
==== Front
Prev Med Rep
Preventive Medicine Reports
2211-3355

S2211-3355(24)00274-2
10.1016/j.pmedr.2024.102859
102859
Short Communication
Preventing diabetes: What overweight and obese adults with prediabetes in the United States report about their providers’ communication and attempted weight loss
Demosthenes Emmanuella J. Emmanuella.Demosthenes@umassmed.edu
a
Freedman Jason a
Hernandez Camila a
Shennette Lisa b
Frisard Christine F. a
Lemon Stephenie C. a
Gerber Ben S. a
Amante Daniel J. Daniel.Amante@umassmed.edu
a⁎
a UMass Chan Medical School, Department of Population and Quantitative Health Sciences, United States
b U.S. Department of Veteran's Affairs, United States
⁎ Corresponding author. Daniel.Amante@umassmed.edu
11 8 2024
10 2024
11 8 2024
46 10285926 4 2024
9 8 2024
10 8 2024
© 2024 The Authors. Published by Elsevier Inc.
2024

https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Objective

To investigate what overweight or obese adults with prediabetes in the United States report being told by providers about 1) having prediabetes, 2) diabetes risk, and 3) losing weight and the associations of these communications with attempted weight loss.

Methods

Data from 2015 to 2018 National Health and Nutrition Examination Surveys (NHANES) for adults with a body mass index in the overweight or obesity ranges and HbA1c in the prediabetes range were examined (n = 2085). Patient reported data on what providers told them about having prediabetes, being at risk for diabetes, and losing weight were compared with attempted weight loss.

Results

Most participants (66.4%) reported never being told they had prediabetes nor being at risk for diabetes, 13.0% reported being told they had prediabetes, 10.6% at risk for diabetes, and 8.0% both messages. 18.3% of participants reported being told to lose weight. Participants who reported being told they had prediabetes and at increased diabetes risk were more likely to report attempted weight loss (adjusted odds ratio (AOR) 1.8, 95% confidence interval (CI) 1.1–3.2). Reporting that they were told to lose weight was not significantly associated with an increase in reported weight loss attempts.

Conclusions

In this cohort of individuals with overweight/obesity and prediabetic HbA1c values, low rates communications with providers about prediabetes and diabetes risk were reported. When both were discussed, patients reported greater attempted weight loss. These findings draw attention to the potential impact that provider communications about prediabetes and diabetes risk may have on lifestyle behavior change.

Abbreviations

AOR Adjusted Odds Ratio

BMI Body Mass Index

CI Confidence Interval

DPP Diabetes Prevention Program

HbA1c Hemoglobin A1c

NHANES National Health and Nutrition Examination Survey
==== Body
pmc1 Introduction

Prediabetes, defined by a fasting glucose level of 100-125 mg/dL, a glucose level of 140-199 mg/dL measured 2h after a 75-g oral glucose load, or glycated hemoglobin level (HbA1c) of 5.7% to 6.4%, affects about one out of every three adults in the United States (US) (Echouffo-Tcheugui et al., 2023). In a national US sample, over 80% of individuals in the US with prediabetes were overweight or had obesity (Liu et al., 2020). Clinical trials such as the Diabetes Prevention Program (DPP) have demonstrated that even small reductions in body weight reduces the risk of developing type 2 diabetes in overweight adults with prediabetes (Diabetes Prevention Program Research et al., 2009). While weight loss interventions can be effective, challenges persist in implementing these programs. Noted barriers include limited provider time for counseling (Cronin et al., 2019).

Individuals with prediabetes often report wanting tailored information from their clinicians on strategies for weight loss (Roper et al., 2019). Primary care physicians report limited knowledge of diagnostic criteria, risk factors, and screening tests, as well as a lack of implementation of evidenced-based recommendations for prediabetes treatment, specifically referral to programs such as the DPP (Tseng et al., 2019). Providers’ perceptions on diagnosing and treating prediabetes vary, with some finding limited utility in diagnosing and treating prediabetes due to lack of patient motivation to make lifestyle changes or barriers to adhering to medication therapy (Kandula et al., 2018). Others find that a prediabetes diagnosis is important for managing patient health and determining whether to treat comorbid conditions (Tseng et al., 2019). Prior research suggests that physician advice can have a “priming effect” on patient responses to behavior change (Kreuter et al., 2000). According to the Health Belief Model, risk perception of a health condition (i.e. perceived susceptibility and severity) influences the likelihood of engaging in desired health behaviors (Champion and Skinner, 2008). The period in time in which a person’s blood glucose levels are in the prediabetes range represents a critical window of opportunity for providers to communicate risk of developing diabetes and refer patients to evidence-based strategies to reduce risk, including weight loss (Diabetes Prevention Program Research et al., 2009).

The objectives of this study were to (1) examine provider communication with individuals about prediabetes, diabetes risk, and advice to lose weight, and (2) evaluate the associations between those messages and patient-reported attempted weight loss in a nationally representative sample of adults with overweight or obesity and prediabetes in the US. We performed a cross-sectional analysis of the National Health and Nutrition Examination Survey (NHANES). We analyzed the 2015–2018 datasets as they were the most recent versions that included all the exposure variables of interest.

2 Methods

2.1 Data source

The publicly available datasets were downloaded directly from NHANES website (www.cdc.gov/nchs/nhanes). NHANES is a nationally representative program of studies designed to assess the health and nutritional status of adults and children in the United States (CDC, 2023). The NHANES datasets include interview responses and results from physical examinations. The interviews include information on demographic and health-related questions. The physical examinations include laboratory tests and physiological measurements conducted by trained staff. All NHANES protocols are approved by the National Center for Health Statistics ethics review board. As this study was based on publicly available anonymized databases, it is considered exempt from ethical compliance by the UMass Chan Medical School Institutional Review Board.

2.2 Study sample

We conducted cross-sectional analyses on a sample of 2085 adults (aged 18 years and older) from the 2015–2018 cohorts with body mass index (BMI) in the overweight or obesity ranges and HbA1c values in the prediabetes range who had not previously been told that they had diabetes. Inclusion criteria included having a BMI value in the following ranges, 25–29.9kg/m2 (overweight) and ≥30kg/m2 (obese) for non-Asian participants and 23–27.5 kg/m2 (overweight) and ≥27.5kg/m2 (obese) for Asian participants, and HbA1c values ≥5.7% and ≤6.4%. Exclusion criteria included reporting yes to ever been told by a doctor or health professional that they have diabetes.

2.3 Exposure variables

The self-reported variables of interest were: (1) ever told you have prediabetes, (2) ever told you have health risk for diabetes, and (3) told to lose weight. These variables were assessed by the following interview questions:• “Have you ever been told by a doctor or other health professional that you have any of the following: prediabetes, impaired fasting glucose, impaired glucose tolerance, borderline diabetes or that your blood sugar is higher than normal but not high enough to be called diabetes or sugar diabetes?”

• “Have you ever been told by a doctor or other health professional that you have health conditions or a medical or family history that increases your risk for diabetes?”

• “During the past 12 months have you ever been told by a doctor or health professional to control your weight or lose weight?”

The response options for all three questions were “Yes”, “No”, “Refused”, or “Don’t know”. Responses for each question were categorized into 2 groups, either “Yes” or “No, Refused, or Don’t know”.

2.4 Outcome variable

The outcome variable of interest was recent attempted weight loss. This was assessed by the response to the question “During the past 12 months, have you tried to lose weight?”. Response options were “Yes”, “No”, “Refused”, or “Don’t know”. Responses were categorized into 2 groups, either “Yes” or “No, Refused, or Don’t know”.

2.5 Statistical analyses

To characterize provider communication on prediabetes status and diabetes risk, we considered four groups − those who reported being: (1) told of having prediabetes or abnormal glucose levels; (2) told of increased diabetes risk; (3) told neither, or (4) told both messages. To characterize provider communication on losing weight, we considered those who reported being told to control or lose weight with those who did not.

Differences in demographic characteristics, mean HbA1c, and weight category (overweight, obese) covariates were assessed by Pearson χ2 or one-way ANOVA, as appropriate. Weighted logistic regressions, unadjusted and adjusted for the covariates described above, assessed the associations between what the study population reported being told about their prediabetes status, diabetes risk, and instructions to lose weight and their self-reported recent weight loss attempts. The reference group for each regression was those who reported being told neither that they had prediabetes nor that they had health risk for diabetes. Analyses were performed using Stata version 15.1.

3 Results

In this analysis, 51.9% of participants were female, the majority of the population were non-Hispanic White (56.8%) and the mean age was 55.3 years (SD 15.6). Mean HbA1c was 5.9% (SE 0.2) with 39.1% of the population with BMI in the overweight range and 60.9% with BMI in the obesity range. Most individuals (68.4%) reported that they were not told by their health care provider that they had prediabetes nor that they were at risk for diabetes. In comparison, 13.0% of individuals reported being told they had prediabetes, 10.6% reported being told that they were at risk for diabetes, and 8.0% reported being told both messages. Among the population, 18.3% of participants reported being told by their provider to control or lose weight. (Table 1).Table 1 Population characteristics of U.S. adults with overweight or obesity and with HbA1c in prediabetes range by healthcare provider communication in NHANES 2015–2018.

	Total	Told neither	Told of prediabetes	Told at risk for diabetes	Told both	P*	
N = 2085	n = 1427 (68.4 %)	n = 271 (13.0 %)	n = 220 (10.6 %)	n = 167 (8.0 %)		
Gender						0.55	
n (%)	
Male	997 (48.1)	741 (50.1)	106 (44.1)	84 (44.9)	66 (44.7)		
Female	1088 (51.9)	686 (49.9)	165 (55.9)	136 (55.1)	101 (55.3)		


	
Age						<.001	
mean (SE)	53.6 (17.3)	54.8 (0.7)	61.1 (1.1)	48.6 (1.0)	53.1 (1.4)		


	
Race						0.04	
n (%)	
Non-Hispanic White	591 (56.8)	396 (55.6)	89 (64.7)	65 (51.5)	41 (59.9)		
Mexican American	354 (10.4)	233 (10.4)	44 (7.6)	50 (16.0)	27 (7.6)		
Other Hispanic	249 (7.1)	171 (7.3)	30 (6.5)	18 (4.6)	30 (9.0)		
Non-Hispanic Black	554 (15.1)	399 (16.5)	64 (11.8)	57 (15.7)	34 (9.4)		
Non-Hispanic Asian	258 (6.1)	178 (6.3)	36 (6.7)	18 (3.9)	26 (6.0)		
Other/Multi-Race	79 (4.6)	50 (3.8)	8 (2.8)	12 (8.3)	9 (8.1)		


	
Insurance						0.02	
n (%)	
No Insurance	324 (12.9)	250 (14.4)	20 (6.4)	36 (13.6)	18 (11.2)		
Private Insurance	708 (40.3)	468 (41.2)	80 (30.2)	86 (44.4)	74 (43.5)		
Public Insurance	735 (29.5)	497 (27.8)	108 (35.0)	77 (30.9)	59 (31.6)		
Dual Coverage	287 (17.3)	193 (16.6)	57 (28.4)	2 (11.1)	15 (13.7)		


	
Weight						0.16	
n (%)	
Overweight	884 (39.2)	646 (41.3)	113 (36.6)	70 (28.7)	55 (40.8)		
Obese	1290 (60.)	846 (58.7)	163 (63.4)	165 (71.3)	116 (59.2)		


	
HbA1c %						<.001	
mean (SE)	5.9 (0.19)	5.9 (0.01)	5.9 (0.02)	5.9 (0.01)	5.9 (0.02)		
Told to lose weight n (%)						<.001	
No	1789 (81.9)	1292 (86.3)	214 (83.8)	174 (78.0)	109 (55.2)		
Yes	385 (18.1)	200 (13.7)	62 (16.2)	61 (22.0)	62 (44.8)		


	
Attempt Weight Loss						0.02	
n (%)							
No	965(50.1)	731 (54.0)	105 (49.3)	79 (41.0)	50 (35.4)		
Yes	908(49.9)	565 (46.0)	134 (50.7)	115 (59.0)	94 (64.6)		
* P values calculated by Adjusted Wald test or Pearson χ2 test.

In the unadjusted logistic regression models, individuals who reported being told they were at risk for diabetes were more likely to report attempted weight loss (OR 1.7, 95% CI 1.1–2.7) than those who reported not being told any messages. Individuals who reported only being told that they had prediabetes were not more likely to report attempted weight loss compared to those who reported being told neither message. Those who reported being told that they both had prediabetes and were at risk for diabetes were more than twice as likely to report attempted weight loss (OR 2.1, 95% CI 1.1–4.1). Individuals who reported being told to lose weight also reported greater attempted weight loss (OR 2.0, 95% CI 1.4–2.7).

After accounting for demographic characteristics, HbA1c, and BMI, the adjusted multivariate logistic regression model showed no significant differences in self-reported weight loss attempts between those who reported being told that they had prediabetes or were at increased risk of developing diabetes compared to those who were told neither message. Those who reported being told both that they had prediabetes and were at risk for diabetes were almost twice as likely to report recent attempted weight loss (adjusted OR 1.8, 95% CI 1.1–4.1). Those who reported being told to lose weight did not report significantly more attempted weight loss in the adjusted model (OR 1.4, 95% CI 0.98–2.1). (Table 2).Table 2 Logistic regressions examining associations between provider communication and attempted weight loss among U.S. adults with overweight or obesity and HbA1c in prediabetes range in NHANES 2015–2018.

	Attempted Weight Loss	
	Unadjusted OR (95% CI)	Adjusted OR* (95% CI)	
Report of provider communication on pre-diabetes and diabetes risk	
Told neither	Reference	Reference	
Told have prediabetes	1.2 (0.9–1.7)	1.2 (0.8–1.7)	
Told at risk for diabetes	1.7 (1.1–2.6)	1.4 (0.9–2.3)	
Told both have prediabetes and at risk for diabetes	2.1 (1.1–4.1)	1.8 (1.03–3.2)	


	
Report of provider communication on instructions to lose weight	
Not reported	Reference	Reference	
Told to lose weight	2.0 (1.4–2.7)	1.4 (0.98–2.1)	
* Adjusted for gender, age, race/ethnicity, insurance type, weight category, and HbA1c (%)

4 Discussion

We found that a majority (68.4%) of overweight or obese adults with an HbA1c value in the prediabetes range reported that their providers had not told them that they have prediabetes nor that they were at increased risk for diabetes. Surprisingly, less than half of the people who reported being told they had prediabetes also reported being told of having increased risk for developing diabetes. This suggested lack of knowledge and/or patient-provider communication about prediabetes is consistent with other studies (Li et al., 2021, Murillo et al., 2019, Joiner et al., 2022). It may reflect a lack of formal screening efforts or inadequate communication about prediabetes status and diabetes risk. With the 2022 U.S. Preventive Services Task Force (USPSTF) recommendations of lowering the starting age of screening for diabetes from 40 to 35 years old in all adults who are overweight or obese, the incidence of prediabetes diagnoses will likely increase (Force et al., 2022). How providers then communicate with patients about their prediabetes status and risk for developing diabetes becomes critically important.

Providers’ perspectives on the use of the term “prediabetes” vary. Some providers report that they favor explaining an individual’s risk of developing diabetes instead of using the term ‘prediabetes’ (Herman, 2023). Providing people with an individualized risk score at the time of prediabetes diagnosis has been found to be better processed, understood, and more likely to lead to behavior change than using population risk estimates (Edwards et al., 2000). Using individually tailored risk models may facilitate patient understanding of modifiable healthy behaviors such as weight loss and its relation to reducing diabetes risk.

Tailored risk feedback to individuals, accompanied by motivational interviewing or other forms of counseling, may also support behavior change. Communication style and messaging will likely influence outcomes (Albury et al., 2023). Gain-framed messages (e.g. positive consequences of performing a behavior) may be more effective in promoting health-affirming behaviors like attempted weight loss, while loss-framed messages (e.g. negative consequences of not performing a behavior) may be more effective to promote screening behaviors (Gallagher and Updegraff, 2012). In prior studies, physicians discussing weight status with their patients has been associated with clinically meaningful weight loss (Pool et al., 2014). Increasing provider knowledge of existing behavior change programs, like the DPP, and establishing referral pathways for providers to facilitate patient engagement may further improve outcomes. This is particularly important among those with prediabetes, where there has been very low engagement observed (Ali et al., 2019). Beyond intensive lifestyle interventions, personal knowledge of diabetes risk may be relevant in provider discussions of risks and benefits of newer anti-obesity medications (such as glucagon-like peptide 1 analogues), which are increasingly prescribed.

4.1 Limitations

This report has several limitations. Due to the cross-sectional design of the analyses, we are unable to determine causality or capture trends in behavior change over time. The possibility of recall bias means patient-reported data may not accurately portray what providers communicated to individuals about having prediabetes, being at risk for diabetes, instructions to lose weight and their attempted weight loss over the previous 12 months. An additional limitation is that single BMI and HbA1c values taken when enrolled in NHANES were used to define the study population as having overweight/obesity and blood glucose values in the prediabetes range. Additionally, this study examined only one aspect of preventive behavior change (attempted weight loss) and there is potential for social desirability bias in self-reporting of this behavior.

4.2 Conclusions

Most individuals with overweight or obesity and prediabetes report that their clinical providers did not tell them about their prediabetes status nor their increased risk of developing diabetes. We also found that most overweight or obese people report that their providers had not told them to lose weight. However, individuals who report being told both prediabetes and consequent risk messages are more likely to also report attempted weight loss. This highlights the potential impact that provider communication about prediabetes and diabetes risk may have on lifestyle behavior change.

Disclosure of funding and conflicts of interest

The authors of this manuscript certify that they have no conflicts of interest to report. Funding for this work was partially supported by NIH 5K01DK131318.

CRediT authorship contribution statement

Emmanuella Demosthenes: Writing – review & editing, Project administration. Jason Freedman: Writing – original draft, Conceptualization. Camilla Hernandez: Writing – original draft. Lisa Shennette: . Christine F. Frisard: Formal analysis. Stephenie C. Lemon: Writing – review & editing, Supervision, Methodology. Ben S. Gerber: . Daniel J. Amante: Writing – original draft, Supervision, Methodology, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.
==== Refs
References

Albury C. Webb H. Stokoe E. Relationship between clinician language and the success of behavioral weight loss interventions: a mixed-methods cohort study Ann Intern Med. 176 11 2023 1437 1447 37931269
Ali M.K. Bullard K.M. Imperatore G. Reach and use of diabetes prevention services in the United States, 2016–2017 JAMA Network Open. 2 5 2019 e193160 e 31074808
CDC, 2023. About the National Health and Nutrition Examination Survey. https://www.cdc.gov/nchs/nhanes/about_nhanes.htm. Updated May 31 2023. Accessed November 17, 2023.
Champion, V.L., Skinner, C.S., 2008. The health belief model. Health behavior and health education: theory, research, and practice. 2008;4:45-65.
Cronin R.M. Jerome R.N. Mapes B. Development of the initial surveys for the all of us research program Epidemiology. 30 4 2019 597 608 31045611
Diabetes Prevention Program Research G. Knowler W.C. Fowler S.E. 10-year follow-up of diabetes incidence and weight loss in the diabetes prevention program outcomes study Lancet 374 9702 2009 1677 1686 19878986
Echouffo-Tcheugui J.B. Perreault L. Ji L. Dagogo-Jack S. Diagnosis and management of prediabetes: a review Jama 329 14 2023 1206 1216 37039787
Edwards A. Hood K. Matthews E. The effectiveness of one-to-one risk-communication interventions in health care: a systematic review Med. Decis. Making. 20 3 2000 290 297 10929851
Force U.S.P.S.T. Mangione C.M. Barry M.J. Screening for prediabetes and type 2 diabetes in children and adolescents: US preventive services task force recommendation statement JAMA 328 10 2022 963 967 36098719
Gallagher K.M. Updegraff J.A. Health message framing effects on attitudes, intentions, and behavior: a meta-analytic review Ann. Behav. Med. 43 1 2012 101 116 21993844
Herman W.H. Prediabetes diagnosis and management JAMA 329 14 2023 1157 1159 37039802
Joiner K.L. Adams M.P. Lee K.A. Piatt G. Davis M.A. Perceived risk for diabetes among U.S. adults with undiagnosed prediabetes Prev. Med. 160 2022 107089
Kandula N.R. Moran M.R. Tang J.W. O’Brien M.J. Preventing diabetes in primary care: providers’ perspectives about diagnosing and treating prediabetes Clin. Diab. 36 1 2018 59 66
Kreuter M.W. SGC Bull F.C. How does physician advice influence patient behavior? Evidence for a priming effect Arch. Fam. Med. 9 5 2000 426 433 10810947
Li E. Silverio A. Cunningham A. LaNoue M.D. Mills G. Association of prediabetes status awareness with behaviors and perception of health J. Am. Board Fam. Med. 34 1 2021 224 230 33452101
Liu C. Foti K. Grams M.E. Shin J.-I. Selvin E. Trends in self-reported prediabetes and metformin use in the USA: NHANES 2005–2014 J. Gen. Intern. Med. 35 2020 95 101 31637644
Murillo R. Katic B.J. Gonzalez T. Vasquez E. Echeverria S. The association of prediabetes and diabetes risk perception with leisure-time physical activity and weight loss Am J Health Promot. 33 4 2019 534 540 30270634
Pool A.C. Kraschnewski J.L. Cover L.A. The impact of physician weight discussion on weight loss in US adults Obes. Res. Clin. Pract. 8 2 2014 e131 e139 24743008
Roper K.L. Thomas A.R. Hieronymus L. Brock A. Keck J. Patient and clinician perceptions of prediabetes: A mixed-methods primary care study Diabetes Educ. 45 3 2019 302 314 31018784
Tseng E. Greer R.C. O’Rourke P. National survey of primary care physicians’ knowledge, practices, and perceptions of prediabetes J. Gen. Intern. Med. 34 11 2019 2475 2481 31502095
