Dietary treatment of patients with diabetes mellitus gestationel
English version- and original Danish version attached subsequently
by
Finn Friis Lauszus
Centerof Diabetes and Pregnancy,
Gynecology / Obstetrics dept. Y,
AarhusUniversityHospital .
Ph.D. dissertation,
The Faculty of Health, AarhusUniversity, 1996.
The PhD dissertation is the result of three years of workwith pregnant women who developed gestational diabetes. Without their goodwill to enter the trial and the patience to take on the extra hassle, I would have not been able to succeedwith my studies. All pregnant women were examined in outpatient clinic in the Department Y, AarhusMunicipalHospitaland performed during my employment as a resident doctor and later as a clinical assistant at the Department of Experimental Clinical Research.
I am Joachim G. Klebe, assistant professor and senior consultant at Dept Y, deeply grateful for his unfailing interest in my project. His enthusiasm as my supervisor has been a great support throughout the process, which he instead of corrections rather came up with great ideas. His mood and energy was contagious.
In addition to Dr. Klebe I owe thanks to my friend, the first registrar, Ph.D., Dr.Med.Scie., Ole Winther Rasmussen, AarhusUniversityHospital, for professional guidance in the endocrine and dietary field. He envisioned the idea for the project and has followed it during implementation and processing, as his knowledge and ideas have been vital to the project.
I am also grateful for the support I have received from Professor Hans Ørskov, Department of Experimental Clinical Research, and senior consultant Kjeld Hermansen, Medical Section C, AarhusCountyHospital.
Dietician Lillian Jensen, Department Y, AarhusUniversityHospital, I owe a great deal of credit in encouraging pregnant women to continue with the diet and record their dietary habits.
Research Laboratory Eva Sejr, Medical Research Lab, has made a great effort to keep track of my many measurements.
Associate Professor Kim Overvad, Department of Epidemiology and Social Medicine, AarhusUniversity and senior consultant, Eivind Thorling, Department of Occupational and Environmental Medicine, AarhusUniversity, helped me with the measurement of fatty acids.
Laboratory Manager Leif Bandholm, Aarhus Oil Factory, is thanked for friendly support and good advice from start to finish the project.
The staff at the clinic, the laboratory, and the maternity ward is thanked for the attention it showed every time one of "my" pregnant appeared.
Finally, thanks to my wife, Karen, for moral, logistic, and technical support.
The project received support and grants from:
Aarhus Oliefabrik A / S (now AarhusKarlshamn AB, in short AAK)
Mimi and Victor Larsen Foundation
Nordisk Insulin Foundation
Novo Nordisk Foundation
Department of Experimental Clinical Research
Danish Medical Research Council's Consultancy Service
For reference:
1. Lauszus FF, Klebe JG, Rasmussen OW, Grøn PL, Lauszus KS. Forandringer i glukosestofskiftet ved gestationel diabetes mellitus under og efter graviditet. Ugeskr f Læger 1997; 159: 4631-36. PMID: 9245038
2. Lauszus FF, Rasmussen OW, Jensen L, Lauszus KS, Hermansen K, Klebe JG. Effect of a high-monounsaturated fatty acid diet on blood pressure and glucose metabolism in women with gestational diabetes. Eur J Clin Nutr 2001; 55: 436-443. doi:10.1038/sj.ejcn.1601193
Abbreviations
BMI: Body Mass Index
BMR: Basal Metabolic Rate
CHO: Carbohydrate
E%: Energy percentage
EFA: Essential fatty acids
GDM: Gestationel diabetes mellitus
HbA1c: Hemoglobin A1c
HDL: High Density Lipoprotein
HOMA: Homeostasis model assessment
IVGTT: Intravenous glucosetolerance test
LDL: Low Density Lipoprotein
MUFA: Mono-unsaturated fatty acids
NIDDM: non-insulin-dependent diabetes mellitus (type 2 d.m.)
OGTT: Oral glukosetolerance test
PUFA: Polyunsaturated fatty acids
SD: Standard deviation
SEM: Standard error of mean
SFA: Saturated fatty acids
TG: Triglyceride
VLDL: Very Low Density Lipoprotein
INTRODUCTION
Definition
Gestational diabetes mellitus (GDM) means glucose intolerance of varying severity with onset or first recognition during pregnancy (1.42). The diagnosis is verified by an oral glucose tolerance test (OGTT). However, there is not a worldwide consensus on the criteria of diagnosis. There are different gestational when to perform the test, the selection of the pregnant women who should be tested and which glucose values that determine whether the pregnant woman has GDM.
Etiology of GDM
The pregnancy induces facilitated anabolism during meals and catabolism during fasting periods, presumably to ensure glucose, fat, and protein for fetal growth (6). Therefore, pregnant women are characterized by a reduced glucosetolerance and increased lipolysis compared to non-pregnant women. Only about 2% of pregnant women in Denmark meets the Danish diagnostic criteria for GDM (34,35). This is because the majority of pregnant women are able to increase their insulin secretion sufficiently to compensate for the reduced insulin sensitivity, thus ensuring normal glucose tolerance. GDM is predominantly seen in obese women, asobesity itself reduces insulin sensitivity, thereby increasing the risk to develop GDM.
Autoantibodies against beta-cells probably play no role in GDM (8,35,36,52). Insulin receptor function and insulin-stimulated glucosetransport in muscle tissue are similar in women with GDM and normal pregnancy (2,15,23). In contrast, insulin secretion is different in the two situations. The glucose-stimulated insulin response is impaired and pro-insulin / insulin ratio increased in GDM women (5,6,54). Thus, it seemsthat insulin resistance and impaired beta-cell function could be present at the same time in GDM.
Compared to normal pregnant women, GDM is accompanied by elevated levels of triglyceride (TG), cholesterol, free fatty acids and beta-hydroxy butyrate. Post partum, the women with GDM have normalized the lipid levels (40). In particular, the free fatty acids and beta-hydroxybutyrateare suspected to exert harmful effects on the fetus: A study showed that the child's later mental development was found to have a negative association to the level of the two metabolites (59). The lipoproteins, HDL, LDL, and VLDL, are only modified to a lesser extent. GDM is associated with an atherogenic lipid profile and other conditions with insulin resistance (2,15,51).
Treatment with diet
GDM is treated primarily with diet. Recommendations follow the usual recommendations for patients with diabetes mellitus. The main goal is a change of meal frequency to frequent, small meals. This prevents formation of ketone bodies seen in periods of fasting. Further,a diet rich in carbohydrate and low on fat is recommended (53). The goal is anintake of calories, which ensures a normal weight gain in pregnancyand improves glucoseregulation.
The advantages of a high-carbohydrate, low fat diet, however, were disputed, as it may result in hypertriglyceridemiacomparedwith a low-carbohydrate, high-fat diet in non-insulin-dependent diabetes mellitus (NIDDM=type 2 diabetes) (21,22,45,50,58). Alternatively thecarbohydrate content may be kept unchanged and saturated fat replaced with monounsaturated or polyunsaturated fat (13,19,46,50). A high-fat, carbohydrate diet with high content of monounsaturated fatty acids (MUFA) appears to improve insulin sensitivity and glucose tolerance and reduce diurnal blood pressurein NIDDM patients compared with the recommended diabetesdiet. Furthermore, there appears to be negative effects on lipid metabolism by a high-fat, low carbohydrate diet rich on MUFA compared to the diabetesdiet (4,7,20,50,56). Epidemiological studies from the Mediterranean area show a lower incidence of cardiovascular disorders, and a lower blood pressure in areas where people consume large amounts of MUFA as olive oil (32,61).
Thus, these studies question whether the diabetic diet is the optimum, which today is used to counsel women with GDM. One might assume that the apparently beneficial effects of MUFA-rich diet in NIDDM patients could be transferred to pregnant women with GDM, where insulin resistance also plays a central role in the pathogenesis.
AIM AND HYPOTHESIS
In pregnant women with GDM a MUFA diet should:
a) improve insulin sensitivity,
b) have a beneficial effect on the lipid profile,
c) reduce blood pressure or keep it unchanged
d) reduce the tendency to increased birth weight,
- Because we expected a development of GDM with the trend toward higher lipid and glucose values and increase in blood pressure in the third trimester.
The design of the study was that the dose of MUFA -and thus fat content should be appropriate high to achieve effect of the relatively short time-span. Oil intake was monitored by diet records and analyzing the fatty acid composition of lipids. The diet with high-fat content should be appetizing and consumed in 6 weeks and cover the need for nutrients and energy during pregnancy.
SCREENING
Screening for GDM is performed at first pregnancy control in 17th- 19th gestational week in theGynecology / Obstetrics Department at AarhusUniversityHospital. The criteria are if the pregnant woman has a history of increased risk of diabetes. In the risk assessment is included pregravid obesity, familial disposition for diabetes, previous childbirth with birth weight> = 4500g, former fetus mortuus of unknown origin, age> 37 and current glucosuria. If the pregnant woman meets one of the above criteria, fastingcapillary glucoseis measured on two different days. Are both fastingcapillary glucose> 4,7mmol / L the pregnant woman is referred to anoral glucose tolerance test (OGTT)with 75g glucose. If the OGTTis normal in week 17-19, arepeat OGTT is performed in 28 30 weeks. If onlyone or no fastingcapillary glucose is found above 4.7mmol / lin week 17-19, repeated measurement of 2 fastingcapillary glucose is performed in 30 weeks. If both values are > 4.7 mmol/lan OGTT is performed. Is this OGTT normal, no further diabetestesting is performed.
On the Gynecologicy / Obstetrics Department at Aarhus University Hospital the following cut-of values were used at the OGTT: fasting 6.4mmol / l, at 30 minutes 13.6mmol / l, at 60 minutes 13.7mmol / l, at 90 minutes 11.0mmol / l, at 120 minutes 10.2mmol / l, at 150 minutes 9.7 mmol / l and at 180 minutes 8.5mmol / l. This corresponds to the average + 3 SD of capillary glucose values. GDM is diagnosed if at least two values are above the cut-off values. This is a modified version of the original one which was measured in venous plasma from a Danish norm made from the results of 40 normal weight, non-pregnant women without a diabetes family history (14).
TREATMENT OF GDM
Diet
Diet constitutes a cornerstone in the treatment of GDM. An increase in meal frequency and fixed caloric intake are important components of the diet. Furthermore, a shift in food energy distribution is recommended to 55% carbohydrate, maximum 30% fat, of which more than 10% as saturated fat, and 15 to 20% protein (53). The fibre content of the food should be at 3g per MJ. Essential fatty acids (EFA) should in pregnant women constitute at least 4.5% energy of the diet. Linolenic acid and its daughter metabolites (n-3 fatty acids) are recommended to contribute at least 0.5% of energy (44).
Insulin
Insulin is used in the treatment of GDM if diet therapy does not lead to satisfactory glucosevalues. At our Department following values are used to decide when o start insulin: Fasting at 6.5mmol / l and day profile with the average of 7.5mmol / l.
Obstetricalfactors
Blood pressure in pregnancy
Blood pressure drops initially during pregnancy and increases from about 33rd gestational week back to pre-pregnancy level (26). In GDM an increased incidence of hypertension and preeclampsia is observed (43.63). As hypertension and glucose intolerance is associated, the insulin resistance and secondary hyperinsulinemia itself may induce hypertension (15,30,47,57).
Fetal implications
Previously GDM was looked upon as a disorder between insulin-dependent diabetes mellitus (IDDM) and the general population; in particular,in matters of incidence of fetus mortuus, macrosomia, and malformations. Since the incidence of perinatal complications in general is reduced to a few percent, it has been increasingly difficult to find earlier marked differences between GDM and the background population. It is uncertain whether GDM today has more pregnancy complications. Part of the explanation may be that the intensive treatment erases the differences in morbidity.
The study population
During the period December 1, 1991 to May 30, 1993, 27 pregnant women with GDM were randomized to enter into two different diet groups. The project was approved by the Ethical Committee (reg.no.1991 / 2089). Inclusion criteria were an abnormal OGTT before the 34th week of pregnancy. Exclusion criteria were medical treatment throughout the trial, especially antidiabetic, cholesterol-lowering or antihypertensive medication. Nine other pregnant women with GDM met the inclusion criteria during the same period. Of these, 3 women did not want to participate, three were Turkish women (language barrier), 1 had endocrine and exocrine pancreatic insuffiency, one was in antihypertensive treatment and one did not attend the controls.
Pregnant women were divided into obese and normal weight and randomized separately. Overweight is defined as 20% above ideal weight. Ideal weight is given by the ideal weight = (height in cm minus 100) times 0.9. Overweight is when the women weighed ideal weighttimes 1.2. One group was assigned to a MUFA diet with an energy distribution of 46% carbohydrate, 37% fat (22% as MUFA) and 16% protein. A control group with GDM was given a diet in accordance with current recommendations (CHO) group.
All the women were instructed according to current guidelines of diabetes diet when diagnosed with an abnormal OGTT at 17th to 33rdweek. The pregnant woman registered her diet in diary before week 33 and once in the 36th/ 37thweek, each lasted three weekdays and one weekend day. In the 33rdweek, the women were randomized. All were primarily put on a calorie intake of 8000kJ. Dietary records were afterwards scrutinized and worked over together with a dietician, after which an individual diet could be planned. At weekly consultations with the dietician the women were encouraged to adhere to their allocated diet. Calorie intake was adjusted depending on the pregnant woman's appetite or lack of weight increase. The women in the MUFA group had baked rolls with nuts given to them with a high MUFA content (8.2g MUFA per roll) to substitute for ordinary bread (0.54g MUFA per 100g). Moreover, they received vegetable oil with high MUFA content (80%) and hazelnuts and almonds for snacks, where it was planned in the diet plan by the dietician. Compliance was assessed by the needed number of rolls and oil from week to week.
On the first test day at 33rd week an intravenous glucose tolerance test (IVGTT) was performed, ambulatory blood pressure measured, fasting blood samples taken and a container for days urine collection given. These measurements were repeated in the 36thand 38th week. The MUFA and CHO group were similar at start of trial (Table 1). There was a trend toward higher weight and glucose values in the MUFA group. Of 27 women with GDM 5 had normal weight and 8 weighed over100kg. The MUFA group included after randomization five women over 100kg and 2 with normal weight.
Table 1 Clinical data of eligible GDM women
------
GroupMUFACHONon-trialp-value
------
Before the trial:
No.13149
age31.429.532.70.87
weight 33rd wk(kg)97.7±6.988.2±4.990.6±5.10.26
BMI 33rd wk (kg/m2)35.3±2.432.2±1.5 31.7±20.27
Auscultatory blood pressure
systolic116±5#115±3#130±50.84
diastolic 77±4# 75±1# 80±40.54
(mmHg)
HbA1c(%)5.6±0.25.3±0.16.1(n=3) 0.06
fasting glucose 5.1±0.3*4.7±0.2*5.4±0.1x 0.28
(mmol/l)
post prandial glucose 7.1±0.5x6.3±0.4x7.3±0.5x0.49
(mmol/l)
fasting insulin 22.7±3.928.2±4.5-0.72
(μU/ml)
creatinine clearance124±7138±11118±19 (n=3)0.54
(ml/min)
Albumin excretion
(mg/24-hour) 10±3 13±4 9±6 (n=3)0.25
During the trial:
weight (kg)
36th wk100.1±6.987.5±5.30.16
38th wk100.3±6.688.5±5.40.75
At delivery100.5±6.588.6±5.10.16
post partum 90.5±7.978.0±5.10.18
BMI (kg/m2)
36th wk36.2±2.431.5±1.50.10
38th wk36.0±2.432.2±1.70.66
At delivery36.3±2.232.4±1.70.16
post partum32.6±2.828.5±1.70.20
------
p-values: MUFA versus CHO
#=measured by 24-h monitor
*=full blood. x=capillary
BMI= body mass index = weight in kg/(height in m)2
1/3 of thewomen, who did not want or could not be included in the protocol,were of non-Caucasian ethnicity. They are hence on called the non-trial women. In the MUFAgroup were one Chinese and one Indian woman. In addition, the non-trial women had more fetus mortuus and macrosomia, but had less diabetes family history. This may be due to recallbias, since they have not been specifically interviewed like the included women.
Table 2 shows the OGTT in all groups. The MUFA group is comparable to the CHO group on any capillary glucosevalues. The area of the fasting value is calculated according to the trapezoid method and only considers the positive area (37).
Table 2OGTT results in 36 eligible GDM women
------
GroupMUFACHONon-trialp-value
------
No.13149
Gestational wk28.7±1.228.8±1.027.6±1.5
Range173318331932
Fasting value(mmol/l) 6.0±0.2 5.7±0.3 5.7±0.10.21
30 min10.8±0.412.3±0.5 9.8±0.40.30
60 min13.0±0.512.3±0.511.5±0.50.34
90 min12.7±0.612.4±0.312.6±0.70.69
120 min12.3±0.911.3±0.311.7±0.40.25
150 min9.3±0.49.2±0.39.5±0.20.96
180 min7.6±0.57.4±0.57.3±0.50.80
Area over fasting value
(mmol/l*30 min) 30.1±2.030.0±1.127.6±1.80.98
------
p-value: MUFA versus CHO
All glucose measurements are capillary plasma
Figure 1 shows the average capillary glucose measurements (y-axis: kapillær glucose) at OGTT (0-180 minutes) for the MUFA-() and CHO-group () compared to the diagnostic criteria for OGTT (▲).
Figure 1
METHODS
The intravenous glucose tolerance test(IVGTT) was performed with the administration of 50% glucose solution of 0.3mg / kg body weight. The times of sampling were at-15, 0, 2, 3, 4, 5, 6, 8, 10, 12, 14, 16, 19, 22, 25, 30, 40, 50, 60, 70, 80 , 90, 100, 120, 140, 160 and 180 minutes after glucoseinfusion. The zero value was calculated by the average of minutes -15 and 0.
The first phase insulin secretion was calculated as a positive area of real value from 0 to 10 minutes. The unit is μU/ml * 1min.
K-value was calculated as the slope of the curve by linear regression of logarithmic values between 10 to 40 minutes at the IVGTT. K value reflects the insulin sensitivity and has the unit -10-2 * mmol / l * min-1.
Homeostasis model assessment (HOMA) is the fasting insulin and glucose measures from IVGTT used for the calculation of R. R is an expression of the insulin resistance given by the formula R = insulin / 22.5 * e –ln(glucose) and the unit U / mol (39 ). Insulin in mU/L in serum were considered similar to values in plasma; glucose from venous whole blood was multiplied by 1.16 to give plasmavalues in mmol / l.
Glucose was determined by glucose oxidase method on venous whole blood (10).
Insulin was determined by specific radioimmunoassay in the serum (70). For each time point, the average of 3 measurements of insulin in the same blood sample is given. Intra- and inter-assay variations coefficients were 5 and 10%.