Glycemic Relapse in Type 2 Diabetes

[Guest guest]

Colleagues, the following is FYI and does not necessarily reflect my own

opinion. I have no further knowledge of the topic. If you do not wish to

receive these posts, set your email filter to filter out any messages

coming from @nutritionucanlivewith.com and the program will remove

anything coming from me.

---------------------------------------------------------

Glycemic Relapse in Type 2 Diabetes

http://www.medscape.com/viewarticle/533358

Alan L. Graber, MD, FACE, FACP; Ayumi K. Shintani, PhD, MPH; Kathleen

Wolff, RN, MSN; Anne Brown, RN, MSN; Tom A. Elasy, MD, MPH

Endocr Pract. 2006;12(2):145-151. ©2006 American Association of

Clinical Endocrinologists

Posted 06/05/2006

Abstract and Introduction

Abstract

Objective: To characterize the occurrence of glycemic relapse after

initial improvement in blood glucose levels and to describe predictors

of relapse in patients with type 2 diabetes.

Methods: Occurrence of glycemic relapse was analyzed in 393 consecutive

patients with type 2 diabetes after participation in a 3-month intensive

outpatient intervention. All patients had hemoglobin A1c (A1C) values

≥7% before the intervention and had achieved adequate glycemic control

(nadir A1C <7%) afterward. The median follow-up time was 26.5 months.

Relapse was defined as a subsequent increase in A1C to ≥7%.

Results: The probability of glycemic relapse was 45% at 1 year after the

intervention and was 76% at 3 years. The median time to relapse was 15.2

months. multivariate regression analysis indicated that treatment

with insulin was associated with a greater risk of relapse— hazard ratio

= 1.5 (95% confidence interval, 1.1 to 2.2), after controlling for the

patient's age, sex, race, body mass index, duration of diabetes, weight

change during the intervention, and nadir A1C value. Among those

patients not treated with insulin at the end of the intervention, a

shorter duration of diabetes and weight loss during the intervention

period were significantly associated with decreased risk of relapse.

Conclusion: The majority of study patients with type 2 diabetes who

attained satisfactory glycemic control after intensive outpatient

intervention had a relapse after the end of the intervention period.

Patients receiving insulin therapy were at particular risk of glycemic

relapse. Therefore, such patients should receive high priority for

continuation of intensive care or for other relapse prevention measures.

Introduction

Relapse is defined as " a recurrence of symptoms of a disease after a

period of improvement " .[1] Relapse in diabetes may be defined as the

deterioration of glycemic control after a period of improvement.

Although relapse has seldom been studied in the clinical setting of

diabetes mellitus, prevention of relapse has been the focus of intensive

studies concerned with addictive behaviors and with depression.[2,3]

Many studies have demonstrated that intensive management programs

improve the outcomes of patients with diabetes. Nevertheless, prolonged

maintenance of adequate glycemic control for a several-year period has

seldom been demonstrated, except in volunteers in resource-rich clinical

trials such as the Diabetes Control and Complications Trial,[4] the

United Kingdom Prospective Diabetes Study,[5] and the Kumamoto Study.[6]

As an alternative to the long-term continuous intensive care used in

such clinical trials, we have examined the effects of a 3-month

intensive outpatient intervention in type 2 diabetes, without the

program assuming subsequent long-term management. After the intervention

had achieved satisfactory levels of glycemic control, subsequent

surveillance determined the incidence of relapse. The purposes of this

report were to describe the incidence and time course of glycemic

relapse and to determine which variables might identify those patients

who are most likely to have a relapse.

Subjects and Methods

Research Design and Definition of Terms

Adult patients with type 2 diabetes mellitus were referred by their

primary care physicians or by a consulting endocrinologist to a 12-week

outpatient intervention of intensive diabetes care at an academic

medical center. Patients were referred because of lack of knowledge of

diabetes self-management skills, inadequate self-management practices,

hyperglycemia refractory to efforts by their primary care physicians,

recurrent or severe hypoglycemia, or the need to initiate insulin or

insulin pump therapy. The subjects in this report were 393 consecutive

patients with type 2 diabetes who had completed the intervention by

December 31, 2002, and had achieved a nadir hemoglobin A1c (A1C) of <7%

after completion of the intervention. All study patients had A1C values

≥7% at entry, a level at which the American Diabetes Association

suggests additional action.[7] Patients who had not achieved a nadir A1C

of <7% at completion of the intervention were excluded from this study.

Nadir A1C was defined as the lowest A1C level achieved within 6 months

after completion of the intervention. A1C was determined at program

entry, at completion of the intervention, and approximately every 6

months after the intervention. Glycemic relapse was defined as an

increase of A1C to ≥7% after the nadir had been reached. A1C was

measured by high-performance liquid chromatography, with a normal range

of 4.0% to 6.1%, in conformity with the National Glycohemoglobin

Standardization Program.

The program was approved by the Institutional Review Board of Vanderbilt

University Medical Center.

Intervention Program

Each patient participated in a 12-week intensive out-patient program,

consisting of instruction and support in diabetes self-management

coupled with adjustment of insulin and orally administered hypoglycemic

medications. The details of the intervention program have been described

previously.[8-10] At least weekly contact occurred between the patient

and an advanced practice diabetes nurse specialist or a dietitian to

recommend changes in insulin dose or other diabetes medications (or

both). The weekly contact consisted of a clinic visit, a telephone call,

or an exchange of electronic mail (e-mail) or faxed messages regarding

home blood glucose results.

Protocols for use of insulin and oral medications were developed and

supervised by an endocrinologist (A.L.G.) and were similar to previously

published recommendations.[11] When glycemic control was not achieved

with lifestyle adjustments and oral hypoglycemic drug therapy, insulin

treatment was begun, usually with a bedtime dose of intermediate- or

long-acting insulin. The usual starting dose was approximately 0.3 U/kg,

and the dosage was adjusted until the glycemic target (fasting plasma

glucose level ≤125 mg/dL) was attained. Subsequently, if post-meal

hyperglycemia persisted, a rapid-acting insulin analogue, lispro or

aspart before meals, was added in patients willing to intensify therapy

and capable of doing so. The starting meal dose was either 0.1 U/kg or

based on the carbohydrate content of the meal. When insulin therapy was

initiated, oral administration of hypoglycemic drugs was usually

continued. Every effort was made to empower patients to self-adjust

their insulin dose on the basis of results of capillary blood glucose

monitoring.[12] After completion of the program, patients returned to

the usual customary care provided by their primary care physician or

endocrinologist.

Statistical Analysis

Data for characteristics of patients were presented as mean values and

SD for normally distributed continuous variables and as median values

and interquartile range (IQR) for nonnormally distributed continuous

variables. For categorical variables, frequencies and percentages were

used. Wilcoxon rank sum tests were used to compare distributions of

continuous variables, and chi-square tests were used to compare the

proportions of categorical variables between insulin-treated and

non-insulin-treated patients.

All patients underwent follow-up for 3 years from the nadir A1C value or

until the time of last clinic visit within the 3-year follow-up period.

For all analyses, the primary outcome variable was the time (months from

the nadir A1C) to the first event of glycemic relapse. Patients were

censored (removed from the analysis) at the last clinic visit or at the

end of the follow-up period.

Kaplan-Meier curves were used to present the probability of a

relapse-free condition over time, and the log-rank test was used to

assess univariate association between a risk factor and the outcome

variable. proportional hazards regression models were used for

multivariate analysis, and proportionality of the hazard was assessed

graphically.

All analyses were conducted among all 393 patients and were also

stratified by use of insulin at the end of the intervention. Risk

factors used in the analysis included patient's age, sex, race, body

mass index (BMI), duration of diabetes in years, history of hypoglycemia

before the intervention, area of residence (Nashville or outside

Nashville), A1C value at the nadir, weight change during the

intervention, and insurance status.

Patient's age, sex, and race were chosen a priori to be included in the

multivariate models, and BMI, duration of diabetes, insulin use at the

end of the intervention, weight change during the intervention, and the

nadir A1C value were chosen to be in the multivariate models because

their P value was less than 0.2 in the univariate analysis of the entire

study cohort. All continuous variables were assessed by using nonlinear

splines in the regression as well as categorization with clinically

meaningful cutoff values.[13]

During the follow-up period, the frequency of A1C measurements varied

among patients. Therefore, the results from the time to event analysis

could be confounded by the monitoring bias. For example, time to relapse

could be interpreted as shorter if a patient were monitored more

frequently. Therefore, we assessed frequencies of A1C measurements by

the baseline risk factors referred to in the aforementioned analysis

with use of Poisson regression.

For all analyses, a 2-sided significance level of 5% was used for

statistical inferences, and SAS version 8.2 and SPSS version 12.0 were used.

Results

Patient Characteristics

The mean ± SD age was 54.8 ± 12.8 years. Of the overall study group of

393 patients, 50% were female, and 25% were African American. The median

duration of diabetes (and IQR) was 1.1 (0.1, 5.7) years. Before

intervention, the mean BMI was 35.0 ± 8.2 kg/m2, and the mean A1C was

9.3 ± 1.9%. After intervention, the mean A1C at nadir was 6.3 ± 0.5%.

The median (IQR) follow-up time was 26.5 (16.9, 36.0) months.

Twenty-three percent of patients (N = 90) were receiving insulin therapy

before intervention, 13% (N = 52) initiated insulin treatment during the

intervention, and 4% (N = 16) discontinued the use of insulin during the

intervention; thus, 32% (126 patients) were receiving insulin treatment

at the conclusion of the intervention. For those patients taking insulin

at the end of the intervention, the mean total daily insulin dose was

0.58 U/kg. Patient characteristics were similar between those receiving

and those not receiving insulin at the end of the intervention, except

for duration of diabetes and weight change during the intervention. The

median duration of diabetes (and IQR) was 0.3 (0.1, 3.6) year for

patients not receiving insulin at the end of the intervention and was

5.3 (1.9, 10.0) years for those receiving insulin. The mean weight

change during the intervention period was a reduction of 4.8 ± 8.1

pounds for patients not receiving insulin and a gain of 2.4 ± 11.2

pounds for patients receiving insulin at the end of the intervention (

Table 1 ).

Glycemic Relapse

The probabilities of glycemic relapse were 45% and 76% at 1 and 3 years,

respectively, and the median time to relapse (and IQR) was 15.2 (5.3,

35.7) months. The incidence rate of relapse was 0.52 per person-year.

Multivariate regression analysis indicated that treatment with

insulin was associated with a greater risk of relapse—hazard ratio (HR)

= 1.5 (95% confidence interval [CI], 1.1 to 2.2), after controlling for

the patient's age, sex, race, BMI, duration of diabetes, weight change

during the intervention period, and nadir A1C value. For the patients

who were not receiving insulin therapy at the end of the intervention,

the probabilities of glycemic relapse were 38% and 70% at 1 and 3 years,

respectively, and the median time to relapse (and IQR) was 19.9 (6.2,

upper extent of IQR not available [NA]) months. The incidence rate of

relapse was 0.42 per person-year for this subgroup. For the patients who

were receiving insulin therapy at the end of the intervention, the

probabilities of glycemic relapse were 56% and 86% at 1 and 3 years,

respectively, and the median time to relapse (and IQR) was 7.7 (3.4,

20.3) months. In this subgroup, the incidence rate of relapse was 0.84

per person-year (Fig. 1).

Among those patients not treated with insulin at the end of the

intervention, longer duration of diabetes was associated with increasing

risk of glycemic relapse. The adjusted HR was 2.0 (95% CI, 1.2 to 3.2)

for longer than 5 years' duration versus less than 1 year, and it was

1.6 (95% CI, 1.0 to 2.5) for a duration of diabetes of 1 to 5 years

versus less than 1 year. The median time to relapse (and IQR) was 30.4

(11.7, NA) months for less than 1-year duration of diabetes, 12.9 (6.0,

NA) months for 1 to 5 years, and 9.0 (4.1, 21.0) months for longer than

5 years' duration (Fig. 2). No significant association was observed

between the duration of diabetes and the risk of glycemic relapse among

patients receiving insulin at the end of the intervention.

Weight loss during the intervention period was associated with decreased

risk of glycemic relapse among patients not treated with insulin. The

adjusted HR was 0.6 (95% CI, 0.3 to 1.0) for losing 10 pounds or more

versus weight gain and was 0.6 (95% CI, 0.4 to 0.9) for losing up to 10

pounds versus weight gain. The median time to relapse (and IQR) was 13.2

(6.2, 31.4) months for those with weight gain, 21.0 (7.2, NA) months for

those with 0 to 10 pounds of weight loss, and 23.7 (11.5, NA) months for

those with 10 pounds or more of weight loss (Fig. 3). No significant

association was observed between the weight change during the

intervention and the risk of glycemic relapse among patients receiving

insulin at the end of the intervention.

Analysis of the frequency of A1C measurements indicated that none of the

risk factors considered in the multivariate regression analysis was

associated with frequency of A1C measurements.

Discussion

Our data demonstrate that most study patients had a glycemic relapse

after the intensive outpatient intervention, when they returned to their

previous customary care. This study extends our preliminary analysis[14]

by including a larger cohort of patients, a longer follow-up period, and

different criteria for relapse, which corresponded to the lower glycemic

targets that have been adopted by the American Diabetes Association.[7]

We were concerned that establishing these criteria for relapse could

include patients whose A1C increased only slightly from nadir to ≥7%,

but when we analyzed the data using criteria that defined glycemic

relapse as both A1C ≥7% and an increase from nadir by at least 1%, the

results were similar.

In this cohort, insulin therapy identified a group of patients at

particularly high risk for relapse. The difference in probability of

relapse between the 2 groups (use of insulin versus no insulin) is

evident within the first year (Fig. 1 . The patients requiring insulin

had a much longer duration of diabetes (median of 5.3 years for insulin

users versus 0.3 year for nonusers) ( Table 1 ). Therefore, the nonusers

of insulin were largely a group of patients who had been referred to the

interventional program shortly after the diagnosis of diabetes. In these

patients not using insulin, the observed significant association between

the duration of diabetes and the risk of glycemic relapse suggests a

process that worsens with progressive duration. In the group of patients

requiring insulin therapy, we no longer observed a duration-associated

difference in median time to relapse (Fig. 2).

When glycemic control improves, patients with diabetes have a well-known

tendency for weight gain.[11] In our patients not receiving insulin, the

median time to glycemic relapse was shorter in those patients who gained

weight during the intervention and was longer in those who lost weight

during this period (Fig. 3). This finding indicates a need for

concomitant intensive medical nutrition therapy to accompany any regimen

that improves glycemic control in type 2 diabetes.

Patient acquisition of adequate skills in the long-term management of

insulin therapy may necessitate more than a 12-week period of intensive

contact with providers. It is possible that a more intensive insulin

regimen, with use of higher doses of insulin and resulting in lower

nadir A1C values, or a longer period of intensive intervention may have

decreased the incidence of relapse in this cohort of patients. Several

published studies have suggested that maintenance of intensive care is

necessary to prevent glycemic relapse in patients with type 1

diabetes.[15-20] In contrast, no previous studies have systematically

addressed the incidence and causes of glycemic relapse in type 2

diabetes. Unlike type 1 diabetes, variation in glycemic control in

insulin-using adults with type 2 diabetes has not been readily explained

by personal characteristics, demographics, family system variables, or

psychosocial characteristics.[21,22]

Glycemic deterioration in type 2 diabetes is at least partly attributed

to progressive decline in beta cell function. The United Kingdom

Prospective Diabetes Study[5] and the Belfast Diet Study[23] established

that ongoing beta cell deterioration led to progressive deterioration of

diabetes control. A recent British study demonstrated that progressive

deterioration in A1C, observed during metformin monotherapy, resumed

within 6 months after a sulfonylurea drug was added to the therapeutic

regimen.[24] Because type 2 diabetes is a disease of progressive

severity, delayed or insufficient intensification of pharmacologic

management is a frequent component underlying failure to achieve

glycemic goals.[25]

The high incidence of relapse in the current study indicates that

patients with type 2 diabetes receiving insulin therapy should receive

higher priority than non-insulin-treated patients for continuation of

intensive management or for inclusion in a relapse prevention program.

Social-learning theory suggests that behavioral relapse in patients with

diabetes is similar to relapse in those with alcoholism, smoking, and

obesity and is determined by an interaction of environmental,

physiologic, cognitive, and affective variables. The relapse prevention

model suggests that relapse can be avoided by self-management training

that includes identification of high-risk situations, training in

problem-solving strategies, and development of cognitive coping

techniques.[2]

This study has important limitations. It is possible that the relatively

high incidence of relapse in our patients is related to bias influenced

by the method of referral, inasmuch as a large proportion of the

insulin-treated patients had proved to be refractory to care provided by

their primary care physicians before referral. In addition, several

potential predictors of relapse, including C-peptide levels, antibodies

to islet cells, frequency of glucose monitoring, physical activity,

adherence to diet, family support, knowledge, attitudes, and other

psychosocial and socioeconomic variables, were not systematically

measured in this study. Further investigations are needed to clarify the

relative contributions of these variables.

Table 1. Characteristics of Study Patients With Type 2 Diabetes, Shown

as Overall Group and Stratified by Use of Insulin*

Factor All patients (N = 393 Using insulin at end of intervention

(N = 126) Not using insulin at end of intervention (N = 267) P value†

comparing use and no use of insulin

Age (yr), mean ± SD 54.8 ± 12.8 53.7 ± 11.6 55.3 ± 13.3 0.18

Sex 0.86

Male, % (no.) 50 (197) 51 (64) 50 (133)

Female, % (no.) 50 (196) 49 (62) 50 (134)

Race 0.66

Black, % 25 24 26

Nonblack, % 75 76 74

Body mass index before intervention

(kg/m2), mean ± SD 35.0 ± 8.2 35.7 ± 8.5 34.6 ± 8.0 0.21

Duration of diabetes (yr), median (IQR) 1.1 (0.1, 5.7) 5.3 (1.9,

10.0) 0.3 (0.1, 3.6) <0.001

Weight change during intervention

(lb), mean ± SD -2.4 ± 9.9 +2.4 ± 11.2 -4.8 ± 8.1 <0.001

History of hypoglycemia before

intervention 0.18

Yes, % (no.) 6 (14) 9 (7) 5 (7)

No, % (no.) 94 (215) 91 (70) 95 (145)

A1C at nadir (%), mean ± SD 6.3 ± 0.5 6.4 ± 0.5 6.3 ± 0.5 0.02

Residence 0.06

Nashville, % (no.) 39 (154) 33 (41) 43 (113)

Non-Nashville, % (no.) 61 (236) 67 (84) 57 (152)

*A1C = hemoglobin A1c; IQR = interquartile range.

†P values were obtained with use of Wilcoxon rank sum tests or

chi-square tests.

References

1. Merriam-Webster's Collegiate Dictionary. 11th ed. Springfield,

MA: Merriam-Webster, Incorporated, 2004: 1050.

2. Marlatt GA, Gordon JR, eds. Relapse Prevention: Maintenance

Strategies in the Treatment of Addictive Behaviors. New York, NY:

Guilford Press, 1985.

3. Sackeim HA, Haskett RF, Mulsant BH, et al. Continuation

pharmacotherapy in the prevention of relapse following electroconvulsive

therapy: a randomized controlled trial. JAMA. 2001;285:1299-1307.

4. Diabetes Control and Complications Trial Research Group. The

effect of intensive treatment of diabetes on the development and

progression of long-term complications in insulin-dependent diabetes

mellitus. N Engl J Med. 1993;329:977-986.

5. UK Prospective Diabetes Study (UKPDS) Group. Intensive

blood-glucose control with sulphonylureas or insulin compared with

conventional treatment and risk of complications in patients with type 2

diabetes (UKPDS 33) [erratum in Lancet. 1999;354:602]. Lancet.

1998;352:837-853.

6. Ohkubo Y, Kishikawa H, Araki E, et al. Intensive insulin therapy

prevents the progression of diabetic microvascular complications in

Japanese patients with non-insulin-dependent diabetes mellitus: a

randomized prospective 6-year study. Diabetes Res Clin Pract. 1995;

28:103-117.

7. American Diabetes Association. Standards of medical care in

diabetes. Diabetes Care. 2004;27(Suppl 1):S15-S35.

8. Brown AW, Wolff KL, Elasy TA, Graber AL. The role of advanced

practice nurses in a shared care diabetes practice model. Diabetes Educ.

2001;27:492-496, 498-500,502.

9. Graber AL, Elasy TA, Quinn D, Wolff K, Brown A. Improving

glycemic control in adults with diabetes mellitus: shared responsibility

in primary care practices. South Med J. 2002;95:684-690.

10. Quinn DC, Graber AL, Elasy TA, J, Wolff K, Brown A.

Overcoming turf battles: developing a pragmatic, collaborative model to

improve glycemic control in patients with diabetes. Jt Comm J Qual

Improv. 2001;27: 255-264.

11. DeFronzo RA. Pharmacologic therapy for type 2 diabetes mellitus.

Ann Intern Med. 1999;131:281-303.

12. RM, Funnell MM, PM, Arnold MS, Fitzgerald JT,

Feste CC. Patient empowerment: results of a randomized controlled trial.

Diabetes Care. 1995;18:943-949.

13. Harrell FE. Regression Modeling Strategies With Applications to

Linear Models, Logistic Regression, and Survival Analysis. New York, NY:

Springer-Verlag, 2001.

14. Elasy TA, Graber AL, Wolff K, Brown A, Shintani A. Glycemic

relapse after an intensive outpatient intervention for type 2 diabetes

[letter]. Diabetes Care. 2003;26:1645-1646.

15. Diabetes Control and Complications Trial/ Epidemiology of

Diabetes Interventions and Complications Research Group. Retinopathy and

nephropathy in patients with type 1 diabetes four years after a trial of

intensive therapy [erratum in N Engl J Med. 2000;342:1376]. N Engl J

Med. 2000;342:381-389.

16. Lawson ML, MR, Fry MK, Perlman K, Sochett EB, Daneman D.

Intensive diabetes management in adolescents with type 1 diabetes: the

importance of intensive follow-up. J Pediatr Endocrinol Metab.

2000;13:79-84.

17. Couper JJ, J, Fotheringham MJ, Sawyer M. Failure to

maintain the benefits of home-based intervention in adolescents with

poorly controlled type 1 diabetes. Diabetes Care. 1999;22:1933-1937.

18. Rosilio M, Cotton JB, Wieliczko MC, et al (French Pediatric

Diabetes Group). Factors associated with glycemic control: a

cross-sectional nationwide study in 2,579 French children with type 1

diabetes. Diabetes Care. 1998;21:1146-1153.

19. Spiess K, Sachs G, Moser G, Pietschmann P, Schernthaner G, Prager

R. Psychological moderator variables and metabolic control in recent

onset type 1 diabetic patients—a two year longitudinal study. J

Psychosom Res. 1994;38:249-258.

20. Bott U, ns V, Grusser M, Bender R, Mulhauser I, Berger M.

Predictors of glycaemic control in type 1 diabetic patients after

participation in an intensified treatment and teaching programme. Diabet

Med. 1994;11:362-371.

21. Nichols GA, Hillier TA, Javor K, Brown JB. Predictors of glycemic

control in insulin-using adults with type 2 diabetes. Diabetes Care.

2000;23:273-277.

22. Trief PM, Grand W, Elbert K, Weinstock RS. Family environment,

glycemic control, and the psychosocial adaptation of adults with

diabetes. Diabetes Care. 1998;21:241-245.

23. Levy J, Atkinson AB, Bell PM, McCance DR, Hadden DR. Beta-cell

deterioration determines the onset and rate of progression of secondary

dietary failure in type 2 diabetes mellitus: the 10-year follow-up of

the Belfast Diet Study. Diabet Med. 1998;15:290-296.

24. Cook MN, Girman CJ, Stein PP, CM, Holman RR. Glycemic

control continues to deteriorate after sulfonylureas are added to

metformin among patients with type 2 diabetes. Diabetes Care.

2005;28:995-1000.

25. Cook CB, Lyles RH, El-Kebbi I, et al. The potentially poor

response to outpatient diabetes care in urban African-Americans.

Diabetes Care. 2001;24:209-215.

Funding Information

This study was supported in part by the National Institute of Diabetes

and Digestive and Kidney Diseases (Grant P60DK20593), the American

Diabetes Association, and the Wood Foundation.

Abbreviation Notes

A1C = hemoglobin A1c; BMI = body mass index; CI = confidence interval;

HR = hazard ratio; IRQ = interquartile range; NA = not available

Reprint Address

Dr. Alan L. Graber, 2558 TVC, Vanderbilt Medical Center, Nashville, TN

37232.

Alan L. Graber, MD, FACE, FACP,1 Ayumi K. Shintani, PhD, MPH,2 Kathleen

Wolff, RN, MSN,3 Anne Brown, RN, MSN,3 Tom A. Elasy, MD, MPH1

1Department of Medicine, Vanderbilt University School of Medicine

2Department of Biostatistics, Vanderbilt University School of Medicine

3Vanderbilt University School of Nursing, Nashville, Tennessee.

--

ne Holden, MS, RD < fivestar@... >

" Ask the Parkinson Dietitian " http://www.parkinson.org/

" Eat well, stay well with Parkinson's disease "

" Parkinson's disease: Guidelines for Medical Nutrition Therapy "

http://www.nutritionucanlivewith.com/