Degree Name

Doctor of Philosophy


University of Wollongong. Dept. of Public Health and Nutrition


Type 1 diabetes mellitus is a chronic metabolic disorder affecting both young and old. Treatment goals are becoming standardised, but the degree towhich patients with type 1 diabetes meet their goals in practice is uncertain.

The aim of this study was to evaluate the nutrition-related outcomes ofpatients with type 1 diabetes diagnosed since 1984 in the Illawana and at least oneyear post-diagnosis. Subjects were recruited from the list of patients registered atthe Diabetes Education Unit (DEU) of the Illawarra Area Health Service.

All patients were aged between 18-40 years. For the purpose ofcomparison a similar nondiabetic group who were either relatives/friends of patients or volunteers was recruited. In these two groups, anthropometric measurements,blood pressure, glycosylated hemoglobin (HbAlc%), lipids, urine albumin:creatinine ratio, dietary intake as well as quality of life were assessed.

Dietary intake of subjects was assessed by a meal-based diet history (Burkemethod), and food pattern and food preparation questionnaires (based on validated DCCT questionnaires). The results were analysed using Diet 1 (V4), which uses the Australian NUTTAB 1995 food composition database. The findings were compared to the Australian and American dietary goals and recommendations.

In total 55 patients (18 female and 37 male) and 47 controls (17 female and 30 males) were seen. The means of body measurements including body massindex (BMI), waist to hip ratio ( WHR ) and body fat%, blood pressure and blood lipids and lipoproteins were within normal range and not significantly different between the two groups. However blood pressure was higher in overweight subjects in both groups. Also obesity or overweight was associated with an adverse lipid and lipoprotein pattern. Plasma fibrinogen concentration in peoplewith type 1 diabetes was significantly higher than in control subjects. Higher plasma fibrinogen levels were associated with higher B M I and elevated plasmatriglyceride and decreased HDL cholesterol concentrations.

The albumin: creatinine ratios in spot and early morning urine samples of patients with type 1 diabetes were significantly higher than these ratios in the control group. The negative relationship of early morning urine sample with HDL cholesterol was noted.

In this study the degree of glycemic control was evaluated by HbAlc%. Mean values were 8.52% (SD 2.21) and 8.56% (SD 1.91) for females and males respectively, higher than in the DCCT intensively treated patients and lower than in the DCCT conventionally treated patients. A number of variables including anthropometric measurements, blood pressure, diabetes history, blood and urine results were evaluated as independent variables in several models with HbAlc% in the diabetic group. No statistically significant associations were found. It was shown however that those with the least education had the poorest diabetic control but this also was not found to be significant.

The dietary intake results indicated that only starch intake (in males andfemales) and dietary fibre intake (in males) were higher among patients than in controls when the macronutrient and alcohol intakes were compared in two groups. None of the same sex comparisons in diabetic and control subjects showed asignificant difference in regard to the energy intake and the contribution of macronutrients. No associations were found between dietary intakes and glycemiccontrol in diabetic subjects.

The findings were compared to the recommended dietary intakes (RDI). Significantly more subjects with diabetes than controls reached the guidelines for fat consumption, however the majority of those with diabetes consumed saturated fatty acids at a level greater than the recommendations. Protein intake of patients with type 1 diabetes in terms of grams per body weight and its contribution to total energy intake was higher than recommended. The average carbohydrate (energy%) was similar in both groups and it was lower than the recommendations. Their micronutrient intake for both diabetic and control subjects were higher than the recommendations. Sodium intake in diabetic subjects was higher than in controls.

The overall diet quality of the subjects was scored on the basis of several dietary recommendations. It was found that the higher the diet score (ie the moreclosely the diet adhered to the recommendations) the better the glycemic control but this association was not significant. Educated people in both groups had higher diet scores. Also in regard to consumption of basic food groups, it was found that diabetic subjects had more fruit than controls did. Both groups did not have legumes and nuts (low glycemic index foods) in their diets regularly.

Not all the patients complied with their dietary prescriptions. Those with better dietary adherence had better glycemic control, although the correlation did not reach significance level.

Quality of life of patients and controls was evaluated with the DiabetesQuality of Life Measure (applied in the DCCT ) and SF-36 (a tool which measures non-disease-specific aspects of quality of life). Our results were comparable with those from the DCCT , showing neither group of patients had severe complications. Diabetes worry subscale on the Diabetes Quality of Life Measure was associated negatively with the duration of diabetes and HbAlc%. The result of SF-36subscales were similar to controls, in both groups these subscales were inverselyrelated to either BMI and/or WHR.

The results of this study highlights the possibility of improving diabetic control and as a result preventing the related complications in this group of patients by nutritional management. However caution should be taken in generalizing from these results, given the relatively small sample size. An up to date frequency counting of this disease is recommended. Then larger scale studies that involve the majority of the patients in this area and/or nutrition intervention programs (like nutrition education) are needed to confirm the generalisability of this study.