How effective is MNT provided by Registered Dietitians in the management of type 1 and type 2 diabetes?

Background

Clinical trials and outcomes research has provided evidence for the effectiveness of Medical Nutrition Therapy (MNT) implemented by Registered Dietitians (RDs) on metabolic and behavioral outcomes.

Eighteen studies were reviewed to evaluate the effectiveness of medical nutrition therapy provided by an RD: eight randomized controlled trials (all positive-quality); two retrospective cohort studies (one positive-quality, one neutral-quality); one longitudinal study (neutral-quality); three time series studies (all neutral-quality); one descriptive study (negative-quality); one before-and-after study (neutral-quality); one narrative review (neutral-quality and one consensus report (neutral-quality).

Conclusion Statement

MNT has its greatest impact at diagnosis of diabetes (Monk et al, 1995; Delahanty et al, 1998).

Eight studies (Franz et al, 1995; DAFNE Study Group, 2002; Graber et al, 2002; Miller et al, 2002; Goldhaber-Fiebert et al, 2003; Wilson et al, 2003; Lemon et al, 2004; Gaetke et al, 2006), evaluating the effectiveness of diabetes MNT at three to six months, reported reductions in A1C, ranging from 0.25% to 2.9%, depending on the type and duration of diabetes. Individual sessions ranging from one to five or a series of 10 to 12 group sessions were employed. A variety of nutrition therapy interventions, such as a reduced energy and fat intake, carbohydrate counting, simplified meal plans, healthy food choices, individualized meal planning strategies, exchange lists, insulin-to-carbohydrate ratios and behavioral strategies were implemented. The number of initial and follow-up sessions varies in all the studies.

Studies reporting on effectiveness of MNT from six to twelve months (Lemon et al, 2004; DAFNE Study Group, 2002; Franz et al, 1995; Wolf et al, 2004; Banister et al, 2004; Chima et al, 2005; Bray et al, 2005) report a variety in the number and type of MNT sessions that lead to improved outcomes. Therefore, the RD needs to determine what is appropriate for individual clients.

Seven studies (DCCT, 1993; Laitinen et al, 1993; Maislos et al, 2002; Banister et al, 2004; Wolf et al, 2004; Bray et al, 2005; Chima et al, 2005) report sustained improvements in A1C at 12 months and longer. All involved regular sessions with an RD, ranging from monthly to three sessions per year. Seven studies (DCCT, 1993; Franz et al, 1995; Goldhaber-Fieber et al, 2003; Banister et al, 2004; Lemon et al, 2004; Wolf et al, 2004; Gaetke et al, 2006) report improvements in other outcomes, such as improved lipid profiles, weight management, decreased need for medications and reduced risk for onset and progression of comorbidities.

Evidence Summary

MNT at Diagnosis of Diabetes

In a neutral-quality narrative review by Delahanty et al, 1998, MNT implemented according to the Nutrition Practice Guidelines resulted in significantly greater reductions in HbA1c levels at 3 months than usual care (-1.00% vs -0.33%).

In a neutral-quality consensus report by Monk et al, 1995, Practice Guidelines Nutrition Care recommends that patients with non-insulin-dependent diabetes mellitus be referred to a dietitian within the first month after diagnosis.

Short-Term Effectiveness (Shorter than Six Months)

In a positive-quality randomized controlled trial by Miller et al, 2002, 92 individuals aged 65 years and older with type 2 diabetes for at least one year, found improved glycemic control and metabolic outcomes after 10 weekly sessions of RD intervention, incorporating principles from information processing, learning theory and Social Cognitive Theory and including written materials, reading food labels, grocery store visits, goal-setting, etc. HbA1c, in the experimental group, decreased from 7.2±0.1% at baseline to 6.7±0.1% after 10 weeks, while in the control group, HbA1c did not change (7.4±0.1% at both baseline and 10 weeks).

Goldhaber-Fiebert et al, 2003: In a positive-quality randomized controlled trial, 61 individuals with type 2 diabetes completed a 12-week lifestyle intervention program with tri-weekly walking groups and 11 weekly 90-minute nutrition classes by trained MS nutrition students. The course focused on portion control, use of healthier food substitutes, food groups, carbohydrates, fats and proteins, hidden calories in beverages and micronutrient and fiber content of fruits and vegetables. The intervention group lost weight and decreased plasma glucose, while the control group gained weight and increased plasma glucose. The change in glycosylated hemoglobin was -1.8±2.3% in the intervention group and -0.4±2.3% in the control group. In terms of weight change, the intervention group lost 1.0±2.2kg, while the control group gained 0.4±2.3kg.

Gaetke et al, 2006, completed a neutral-quality retrospective cohort study of 175 adults with type 2 diabetes and CVD, free of major medical complications, who had been referred to the same RD, not on lipid-lowering or hypoglycemic medications and who had received a single nutrition counseling session, composed of individualized instruction with the provision and explanation of printed materials. Records were reviewed to determine changes in clinical and anthropometric measures before seeing an RD and at a physician visit three months after seeing an RD. A significant number improved fasting blood glucose, their HbA1c and their total cholesterol, for a statistically significant difference at three months. Adjusted HbA1c for the MNT group at baseline was 10.1±0.3% and 7.2±0.3% at three months, for a mean difference of -2.9%. For the control group, adjusted HbA1c at baseline was 8.5±0.3% and 9.6±0.2% at three months, for a mean difference of +1.1%.

Wilson et al, 2003, in a positive-quality retrospective cohort study of Indian Health Service Diabetes Care and Outcomes Audit data from 2001, reviewed 7,490 records with clinical care data examined to evaluate effectiveness of clinical nutrition education when delivered by an RD, compared to a non-RD; clinical nutrition education by an RD is generally longer in duration and involves a nutrition assessment, goal setting, intervention and plans for follow-up. The measurement interval was 132 days. Clinical nutrition education was associated with improving HbA1c values. The mean difference in HbA1c for nutrition therapy when provided by an RD was -0.23±0.12%, when provided by both RD and non-RD, it was -0.18±0.15% and when provided by a non-RD, it was -0.10±0.09%.

Graber et al, 2002, in a neutral-quality before-and-after study, analyzed an initial pulse of intensive outpatient care, followed by subsequent surveillance, to determine long-term benefits of this pulse. The 12-week intensive program was composed of instructions and support in diabetes self-management, provided by diabetes educators and supervised by endocrinologists. There was an initial one-hour visit with a nurse, diabetes educator and dietitian and weekly contact was made through a clinic visit, a phone call, e-mail or fax exchange. The educational intervention content included diet, exercise, self-monitoring of blood glucose and medication adherence, as well as preventive measures, such as foot care and screening for complications. The population was initially 568 subjects, with 350 in this report (111 are not yet finished and are not included in this report), with type 1 and 2 diabetes and unsatisfactory glycemic control, frequent hypoglycemia and inadequate self-management. The mean number of nurse-educator visits was 3.8±1.7 and the mean number of dietitian visits was 1.7±1.1, for a mean total of clinic visits of 5.5±2.3. The mean decrease in HbA1c was 1.7% (95% confidence interval: 1.4% to 1.9%). Individuals with type 1 diabetes had the smallest mean decrease in HbA1c (1.1%), those with type 2 diabetes taking insulin had a mean decrease of 1.4% and those with type 2 diabetes not taking insulin had the greatest mean decrease of 2.3%.

Effectiveness at Six Months

In a positive-quality randomized controlled trial by Franz et al, 1995, 203 subjects with type 2 diabetes completed a six-month assignment to either MNT basic care (single RDvisit to establish nutrition care plan) or practice guidelines nutritional care (initial visit with RD, followed by two visits during the first six weeks of the study period, where the RD determined the appropriate nutrition prescription, educational intervention, materials and evaluation). At six months, both groups had similar improvements in blood glucose control (8.3% to 7.4% vs. 8.3% to 7.6%, respectively; both P<0.001) and weight loss (93.9kg to 92.4kg and 93.7kg to 92.0kg, respectively), but only participants in the practice guidelines nutritional care group had significant improvements in cholesterol level [reduction from 5.6 ± 1.2mmol per L (216.7 ± 46.4mg/dl) to 5.4 ± 1.0mmol per L (209.0 ± 38.7mg/dl).

In a positive-quality randomized controlled trial by the DAFNE Study Group, 2002, 136 individuals with type 1 diabetes were in a five-day course taught by trained diabetes dietitans, nurse specialists and dietitians to teach skills and confidence to adjust insulin to lifestyle, rather than being told to adapt the timing and content of meals to more fixed doses of insulin, with a follow-up at six months following the intervention. DAFNE training significantly improved hyperglycemia with no significant increase in severe hypoglycemia and produced sustained positive effects on quality of life. HbA1c at baseline for the initial intervention group was 9.4% and was reduced to 8.4% after six months. For the comparison group, HbA1c at baseline was 9.3% and 9.4% after six months.

Lemon et al, 2004, in a neutral-quality time series study of 244 physician-referred adults with type 2 diabetes at start (205 at three months, 208 at six months and 184 measured at baseline, three months and six months), investigated RDs providing nutrition education and counseling, according to the facility's policy. Use of standardized practice guides was not required. All subjects received face-to-face nutrition education and counseling at baseline. Additional sessions were not required, but were scheduled at the discretion of dietitian and subject. Intervention records were kept. Weight and glycemic control, coronary heart disease risk and self-management behaviors improved significantly between baseline and three months and between baseline and six months. Over six months, weight loss was -6.2±14.6lbs, reduction in HbA1c was -1.7±2.0% and coronary heart disease risk reduction was -3.5±6.1%.

Effectiveness at One Year

Wolf et al, 2004, in a positive-quality randomized controlled trial of 118 adults with type 2 diabetes, using medications, with BMI above 27, the RD case manager met with participants individually (six times, four hours total), in groups (six one-hour sessions) and by phone (monthly contact) for 12 months. Goals were tailored, but based on weight loss (5% of initial weight) and dietary intake as well as physical activity, reflecting national recommendations. At six months, more individuals in case management decreased total medications and fewer individuals increased total medications. At 12 months, case management participants were taking fewer total medications per day and reduced diabetes medications. Maximal weight loss occurred in a period of more frequent contact with RDs. Weight regain in the last four months suggests a need for ongoing lifestyle coaching. Weight loss for the intervention group was greatest at eight months (4.0kg) and by 12 months, the average weight loss was 2.4kg. The usual care group gained 0.6kg at 12 months. The difference in HbA1c between groups was greatest at four months (-0.57%), while the difference at eight months was -0.35% and -0.20% at 12 months.

Chima et al, 2005, in a negative-quality descriptive study, described a technology to evaluate the effectiveness of specific interventions and teaching methods. Sessions were taught by dietitians or nurse diabetes educators, with program options of including education on use of the blood glucose monitor, insulin teaching, heart-healthy classes and preconception guidelines. Outcomes were evaluated for patients who had participated in at least one class session, were one year out from program entry and had a HbA1c done at baseline and approximately one year. The population is described in relation to the tracking program: 438 entered the system, meaning they had at least one class; 216 had pre- and post-test HgbA1c; 72 were evaluated with pre-program data and one year follow-up; plus 200 random sample. Participants in the program appeared to achieve better blood glucose control than the general diabetes population. Mean HbA1c at baseline was 9.8±2.9% and 7.4±1.7% at approximately one year.

In a neutral-quality longitudinal study by Banister et al, 2004, 70 adults with type 2 diabetes attended a community clinic for a one-year program. Program participation varied between two and 12 months. During this time, subjects received education through a four-hour class, followed by individual dietitian consults and monthly support meetings. Class topics covered were a definition of diabetes, role of insulin, target ranges for fasting glucose, post-prandial glucose and HbA1c, symptoms of hypoglycemia and hyperglycemia, basic nutrition concepts, exercise strategies, foot care procedures, self-monitoring of blood glucose and goal-setting. Diabetes self-management education in a community clinic was associated with a 15% decrease in HbA1c, maintenance to decrease in medication use and limited average weight gain in the presence of medicines known to cause weight gain. 54 of the initial 70 participants met with the RD and initial HbA1c values ranged from 5.2% to 16.2%. Mean HbA1c improved from 9.7±2.4% to 8.2± 2.0%.

In a neutral-quality time series study, Bray et al, 2005, studied 160 rural African-Americans with type 2 diabetes with one of the following on the day of visit: HbA1c above 7.0, BP above 135/85 or evidence of high risk of end-stage organ disease. Over 12 months, a care re-design process with access to the full range of diabetic management services was undertaken. Patients were assigned to a group numbering from three to 12 subjects, who met for four two-hour group sessions over two months. These visits were led by a nurse, a physician, a pharmacist and a nutritionist and focused on an overview of diabetes, nutrition, medication and self-management and goal-setting. For the control group, charts were reviewed to identify those with the same conditions. More than 60% of patients in the intervention group responded with some reduction in their HbA1c. For the intervention group, median HbA1c at baseline was 8.2±2.6% and at the 12-month follow-up was 7.1±2.3%. For the control group, median HbA1c at baseline was 8.3±2.0% and 8.6±2.4% at 12 months. There were no significant differences in mean weight or blood pressure measurements between baseline and follow-up in the intervention group.

Long Term Effectiveness (Beyond One Year)

In a positive-quality randomized controlled trial by the Diabetes Control and Complications Trial Research Group, 1993, 1,398 subjects with type 1 diabetes received either conventional therapy (one or two daily insulin injections, clinic visits every three months) or intensive treatment (insulin pump or insulin more than three times per day, monthly clinic visits including RD visits), with education about diet and exercise. Subjects were followed for a mean of 6.5 years. Intensive diabetes therapy reduced the overall risk for onset and progression of diabetic retinopathy, nephropathy and neuropathy by approximately 50%.

In a positive-quality randomized controlled trial by Laitinen et al, 1993, 86 obese subjects with type 2 diabetes were assigned to the conventional group (follow-up every two to three months for usual education) or intervention group (clinic every two months, for a total of six times, for education by the physician, diabetes nurse and clinical nutritionist) for 15 months. Education was focused on principles of the diabetic diet, fat, carbohydrates and fiber, use of sweeteners, advantages and disadvantages of special products for diabetic patients, behavior modification and food preparation practice. In the first three months, fasting blood glucose and hemoglobin A1c decreased significantly in both groups. Between three and 15 months, the change in fasting blood glucose and HbA1c were significantly lower in the intervention group, compared to the conventional group (-7.2 mg/dl vs. 0.0 mg/dl and -0.6% vs. -0.3%, respectively).

In a positive-quality randomized controlled trial by Ash et al, 2003, 51 men with type 2 diabetes were randomized into one of three isocaloric groups and seen by a dietitian and physician weekly for 12 weeks, with a follow-up at 18 months. The three dietary interventions were focused on the removal of food cues (intermittent energy restriction group), the removal of food preparation cues (pre-portioned meals group) and usual dietetic intervention (self-selected meals group). The method of implementation made no difference at 12 weeks or 18 months. Intensive weekly contact with a health professional, in combination with moderate calorie restriction, facilitated success of short-term results. HgbA1c at baseline was 7.9±2.0% (N=50), at 12 weeks was 6.7±1.5% (N=39) and at 18 months was 8.3±2.3% (N=27).

Maislos et al, 2002, in a neutral-rated time series study of 492 subjects, who attended the Western Negev Mobile Diabetes Care Clinic, with two visits (first and last) within the two calendar years 1998 and 1999, reviewed and described a mobile diabetes clinic aimed to provide comprehensive interdisciplinary care to patients with diabetes in a semi-rural area, most of whom had never visited a diabetes center. A nurse, physician and dietitian saw the patient at the first visit. Educational programs were individualized and focused on the acquisition of skills using devices, the relevance of physical activity, appropriate diet and energy balance, foot care, etc. Mean HbA1c at the first visit was 9.6±2.4% and at the last visit was 8.7±3.3%.

Summary of the Research Regarding Effectiveness of MNT in Diabetes Management, Listed in Order of Study Type and Rating

View table in new window