The relationship between glycemic intake and insulin resistance in older women

O'Sullivan, Therese Anne (2008) The relationship between glycemic intake and insulin resistance in older women. PhD thesis, Queensland University of Technology.


Glycemic intake influences the rise in blood glucose concentration following consumption of a carbohydrate containing meal, known as the postprandial glycemic response. The glycemic response is a result of both the type and amount of carbohydrate foods consumed and is commonly measured as the glycemic index (GI) or glycemic load (GL), where the GI is a ranking in comparison to glucose and the GL is an absolute value encompassing both the GI and amount of carbohydrate consumed. Evidence from controlled trials in rat models suggests that glycemic intake has a role in development of insulin resistance, however trials and observational studies of humans have produced conflicting results. As insulin resistance is a precursor to type 2 diabetes mellitus, lifestyle factors that could prevent development of this condition have important public health implications. Previous observational studies have used food frequency questionnaires to assess usual diet, which could have resulted in a lack of precision in assessment of individual serve sizes, and have been limited to daily measures of glycemic intake. Daily measures do not take fluctuations in glycemic intake on a per meal basis into account, which may be a more relevant measure for investigation in relation to disease outcomes. This PhD research was conducted in a group of Brisbane women aged 42 to 81 years participating in the multidisciplinary Brisbane Longitudinal Assessment of Ageing in Women (LAW study). Older women may be at particular risk of insulin resistance due to age, hormonal changes, and increases in abdominal obesity associated with menopause, and the LAW study provided an ideal opportunity to study the relationship between diet and insulin resistance. Using the diet history tool, we aimed to assess the glycemic intake of the population and hypothesised that daily GI and daily GL would be significantly positively associated with increased odds of insulin resistant status. We also hypothesised that a new glycemic measure representing peaks in GL at different meals would be a stronger predictor of insulin resistant status than daily measures, and that a specially designed questionnaire would be an accurate and repeatable dietary tool for assessment of glycemic intake. To address these hypotheses, we conducted a series of studies. To assess glycemic intake, information on usual diet was obtained by detailed diet history interview and analysed using Foodworks and the Australian Food and Nutrient (AUSNUT) database, combined with a customised GI database. Mean ± SD intakes were 55.6 ± 4.4% for daily GI and 115 ± 25 for daily GL (n=470), with intake higher amoung younger participants. Bread was the largest contributor to intakes of daily GI and GL (17.1% and 20.8%, respectively), followed by fruit (15.5% and 14.2%, respectively). To determine whether daily GI and GL were significantly associated with insulin resistance, the homeostasis model assessment of insulin resistance (HOMA) was used to assess insulin resistant status. Daily GL was significantly higher in subjects who were insulin resistant compared to those who were not (134 ± 33 versus 114 ± 24 respectively, P<0.001) (n=329); the odds of subjects in the highest tertile of GL intake being insulin resistant were 12.7 times higher when compared with the lowest tertile of GL (95% CI 1.6-100.1, P=0.02). Daily GI was not significantly different in subjects who were insulin resistant compared to those who were not (56.0 ± 3.3% versus 55.7 ± 4.5%, P=0.69). To evaluate whether a new glycemic measure representing fluctuations in daily glycemic intake would be a stronger predictor of insulin resistant status than other glycemic intake measures, the GL peak score was developed to express in a single value the magnitude of GL peaks during an average day. Although a significant relationship was seen between insulin resistant status and GL peak score (Nagelkerke’s R2=0.568, P=0.039), other glycemic intake measures of daily GL (R2=0.671, P<0.001) and daily GL per megajoule (R2=0.674, P<0.001) were stronger predictors of insulin resistant status. To develop an accurate and repeatable self-administered tool for assessment of glycemic intake, two sub-samples of women (n=44 for the validation study and n=52 for the reproducibility study) completed a semi-quantitative questionnaire that contained 23 food groupings selected to include the top 100 carbohydrate foods consumed by the study population. While there were significant correlations between the glycemic intake questionnaire and the diet history for GL (r=0.54, P<0.01), carbohydrate (r=0.57, P<0.01) and GI (r=0.40, P<0.01), Bland-Altman plots showed an unacceptable difference between individual intakes in 34% of subjects for daily GL and carbohydrate, and 41% for daily GI. Reproducibility results showed significant correlations for daily GL (r=0.73, P<0.001), carbohydrate (r=0.76, P<0.001) and daily GI (r=0.64, P<0.001), but an unacceptable difference between individual intakes in 25% of subjects for daily GL and carbohydrate, and 27% for daily GI. In summary, our findings show that a significant association was observed between daily glycemic load and insulin resistant status in a group of older women, using a diet history interview to obtain precise estimation of individual carbohydrate intake. Both the type and quantity of carbohydrate are important to consider when investigating relationships between diet and insulin resistance, although our results suggest the association is more closely related to overall daily glycemic intake than individual meal intake variations. A dietary tool that permits precise estimation of carbohydrate intake is essential when evaluating possible associations between glycemic intake and individual risk of chronic diseases such as insulin resistance. Our results also suggest that studies using questionnaires to estimate glycemic intake should state degree of agreement as well as correlation coefficients when evaluating validity, as imprecise estimates of carbohydrate at an individual level may have contributed to the conflicting findings reported in previous studies.

Impact and interest:

Citation counts are sourced monthly from Scopus and Web of Science® citation databases.

These databases contain citations from different subsets of available publications and different time periods and thus the citation count from each is usually different. Some works are not in either database and no count is displayed. Scopus includes citations from articles published in 1996 onwards, and Web of Science® generally from 1980 onwards.

Citations counts from the Google Scholar™ indexing service can be viewed at the linked Google Scholar™ search.

Full-text downloads:

1,601 since deposited on 12 Feb 2009
102 in the past twelve months

Full-text downloads displays the total number of times this work’s files (e.g., a PDF) have been downloaded from QUT ePrints as well as the number of downloads in the previous 365 days. The count includes downloads for all files if a work has more than one.

ID Code: 17814
Item Type: QUT Thesis (PhD)
Supervisor: Lyons-Wall, Philippa & Bremner, Alexandra
Keywords: Australia, carbohydrate, diabetes, diet history, dietary intake, food frequency questionnaire, glycemic (or glycaemic) index, glycemic intake, glycemic load, hyperglycemic peak, insulin resistance, insulin sensitivity, postmenopausal women
Divisions: Current > QUT Faculties and Divisions > Faculty of Health
Current > Schools > School of Public Health & Social Work
Institution: Queensland University of Technology
Deposited On: 12 Feb 2009 06:01
Last Modified: 28 Oct 2011 19:52

Export: EndNote | Dublin Core | BibTeX

Repository Staff Only: item control page