| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1177/0145721705284743
Low–Glycemic Index CarbohydratesAn Effective Behavioral Change for Glycemic Control and Weight Management in Patients With Type 1 and 2 DiabetesFrom Nutrition Works, Mendham, New Jersey (Ms Burani), and the University of Massachusetts Dartmouth (Dr Longo). Correspondence to Johanna Burani, MS, RD, CDE, Nutrition Works, 51 Dean Road, Mendham, NJ 07945 (jburani{at}gmail.com).
Purpose This retrospective study evaluated the incorporation of low–glycemic index (GI) carbohydrates into daily meal planning as an effective behavioral lifestyle change to improve glycemic control and weight management in patients with type 1 and 2 diabetes. Methods Twenty-one subjects participated in this study. All office visits and interview sessions took place in a 2-physician private medical practice setting in Wayne, New Jersey. Patients' pre- and postcounseling HbA1c and body mass index (BMI) values and their antidiabetic medication dosages were recorded. Audiotaped interviews were conducted using the 10-question Glycemic Index Foods Quiz (GIFQ) and the 29-question Interview Questionnaire (IQ). The GI values of pre- and postcounseling meals were calculated. Assessment was based on triangulating the subjects' adherence to the low-GI carbohydrate behavioral change and the primary outcome measures: HbA1c and BMI. Results Low-GI medical nutrition therapy (LGI-MNT) counseling reduced HbA1c by 19% (mean drop of 1.5 U) and decreased BMI by 8% (mean loss of 17 pounds). This was accomplished by the participants independently lowering the GI values of their meals by 25% (mean reduction of 15 points). Results were achieved over a time frame of 3 to 36 months from the initial LGI-MNT counseling session. Conclusions Daily incorporation of low-GI carbohydrates in meal planning can be an effective diabetes self-management strategy for glycemic control and weight management. The documented responses to the subjects' conceptual and practical knowledge of the GI confirm their acceptance of this approach as a permanent behavioral lifestyle change and not a "diet." The positive results of this study attest to what worked for these subjects, inviting diabetes educators to consider offering low-GI dietary advice to their diabetes patients.
Receiving a diabetes diagnosis puts the patient on notice that some lifestyle changes should be made to improve health. Health care providers inform the patient about the triad of glycemic management options: diet, exercise, and medication. To improve overall glycemia and preserve quality of life, the patient soon realizes that food choices and weight management will affect these goals. Recent national statistics confirm that Americans are living in concurrent diabetes and obesity epidemics, with no abatement in sight. The prevalence of diabetes in the United States has doubled between 1980 and 2003, from 5.8 million to an estimated 13.8 million Americans,1 and another 41 million are estimated to have prediabetes.2 Of equal concern, 65% of adult Americans are overweight or obese, and more than 60 million of these are obese.3 To best address the gravity of these statistics, diabetes educators look for flexible, sensitive, and creative ways to motivate their patients to initiate behavioral changes that will foster and maintain improved health. The currently recommended approach for diabetes education is to focus on behavior-centered, rather than knowledge-centered, interventions that are feasible in the "real world."4 Following this approach, the educator presents basic information and offers appropriate skills training for self-management. It is then the patient who decides what, when, and how to incorporate this knowledge into his or her lifestyle. What does the patient gain from making self-directed, informed decisions about behavioral changes? Anderson et al5 provide evidence indicating that glycemic control and quality of life may improve with this type of patient empowerment model. When recommended strategies to improve the patient's health are perceived as meaningful and feasible to the patient and the focus moves away from technical information and toward constructive behavioral modifications that the patient identifies and chooses to implement, the educator has succeeded in empowering the patient, and improved outcomes are more likely to result. Diabetes is a disorder of carbohydrate metabolism. Thus, a balanced diet, with careful attention given to the carbohydrates consumed, plays a prominent role in improving glycemia and other diabetic outcomes. As recently as 2004, the carbohydrate recommendations as stated by the American Diabetes Association (ADA) in their clinical practice recommendations for that year6 focused on the absolute importance of the total amount of carbohydrate consumption per meal or snack and attributed little importance to its food source. However, since evidence continues to accumulate on the efficacy of the type of carbohydrate consumed in promoting normal or improved glycemia in diabetes patients,7,8 the 2005 ADA recommendations now indicate that "use of the glycemic index/glycemic load can provide an additional benefit over that observed when total carbohydrate is considered alone."9 First presented by Jenkins et al10 in 1981, the glycemic index (GI) ranks carbohydrates based on their postprandial glucose excursion rate into the bloodstream as compared to a reference food (pure glucose or white bread). Quickly digested high-GI carbohydrates ("gushers") produce a greater and more rapid rise in blood glucose than the more slowly digested low-GI ones ("tricklers").11 Today, approximately 1750 GI values of tested foods can be found on an online database (www.glycemicindex.com). Alternatively, The New Glucose Revolution provides another reference source for published GI values.12 The purpose of this retrospective study was to evaluate the incorporation of low-GI carbohydrates into daily meal planning as an effective behavioral lifestyle change to improve glycemic control and weight management in patients with type 1 and 2 diabetes.
Setting and Sample All individuals recruited for this study were adult patients with type 1 or 2 diabetes referred by their primary care physician or endocrinologist for medical nutrition therapy for diabetes. A review of 199 medical charts identified 23 eligible individuals (12%) who met the inclusion criteria. Of those excluded, 74 patients (37%) did not complete the required 2 hours of medical nutrition therapy (MNT); 50 patients (25%) gained weight, 23 patients (12%) did not show improved HbA1c levels, 15 patients (8%) required increased medication, and 14 patients (7%) either refused participation, could not be contacted, were too sick, or had died. The final number of participants was 21; 2 participated in the study but were subsequently disqualified because of inaccurate initial information that, when corrected, did not meet all of the inclusion criteria. Twenty were living at home, either in urban or suburban northern New Jersey communities, and 1 was a third-year college student living in an off-campus apartment in southern New Jersey. To be eligible for this study, participants had to meet the following criteria: (1) diagnosis of type 1 or type 2 diabetes, (2) stable or improved current HbA1c and body mass index (BMI) values after initiation of low-GI MNT (LGI-MNT) with no increase in number and/or dosages of oral antidiabetic drugs (OADs) and/or insulin, and (3) a minimum of 2 hours of LGI-MNT counseling from the investigator within the time frame of September 1, 2001, through May 1, 2004. Office visits occurred in a 2-physician private medical practice setting in Wayne, New Jersey, where the primary investigator has been an independent nutrition consultant for 16 years. All eligible subjects received a formal written invitation, and upon acceptance, all signed a written informed consent form prior to their participation in the study. An interview date was scheduled by phone. The descriptive and clinical demographic characteristics are represented in Tables 1 and 2.
The time lapse from a subject's initial LGI-MNT session to the interview date was between 3 and 36 months; the average number of MNT office visits was 5, and the average counseling time was 3.6 hours.
Data Collection Procedures
Primary Outcome Measures: HbA1c and BMI
Antidiabetic Medications and GI Values of Meals
Glycemic Index Foods Quiz
Interview Questionnaire A one-on-one structured interview, consisting of the GIFQ and the IQ, was conducted by the investigator with each participant. The interview time varied from 45 to 60 minutes. No subject received any review of GI information. Their answers were based exclusively on what they remembered from their LGI-MNT counseling sessions, 62% of which occurred more than 1 year prior to the interview date, and also on the carbohydrate choices they subsequently had been making based on that knowledge. The investigator was not present while the participants were recording their answers except to assist the 2 octogenarian subjects with discrete visual impairment and 1 Hispanic subject with a minor language barrier. In these cases, every attempt was made through neutral body language and voice tone to neither encourage nor discourage certain responses from the subjects. All responses and comments were audiotaped and transcribed verbatim and coded for accurate data collection.
Enhancing Trustworthiness of the Interview Findings As a measure of quality control, all interview texts and test answers were independently reviewed by 2 registered dietitians, certified diabetes educators who were not associated with the study. Both signed a form attesting to the candor of the investigator's data representation.
Data Analysis Since the GIFQ is a short 10-item measure, an interitem correlation was used to determine the reliability of the GIFQ. A t test was used to compare the test-retest results of the GIFQ for the 6 recalled subjects. Descriptive statistics were used to explain the qualitative findings of the IQ responses.
Summary of Quantitative Data Primary Outcome Measures: HbA1c and BMI Two paired-samples t tests were conducted to evaluate the impact of LGI-MNT counseling on the HbA1c and BMI statuses of the 21 participants (Table 3).
HbA1c
Body Mass Index An 8% reduction in BMI (mean weight loss of 17 lb) was calculated. Sixteen subjects (76%) succeeded in losing weight, 4 subjects (19%) achieved stable weight, and 1 subject (5%) gained weight. This final subject's weight gain was tolerated because the BMI remained less than 25.0. There was a statistically significant change from the precounseling BMI values (mean = 35.1, SD = 11.4) to the postcounseling BMI values (mean = 32.3, SD = 9.1; t[20] = –3.7, P = .002). The  2 statistic (0.40)
indicated a very large effect size. Figure
2 illustrates the BMI improvement of these statistical
differences.
Antidiabetic Medications A comparison of the pre- and postcounseling dosages of antidiabetic medications is highlighted in Table 4. On average, all 3 type 1 subjects reduced their daily insulin dosages (premeals and bedtime) postcounseling between 33% to 58%. One of the type 1 subjects wears an insulin pump and was able to increase his insulin-to-carbohydrate ratio from 1:15 to 1:20. The type 2 subject on insulin also achieved a dosage reduction of 20%. It is noteworthy that all 4 subjects who started with no OADs experienced no change in that status and 2 subjects who were taking an OAD prescription precounseling were able to discontinue them completely. The other notable result is that 2 subjects decreased their OADs from 2 medications to 1 after counseling. Eight subjects experienced no change in their prescriptions; 1 subject's metformin (Glucophage; Bristol-Myers Squibb, Princeton, NJ) dose was increased; this was acceptable because this OAD frequently requires upward titration from the initial dosage based on patient tolerance. There were no consistent differences noted in caloric and carbohydrate intakes between pre- and postcounseling meals: 11 subjects consumed fewer calories, and 8 consumed fewer carbohydrates in their postcounseling meals.
GI Values of Meals
Prior to LGI-MNT instruction, 5 subjects (24%) were haphazardly following a low-GI diet (either their spouse had already received LGI-MNT or some of their personal food preferences happened to be low-GI), 14 (67%) were in the intermediate-GI range, and 2 (10%) were consuming high-GI meals. It is noteworthy that several subjects started making changes in their diets when they were first diagnosed with diabetes even before initiating medical nutrition therapy by eating more whole grains, fruits, and vegetables and less refined starches and concentrated sweets. Twenty subjects (95%) succeeded in lowering the GI values of their daily food intake. One subject, with type 1 diabetes and a remarkable history for severe hypoglycemic episodes, showed a 6% increase in overall GI but still remained in the low-GI category. At the end of the study, all participants were eating low-GI meals and snacks (average postcounseling daily GI value = 45).
Glycemic Index Foods Quiz
Summary of Qualitative Data
This retrospective study set out to evaluate whether LGI-MNT counseling could improve glycemic control and weight management in noninstitutionalized patients with type 1 and 2 diabetes. Five pre- and postcounseling quantitative data sources were compared to attain this objective: HbA1c, BMI, antidiabetic medications, GI values of meals, and GIFQ scores. One qualitative data source was also compared, the IQ. Since the primary inclusion criteria dictated that participants must have achieved improved glycemic control (HbA1c) and a stable or reduced weight status (BMI), only successful patients were tracked in this study to evaluate the impact of LGI-MNT on their success. With regard to weight management, a recently published study14 failed to show a significant improvement with GI education in a 1-year posttreatment weight loss in a behavioral weight management program. However, there are animal15 and human16-18 intervention studies that point to a positive impact of low-GI diets on weight loss, resting energy expenditure, and body fat deposition. These studies support the weight loss results of this intervention. The participants were questioned in depth whether they believed learning about high-/low-GI carbohydrates helped them with their glycemic control and weight management. Their documented responses were overwhelmingly positive about the usefulness of knowledge about the GI. Of paramount interest to this study was the participants' unanimous 100% consensus regarding behavioral changes in their lifestyles: (1) behavioral changes work better than going on a diet for their diabetes control and weight management, (2) choosing low-GI carbohydrates helped them improve their diabetes and weight statuses, and therefore, (3) they plan on continuing to use this efficacious self-management strategy.
Strengths of the Study The subjects are either living at home or away at college, making ad libitum carbohydrate choices at every meal and snack opportunity. This was not a clinical trial but rather a report of what these patients chose to do for themselves "out there" while living their daily lives. Their choices were influenced by knowledge about the GI concept that they received 3 to 36 months prior to the onset of the study. The positive primary outcome measures of these participants (HbA1c and BMI) attest to the efficacy of the patient empowerment model.
Limitations of the Study Hopefully, future studies will address these issues and others, such as waist circumference measurements and the glycemic load of meals.
Implications for Practice Results of this study also support the value of the patient empowerment model. In most cases, patients want to succeed in their diabetes self-management care. When the diabetes educator presents practical behavior-centered information in a nonthreatening manner, the patient is more likely to accept this guidance and feel motivated toward positive change. Learning to incorporate low-GI carbohydrates affords patients a practical skill that is within their grasp. It empowers them to "own" their diabetes and actively contribute to their control over it.
The authors wish to thank Jody Lashen, MS, RD, CDE, and Diane Misner, RD, CDE, for reviewing all interview texts and test answers. Additional appendices are available upon request.
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
 |
Sign In to gain access to subscriptions and/or personal tools.
Top
Abstract
Research Design and Methods
