Roberto Bizzotto (1), Chris Jennison (2), Angus Jones (3,4), Mark Walker (5), Ewan R Pearson (6), Andrea Mari (1) for the IMI DIRECT Consortium
(1) Institute of Neuroscience, National Research Council, Padova, Italy; (2) Department of Mathematical Sciences, University of Bath, Bath, UK; (3) National Institute for Health Research Exeter Clinical Research Facility, University of Exeter Medical School, Exeter, UK; (4) Royal Devon and Exeter NHS Foundation Trust, Exeter, UK; (5) Institute of Cellular Medicine, Newcastle University, Newcastle, UK; (6) Molecular and Clinical Medicine, University of Dundee, Dundee, UK.
Objectives: Many studies have described clinical parameters associated with type 2 diabetes (T2D) onset or “time-to-failure” markers, while parameters associated with progression rate in established and newly-diagnosed T2D are unclear. Here, we have evaluated these associations considering several metabolic parameters from the prospective DIRECT study.
Methods: We studied 732 white T2D patients from northern Europe recruited to the DIRECT study (baseline treatment as diet only or metformin, HbA1c <60mmol/mol, diabetes duration 1c concentrations were collected at months 0, 9, 18, 27 and 36 after start of the study. Mixed-meal tests were performed at 0, 18 and 36 months, for assessment of peripheral insulin sensitivity (as OGIS index), oral insulin clearance (CLIo), β-cell glucose sensitivity (GS) and β-cell rate sensitivity (RS). β-cell function was evaluated via mathematical modelling [3]. We computed the orthogonal longitudinal and cross-sectional components of the HbA1c data and several other relevant laboratory, clinical and functional parameters (including blood lipids, liver enzymes, and measures of regional fat from MRI, n=561). The longitudinal components of the data were described as proportional functions of time, with normally distributed slope (σ) describing underlying progression rate and normally distributed residual unexplained variability. The used approach, named conditional linear mixed-effect modelling, is a tool for characterizing individual longitudinal patterns while minimizing the potential bias introduced by assumptions on and shrinkage of cross-sectional effects [4]. The σHbA1c values were adjusted for changes in body mass index (BMI) and in standardized dosage of antidiabetic treatments; the σOGIS values were adjusted for BMI changes. Modelling was performed with MonolixSuite2016 R1. Multivariate linear and logistic regression of σHbA1c was performed considering the baseline values and progression rates of all available parameters as potential independent variables. In the logistic regression, subjects were divided into fast and average progressors based on a σHbA1c threshold of 2.7 mmol mol-1 y-1 (chosen based on the presence of a right tail in the σHbA1c distribution). Fast progressors were 33.
Results: The progression rates for all considered parameters were successfully estimated for all patients. In the linear analysis of σHbA1c, faster HbA1c progression was independently associated with lower baseline values and faster deterioration of OGIS and GS, with higher baseline values of CLIo and its increase, and with RS reduction. Faster HbA1c progression was also independently associated with lower baseline HDL-cholesterol (HDL) or its slower increase, with a quicker increase of triglycerides, as well as with higher baseline values of BMI and lower baseline values of HbA1c (adjusted R2 = 0.38). The variables with strongest effects were the baseline OGIS value and the progression rates of OGIS, GS and CLIo. The subjects with available MRI revealed an independent effect of visceral or liver fat, replacing the effects of baseline BMI and HDL. The Receiver Operating Characteristic (ROC) AUC from the logistic regression using the same determinants was 0.94 ([0.86, 0.98] 95% CI). Stronger deterioration and a lower baseline value of OGIS and GS, CLIo increase, and HDL reduction were the features independently associated to fast progression.
Conclusions: (1) Conditional linear mixed-effect modelling proved to be a useful way of describing longitudinal data with hierarchical structure. (2) Quickly decreasing peripheral insulin sensitivity, increasing insulin clearance, and deteriorating β-cell function and fat metabolism independently associate with rapid HbA1c progression in white adults from northern Europe with recently-diagnosed T2D. (3) The temporal trajectories of these mechanisms were able to discriminate a subgroup of fast progressors with high performance. (4) Mathematical methods and models for the assessment of insulin sensitivity and β-cell function are effective tools to understand deterioration of glucose homeostasis.
Acknowledgements:
We are grateful to all participants of the DIRECT consortium. The work leading to this publication has received support from the Innovative Medicines Initiative Joint Undertaking under grant agreement n°115317 (DIRECT), resources of which are composed of financial contribution from the European Union’s Seventh Framework Programme (FP7/2007-2013) and EFPIA companies’ in kind contribution.
References:
[1] Koivula RW, Heggie A, Barnett A, et al. Discovery of biomarkers for glycaemic deterioration before and after the onset of type 2 diabetes: rationale and design of the epidemiological studies within the IMI DIRECT Consortium. Diabetologia, 57(6):1132-1142, 2014.
[2] Koivula RW, Forgie IM, Kurbasic A, et al. Discovery of biomarkers for glycaemic deterioration before and after the onset of type 2 diabetes: descriptive characteristics of the epidemiological studies within the IMI DIRECT Consortium. Diabetologia 62(9):1601-1615, 2019
[3] Mari A, Tura A, Gastaldelli A, Ferrannini E. Assessing Insulin Secretion by Modeling in Multiple-Meal Tests: Role of Potentiation. Diabetes 51: S221–S226, 2002.
[4] Verbake G. Conditional Linear Mixed Models. The American Statistician 55: 25–34, 2001.
Reference: PAGE () Abstr 9239 [www.page-meeting.org/?abstract=9239]
Poster: Oral: Drug/Disease Modelling