Research: Epidemiology Diet quality is prospectively associated with incident impaired fasting glucose in older adults B. Gopinath 1 , E. Rochtchina 1 , V. M. Flood 2 and P. Mitchell 1 1 Department of Ophthalmology and Westmead Millennium Institute, Centre for Vision Research, University of Sydney and 2 Faculty of Health and Behavioural Sciences, University of Wollongong, Sydney, NSW, Australia Accepted 18 December 2012 Abstract Aims Dietary modifications may play an important role in the prevention of diabetes. We aimed to assess the temporal association between diet quality and both impaired fasting glucose and Type 2 diabetes among older adults. Methods A total of 2564 participants aged 49+ years at baseline were examined between 1992 and 1994 and 2002– 2004 and had their fasting blood glucose measured. Dietary data were collected using a semi-quantitative food frequency questionnaire. A modified version of the Healthy Eating Index for Australians was developed to determine Total Diet Score. Incident diabetes (or impaired fasting glucose) was defined in participants at risk who were newly diagnosed by a physician during the follow-up or found to have a fasting blood glucose level  7.0 mmol/l (or 6.1–6.9 mmol/l). Results After adjusting for age, sex, current smoking, body mass index, hypertension and serum triglycerides, comparing highest with lowest tertile of total diet score, a significant 75% decrease in risk of incident impaired fasting glucose was observed in men (P trend = 0.02). Also, in men, each two-point increase in Total Diet Score was associated with a 52% reduction in the 10-year incidence of impaired fasting glucose, (OR 0.48, 95% CI 0.33–0.69). No significant associations were observed among women or with the 10-year incidence of diabetes. Conclusions Greater compliance with published dietary guidelines (better diet quality) was associated with a reduced risk of pre-diabetes in men, but not women. Diabet. Med. 30, 557–562 (2013) Introduction Type 2 diabetes is a growing health problem, with nearly 400 million people estimated to suffer from this condition worldwide by 2030 [1]. Prevention of diabetes is therefore, crucial to lowering incidence of disease and minimizing the individual, familial and public health burden [2]. A healthy lifestyle, including diet and physical exercise, is critical to prevent diabetes in the population and delay its onset in susceptible individuals [3]. However, further research is warranted to identify optimal eating patterns for the prevention of Type 2 diabetes [4]. The study of complete dietary patterns represents the most adequate approach to assess the role of diet on the risk of diabetes – more so than examining isolated foods or nutrients [3]. The rationale for this concept is that synergistic or antagonistic effects may exist between the different compo- nents of a food pattern [3,5]. Only a few studies have assessed the relationship between diet quality scores and risk of Type 2 diabetes. A US cohort study of 80 029 women aged 38–63 years showed that higher diet quality scores were associated with a 26% reduced risk of Type 2 diabetes [6]. In addition, a recent prospective US study of 41 615 men showed that several diet quality scores were associated with a lower risk of Type 2 diabetes; a 1 SD increase in diet quality scores was associated with a 9–13% reduced risk of Type 2 diabetes (P < 0.01). Moreover, a Japanese cross-sectional study of men showed that higher diet quality was associated with a reduced likelihood of impaired fasting glucose (P trend = 0.02) [7]. These data suggest a common underlying dietary pattern [8]. However, the findings from these studies may not be generalizable to other populations as they assessed men and women separately. Further studies are needed to clarify the prospective relationship between diet quality and risk of Type 2 diabetes among a representative population of both men and women. We previously developed a tool [9] modelled using both Australian and US diet quality indices [10,11]. In the present study, we used a relatively large cohort of older adults aged 49+ years to examine the temporal association between diet Correspondence to: Paul Mitchell. E-mail: paul_mitchell@sydney.edu.au ª 2013 The Authors. Diabetic Medicine ª 2013 Diabetes UK 557 DIABETICMedicine DOI: 10.1111/dme.12109