OBJECTIVE—The purpose of this study was to describe the weight-loss practices and weight-related issues reported by youth with diabetes, according to sex and diabetes type.
RESEARCH DESIGN AND METHODS—A total of 1,742 female and 1,615 male youth aged 10–21 years with type 1 or type 2 diabetes completed a SEARCH for Diabetes in Youth study visit during which height, weight, and A1C were measured. A survey assessed weight-related issues and weight-loss practices.
RESULTS—Although more common in youth with type 2 diabetes, youth with type 1 diabetes also reported weight-related concerns and had elevated BMI. Among youth who had ever tried to lose weight (n = 1,646), healthy weight-loss practices (diet [76.5%] and exercise [94.8%]) were the most common, whereas unhealthy practices (fasting [8.6%], using diet aids [7.5%], vomiting or laxative use [2.3%], and skipping insulin doses [4.2%]) were less common. In sex-specific multivariable models including age, race/ethnicity, diabetes type, BMI category, and glycemic control, obese females and overweight/obese males were more likely to report ever practicing any unhealthy weight-loss practice than normal-weight youth. These practices were associated with poor glycemic control for female but not male subjects. All unhealthy weight-loss practices except fasting were more common in female than in male subjects. Dieting, fasting, and using diet aids were all more common in youth with type 2 diabetes than in those with type 1 diabetes.
CONCLUSIONS—Given the prevalence of overweight and obesity among youth with type 1 or type 2 diabetes, health care professionals caring for youth with diabetes need to pay particular attention to identifying youth, particularly females, with unhealthy weight-loss practices.
Diabetes is one of the three most prevalent chronic diseases of youth (1), with the majority of affected youth having type 1 diabetes (2). However, type 2 diabetes is being diagnosed more frequently in youth than has been reported in previous decades (2–4). Although youth with type 2 diabetes are likely to be overweight or obese, the increase in overweight in the U.S. population is mirrored among youth with type 1 diabetes (5,6). Strategies used to lose or manage weight include those that are healthy, such as regular physical activity and consuming a healthy diet, as well as those that are unhealthy, such as using over-the-counter diet aids without physician's advice, fasting, taking laxatives or diuretics, and vomiting. In 2005, 12.3% of 9th to 12th graders went without eating for at least 24 hours, 6.3% used diet pills, powders, or liquids, and 4.5% vomited or took laxatives to maintain or lose weight (7). Females were significantly more likely than males to use these unhealthy strategies; some racial/ethnic differences were observed.
Certain features of diabetes and its management, including weight gain after the initiation of insulin treatment, dietary restraint, and the knowledge that withholding insulin can cause weight loss, may trigger eating disturbances in youth with type 1 diabetes (8). Eating disorders have been associated with poor metabolic control and microvascular complications in type 1 diabetic youth (9–12). There is limited information about weight-related concerns among youth with type 2 diabetes. The American Diabetes Association recommends that youth with type 2 diabetes implement lifestyle modifications to reduce their intake of high-fat, high-energy foods and to increase physical activity to optimize glycemic control as well as their cardiovascular risk profile, including their lipid levels and blood pressure (13). At the same time, medical nutrition therapy must take in to account the nutritional needs required to support normal growth and development during childhood and adolescence (13,14). In this article, we describe the approaches to healthy and unhealthy weight-loss practices reported by youth with type 1 or type 2 diabetes by sex. In addition, we explore the associations between any unhealthy weight-loss practice, body weight perception, weight management goal, and worry about weight and glycemic control among youth with type 1 or type 2 diabetes by sex.
RESEARCH DESIGN AND METHODS—
SEARCH for Diabetes in Youth (SEARCH) is a multicenter study that began conducting population-based ascertainment of youth with clinically diagnosed diabetes who were <20 years of age in 2001 and is continuing to enroll youth with newly diagnosed (incident) diabetes (15). SEARCH recruited youth from four geographically defined populations, Indian Health Service beneficiaries from four American Indian populations, and enrollees in several managed health care plans. Institutional review board(s) for each site approved the study protocol. At the time of the SEARCH study visit, informed consent was obtained, physical measurements and fasting blood samples were obtained from metabolically stable participants (no episodes of diabetic ketoacidosis during the previous month) after a minimum 8-h overnight fast, and questionnaires were administered. Nonfasting blood samples were obtained from participants who had not fasted or had not withheld their medications but agreed to give a blood sample.
BMI (weight in kilograms divided by the square of height in meters) was calculated using height/weight measurements taken at the study visit. Percentiles for BMI z-scores specific to sex and age were assessed using algorithms prepared by the U.S. Centers for Disease Control and Prevention (CDC) based on the 2000 CDC growth charts (16,17). Youth were classified on BMI z-scores as follows: ≥95th percentile, obese; 85–94.9 percentiles, overweight; >5th to <85th percentiles, healthy weight; and <5th percentile, underweight (18).
With the use of questions similar to those in the Youth Risk Behavior Surveillance System (YRBSS) (7,19), youth aged ≥10 years were asked about their current weight (very underweight, slightly underweight, about the right weight, slightly overweight, and very overweight [perception of weight]), what they were trying to do about their weight (nothing, lose weight, gain weight, or trying to stay the same [weight management goal]), whether they worried a lot about their weight (yes/no), and whether they had ever tried to lose weight (yes/no).
Youth who reported ever trying to lose weight were asked whether they had tried to do so by any of the following methods: 1) eating less food, fewer calories, or foods lower in fat (dieting); 2) exercising; 3) going without eating for >24 h (fasting); 4) using diet pills, powders, or liquids without a doctor's advice (diet aids); 5) vomiting or taking laxatives; or 6) skipping insulin doses. Response 6 was added to be specific to diabetes. Using terminology consistent with YRBSS publications, diet and exercise were considered “healthy”; fasting, using diet aids without a doctor's advice, vomiting or taking laxatives, and skipping insulin doses were considered “unhealthy” (7,19). We created two dichotomous unhealthy composite measures: one that included fasting, diet aids, and vomiting or laxative use and a second with these three practices plus skipping insulin doses.
A1C was measured at the study visit and used to categorize glycemic control using the American Diabetes Association guidelines as good (A1C <8.0%), marginal (8.0–9.4%), and poor (≥9.5%) (20). Race/ethnicity was obtained through self-report using the standard census questions (21). Youth who reported Hispanic ethnicity were categorized as Hispanic, non-Hispanic youth were categorized by their race, and youth who reported multiple races or did not report race/ethnicity were categorized as other/unknown race/ethnicity. Highest parental education was based on the parent with the highest education as reported on a questionnaire. Type of diabetes was based on the clinical diagnosis by the physician. Youth with a clinical diagnosis of type 1 diabetes were much more likely to have a positive diabetes autoantibody and had much lower fasting C-peptide concentrations than youth with a clinical diagnosis of type 2 diabetes (3).
These analyses were restricted to youth with type 1 or type 2 diabetes who were ≥10 years of age at the time of their study visit. Of the 3,708 youth aged ≥10 years who completed a study visit and whose diabetes was prevalent (2001) or incident (2002–2005), exclusions were made sequentially for having diabetes type other than type 1 or 2 or having type missing (n = 39, 1.1%), height or weight missing (n = 222, 6.0%), and for not answering the weight-related or weight-loss questions (n = 90, 2.4%) for a final analytic sample of 3,357 youth.
Analyses, performed with SAS (version 9.1; SAS Institute, Cary, NC), were stratified by sex. We describe the demographic and clinical characteristics and weight-related issues of all youth in the sample as well as the weight-loss practices of youth who had ever tried to lose weight by diabetes type within sex categories and then compare youth by sex, combining youth with type 1 and type 2 diabetes. Using multiple logistic regression, we calculated adjusted odds ratios (AOR) and 95% CI to explore the associations between diabetes type, sex, and weight-loss practices. Then, associations between demographic characteristics, diabetes type, BMI percentile categories, and glycemic control categories and outcomes of any unhealthy weight-loss practices and the three weight-related outcomes were examined among youth with measured A1C.
The study population comprised 1,742 female and 1,615 male youth with a mean age of 15.0 years and a mean duration of diabetes of just >4 years (Table 1). The majority (84.5%) of the participants had type 1 diabetes. About 43% of the participants were overweight or obese, 39% perceived they were slightly or very overweight, ∼27% worried about their weight, and almost 40% were trying to lose weight. All of these characteristics varied by sex (P < 0.0001). As expected, youth with type 2 diabetes were more likely than youth with type 1 diabetes to be overweight or obese. However, 37.0% of females and 32.0% of males with type 1 diabetes were overweight or obese. Of the youth with type 2 diabetes, 38% (n = 125) of females and 34% (n = 65) of males reported taking insulin.
Just less than half of the youth surveyed reported ever trying to lose weight. Youth with type 2 diabetes were more likely than youth with type 1 diabetes to have ever tried to lose weight (90.5 vs. 52.4% for females and 79.3 vs. 32.1% for males) (Table 1). Of these 1,646 youth who reported ever trying to lose weight, healthy practices, including diet (76.5%) and exercise (94.8%), were very common. Females were more likely to report dieting (81 vs. 66%, P < 0.001), whereas males were more likely to report exercising (97 vs. 94%, P < 0.01) (Table 2). The prevalence of fasting was 8.6%, diet aid use was 7.5%, vomiting or laxative use was 2.3%, and skipping insulin doses (among insulin users) was 4.2%. Females were more likely than males to report any unhealthy weight-loss practices, including (P < 0.01) and excluding (P < 0.001) skipping insulin doses.
Dieting, fasting, and using diet aids were more common among type 2 diabetic youth than among type 1 diabetic youth for both sexes, whereas exercise, vomiting or laxative use, and skipping insulin doses were equally prevalent for youth with either type of diabetes (Table 1). Type 2 diabetic youth were more likely to report unhealthy weight-loss practices than type 1 diabetic youth for both sexes. After adjustment for age, highest parental education, and race/ethnicity, type 2 diabetic females were more likely to diet and fast to lose weight than type 1 diabetic females (Table 2). Although using diet aids, vomiting or laxative use, and skipping insulin were all more common among females with type 2 versus type 1 diabetes, the smaller number of females with type 2 diabetes and the low prevalence of these practices limited our power to detect a difference by type. Only dieting was more common among males with type 2 versus type 1 diabetes. In the model with both sexes combined, dieting, fasting, and using diet aids were all more common in youth with type 2 versus type 1 diabetes.
Correlates of any unhealthy weight-loss practice and weight-related issues
Separate multiple logistic regression analyses for females and males were used to identify the associations between the four outcomes (any unhealthy weight-loss practice, self-perception of overweight, trying to lose weight, and worry about weight) and covariates (age category, BMI category, diabetes type, glycemic control, and race/ethnicity) (Table 3). For both females and males, being obese or overweight (compared with healthy weight) was associated with all outcomes with two exceptions: being overweight was not associated with any unhealthy weight-loss practice for females nor with worrying about weight for males.
Among females, all four outcomes were associated with increasing age and having type 2 versus type 1 diabetes. Poor glycemic control (A1C >9.5%) was associated with reporting any unhealthy weight-loss practice (OR 1.82 [95% CI 1.23–2.70]) and self-perception of overweight (1.47 [1.06–2.04]) but not with trying to lose weight or worry about weight. Race/ethnicity was only associated with perception of being overweight or obese (P = 0.03). Among males, age was only associated with reporting any unhealthy weight-loss practice, with youth aged ≥19 years having a higher odds of ever reporting any practice compared with those aged 10–12 years (2.45 [1.04–5.77]). Having type 2 diabetes was associated with all outcomes. There was no association between any outcome and glycemic control. Race/ethnicity was only associated with worrying about weight (P = 0.01).
To our knowledge, this is the first report of weight-related issues and weight-loss practices comparing youth with type 1 and type 2 diabetes by sex. Whereas desiring to lose weight, worrying about weight, and having ever tried to lose weight were very common and not unexpected findings among type 2 diabetic youth, these characteristics were not uncommon among type 1 diabetic youth either, particularly among females, of whom 11% were obese. We found that reporting any unhealthy weight-loss practice was more common among type 1 than type 2 diabetic youth. Obese females and overweight/obese males were more likely to report any of these unhealthy practices than healthy weight youth. Among females but not males, there was an association between poor glycemic control and reporting any unhealthy weight-loss practice.
We compared the weight-loss practices of the 2,837 type 1 diabetic youth in SEARCH with results from diabetes clinic cohorts from Minnesota (n = 143), eastern Canada (n = 361), and Philadelphia (n = 295) (9,10,22). Exercise for weight-loss was more common in SEARCH participants (females 93.4%; males 96.9%) than among youth in Minnesota (89.9 and 47.9%, respectively) or Philadelphia (77.8 and 62.5%, respectively) for weight loss or maintenance (10,22). In the Canadian study, only 12% of females reported dieting for weight loss compared with 79% in SEARCH. Of the unhealthy practices, for females and males, 7.4 and 1.4% from the Minnesota study and 6.7 and 2.6% from Philadelphia, respectively, reported fasting, compared with 6.2% of females and 5.3% of males in SEARCH. Skipping insulin was more prevalent in the Minnesota (10.3%) than in the SEARCH (2.6%) or Philadelphia (1.5%) females but not in males (1.4, 1.3, and 1.3%, respectively). Skipping insulin was also more common in the female Canadian cohort (11%), although these researchers also included underdosing in their measure (23). Disordered eating in the Minnesota study was also associated with poor glycemic control in both sexes (10). In the Philadelphia cohort, older females with higher BMI and A1C exhibited significantly more weight-control behaviors (22). In SEARCH, we found that glycemic control was associated with any unhealthy weight-loss practice in females but not in males. The composite measure that we used for any unhealthy weight-loss practice in the current article differs from that in other studies. We found no published studies of weight-loss practices among type 2 diabetic youth with which we could compare our results.
The prevalence of healthy and unhealthy weight-loss practices reported by the SEARCH cohort could not be directly compared with the YRBSS because of the subgroups asked the questions (in SEARCH, those who had ever tried to lose weight; in YRBSS, those who tried for weight loss or maintenance) and different time frames specified for these questions (SEARCH, ever; YRBSS, past 30 days). In general, healthy practices, diet and exercise, were more common among SEARCH participants for weight loss than among YRBSS respondents for weight loss or maintenance (7). SEARCH males had a prevalence of unhealthy practices similar to that of those in YRBSS, whereas SEARCH females were more likely to fast and use vomiting or laxatives for weight-loss than females in YRBSS for weight loss or maintenance.
Of the youth aged ≥10 years in the SEARCH study cohort who completed a study visit, ∼21% were overweight and 22% were obese. Data from the National Health and Nutrition Examination Survey for 2003–2006 demonstrated that 16.5% of U.S youth aged 12–19 years were overweight (BMI ≥85th–<95th percentile) and 17.6% were obese (BMI ≥95th percentile), whereas 15 and 17% of youth aged 6–11 years were overweight and obese, respectively (24). The higher prevalence of overweight and obesity among youth in SEARCH compared with the general U.S. population is to be expected, given that 15% of youth in the SEARCH study have type 2 diabetes, a condition that is strongly associated with obesity.
Strengths and limitations
Our study includes >3,000 racially/ethnically diverse youth with diabetes, including >500 youth with type 2 diabetes, probably the largest cohort of its kind in the U.S. Our data allowed us to compare the weight-related issues and weight-loss practices of youth with type 1 diabetes with those of youth with type 2 diabetes and by sex using the same methodology and survey questions. This study has several limitations. Because weight-loss practices may have occurred recently or in the distant past, we were not able to report associations between clinical indicators and individual weight-loss practice. Instead, we used a composite measure of “any unhealthy weight-loss practice” as a marker for a history of such unhealthy behavior. Additionally, we could not assume that the respondents’ current BMI category was reflective of BMI at the time of these practices or whether these practices occurred before or after diabetes diagnosis. Although we characterized eating less food, fewer calories, or foods lower in fat as a healthy weight-loss practice, some youth may have overrestricted their food/calorie intake. Finally, despite extensive efforts to optimize recruitment, ∼47% of eligible youth completed the SEARCH study visit (25).
Summary and clinical implications
In this cohort of youth with type 1 and type 2 diabetes, overweight and obesity were common and 60% of females and 38% of males reported ever trying to lose weight. Healthy weight-loss practices such as dieting and physical activity were quite common among these youth with type 1 and type 2 diabetes. Youth with type 2 diabetes were more likely than those with type 1 diabetes to report using any unhealthy weight-loss practice; females exhibited a higher prevalence of unhealthy practices than did males. In addition, among females, poor glycemic control was associated with reporting any unhealthy weight-loss practice and perceiving that they were overweight.
Given the high prevalence of overweight and obesity among type 2 diabetic youth, the increasing prevalence of these conditions in type 1 diabetic youth (5,6), and the pressures on some overweight and obese youth to lose weight, it is likely that approaches to weight management will not always be healthy ones. Such practices may have a negative effect on diabetes management, including glycemic control, a risk factor for future complications. Physicians and other health care professionals caring for diabetic youth, particularly females, need to identify those with unhealthy weight-loss practices and provide them with more healthy weight-management strategies in the context of their ongoing diabetes management.
SEARCH for Diabetes in Youth is funded by the Centers for Disease Control and Prevention (PA 00097 and DP-05-069) and supported by the National Institutes of Health, Kaiser Permanente Southern California (U01 DP000246), University of Colorado Health Sciences Center (U01 DP000247), Pacific Health Research Institute (U01 DP000245), Children's Hospital Medical Center (Cincinnati) (U01 DP000248), University of South Carolina (U01 DP000254), University of Washington School of Medicine (U01 DP000244), and Wake Forest University School of Medicine (U01 DP000250). In addition, the authors acknowledge the involvement of General Clinical Research Centers at the following institutions: Medical Center of South Carolina (Grant M01 RR01070), Cincinnati Children's Hospital (Grant M01 RR08084), Children's Hospital and Regional Medical Center and the University of Washington School of Medicine (Grants M01RR00037 and M01RR001271), and Colorado Pediatric General Clinical Research Center (Grant M01 RR00069).
Preliminary results of this study were presented in poster format at the 66th annual meeting of the American Diabetes Association, Washington, DC, 9–13 June, 2006.
Published ahead of print at http://care.diabetesjournals.org on 22 September 2008.
The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention and the National Institute of Diabetes and Digestive and Kidney Diseases.
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/licenses/by-nc-nd/3.0/ for details.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.