comorbidity (Table 1). Table 1. Cohort CharacteristicsCharacteristicParticipants, No. (%)Adults with obesity and history of obesity-associated cancer (n = 315)Adults with obesity and no history of obesity-associated cancer (n = 7439)Age, median (IQR), y66.0 (56.0-74.0)50.0 (36.0-63.0)Sex Male62 (19.7)3338 (44.9) Female253 (80.3)4101 (55.1)Race and ethnicity Black non-Hispanic65 (20.6)2129 (28.6) Hispanic80 (25.4)2044 (27.5) White non-Hispanic146 (46.3)2631 (35.4) Othera24 (7.6)635 (8.5)Insurance Medicare178 (57.2)1754 (24.1) Medicaid28 (9.0)801 (11.0) Private71 (22.8)2918 (40.1) Other public13 (4.2)492 (6.8) Uninsured21 (6.8)1318 (18.1)Education ≤High school137 (43.6)3365 (45.3) Some college124 (39.5)2617 (35.2) ≥College53 (16.9)1450 (19.5)BMI, median (IQR)b34.7 (32.0-38.9)34.6 (31.9-38.9)Obesity severity class (BMI) Obesity I (30 to <35)165 (52.4)3936 (52.9) Obesity II (35 to <40)83 (26.3)1931 (26.0) Severe obesity (≥40)67 (21.3)1572 (21.1)Comorbidities Diabetes100 (33.1)1489 (20.8) Prediabetes only50 (16.6)783 (10.9) Hypertension200 (63.5)3523 (47.4) Hypercholesterolemia174 (56.7)2881 (39.1)Smoking status Never171 (54.5)4260 (57.3) Former105 (33.4)1894 (25.5) Current38 (12.1)1280 (17.2)Cancer type (within obesity-associated cancer cohort) Breast129 (41.0)NA Uterine50 (15.9)NA Colorectal57 (18.1)NA Other79 (25.1)NAAbbreviations: BMI, body mass index; NA, not applicable. a Race and ethnicity were self-reported by participants during the National Health and Nutrition Examination Survey in response to standardized questions on race and Hispanic origin. The other race category includes non-Hispanic Asian respondents and non-Hispanic respondents reporting other or multiple races. b Calculated as weight in kilograms divided by height in meters squared. Cohort Characteristics Abbreviations: BMI, body mass index; NA, not applicable. Race and ethnicity were self-reported by participants during the National Health and Nutrition Examination Survey in response to standardized questions on race and Hispanic origin. The other race category includes non-Hispanic Asian respondents and non-Hispanic respondents reporting other or multiple races. Calculated as weight in kilograms divided by height in meters squared. After adjustment, survivors of obesity-associated cancers were no more likely than comparator adults to receive clinician counseling on overweight diagnosis, weight loss, exercise, or reducing fat and/or calories (Table 2). Survivors of obesity-associated cancers were less likely to have attempted weight loss in the past 12 months (survivors: 47.9%; 95% CI, 38.1%-57.8%; nonsurvivors: 59.6%; 95% CI, 57.5%-61.7%). Table 2. Adjusted Clinician Counseling and Weight-Loss Attempts by CohortCharacteristicParticipants, % (95% CI)aAdults with obesity and history of obesity-associated cancerAdults with obesity and no history of obesity-associated cancerClinician counseling Advised overweight72.0 (64.2-79.8)71.8 (70.3-73.4) Advised to lose weight51.1 (40.8-61.5)51.8 (50.1-53.5) Advised to exercise58.5 (50.3-66.6)56.8 (55.0-58.7) Advised to reduce fat or calories44.3 (36.5-52.1)46.4 (44.9-48.0)Attempted weight loss Overall47.9 (38.1-57.8)59.6 (57.5-61.7) Without clinician counseling24.2 (13.8-34.6)53.9 (51.0-56.8) With clinician counseling65.6 (58.1-73.0)65.5 (62.5-68.5) Difference, percentage points41.6 (28.4-54.7)11.6 (7.7-15.5) a Values are adjusted predicted probabilities (percentage) with 95% CIs, estimated by survey-weighted logistic regression adjusting for age, sex, race and ethnicity, insurance, education, body mass index, diabetes status, hypertension, hypercholesterolemia, and smoking status. Clinician counseling outcomes and overall attempted weight loss were each estimated from separate models; estimates of attempted weight loss stratified by counseling status were derived from a model that included a counseling × cohort interaction term. The difference is the adjusted within-cohort difference in the probability of attempted weight loss between respondents who did and did not receive clinician counseling, expressed in percentage points. The difference in the association of counseling between cohorts (difference-in-differences) was 29.8 (95% CI, 16.0-43.5) percentage points. Adjusted Clinician Counseling and Weight-Loss Attempts by Cohort Values are adjusted predicted probabilities (percentage) with 95% CIs, estimated by survey-weighted logistic regression adjusting for age, sex, race and ethnicity, insurance, education, body mass index, diabetes status, hypertension, hypercholesterolemia, and smoking status. Clinician counseling outcomes and overall attempted weight loss were each estimated from separate models; estimates of attempted weight loss stratified by counseling status were derived from a model that included a counseling × cohort interaction term. The difference is the adjusted within-cohort difference in the probability of attempted weight loss between respondents who did and did not receive clinician counseling, expressed in percentage points. The difference in the association of counseling between cohorts (difference-in-differences) was 29.8 (95% CI, 16.0-43.5) percentage points. The association between clinician counseling and attempted weight loss differed between groups. Counseling was associated with a 41.6-percentage point (pp) higher probability (95% CI, 28.4-54.7 pp) of attempted weight loss among cancer survivors vs an 11.6-pp higher probability (95% CI, 7.7-15.5 pp) among comparator adults (difference in differences, 29.8 pp; 95% CI 16.0-43.5 pp). Discussion In this cross-sectional study of US adults with obesity, survivors of obesity-associated cancers were not more likely than comparator adults to receive weight management counseling from clinicians and were less likely to attempt weight loss, consistent with prior studies.4 However, when counseling did occur, we found that it was more strongly associated with weight loss attempts among survivors than in the general population of adults with obesity. Although clinician counseling has been linked to weight-management program enrollment in select cancer survivors, no prior study has quantified its differential association between these groups.5 Survivors of obesity-associated cancers thus appear to be both underengaged at baseline and uniquely responsive to counseling when it does occur. This combination of high need, high responsiveness, and low delivery defines an important opportunity. That survivors are no more likely to receive counseling may reflect fragmentation of obesity management within survivorship care, particularly transitions from oncology follow-up to primary care that diffuse ownership of obesity treatment. It may also reflect limited awareness among clinicians and patients alike that obesity is a modifiable risk factor for cancer recurrence and cancer-specific mortality. Several mechanisms may explain the differential association of counseling between groups. The posttreatment period is recognized as a teachable moment in which survivors are motivated to act on health-promoting advice, and counseling delivered in the context of cancer care may carry greater perceived stakes.6 Limitations include cross-sectional design; self-reported NHANES measures; unknown relationship to subsequent clinical outcomes; and lack of data on time since diagnosis, prognosis, and competing health concerns. Aggregating across cancer types also limits interpretation, as obesity’s prognostic role varies by cancer and the small sample precludes site-specific estimates. Lastly, whether weight loss benefits survivors (ie, the obesity paradox) remains contested.7 Our findings suggest that systematic, standardized delivery of clinician counseling represents a promising opportunity to prompt weight loss attempts among survivors of obesity-associated cancers. With the increasing availability of effective weight management therapies, spanning lifestyle interventions, pharmacotherapy, and bariatric surgery, the potential benefit of such counseling may be greater than ever before.