Examining the relationship between multimorbidity and glycaemia, glycaemic variability and all-cause mortality in people with type 2 diabetes in Taiwan: Findings from a Longitudinal Cohort Study.

Conference:

SAPC ASM 2018 - London

Talk Code:

P2.74

Presenter:

Me (Jason I Chiang)

Twitter:

@Jason_I_Chiang

Co-authors:

Jason I Chiang, Sing-Yu Yang, Jo-Anne Manski-Nankervis, Sharmala Thuraisingam, John Furler, Frances Mair, Barbara I Nicholl, Bhautesh Jani, Tsai-Chuang Li, Cheng-Chieh Lin

Author institutions:

University of Melbourne, China Medical University, University of Glasgow

Problem

Multimorbidity (MM) is highly prevalent in people with type 2 diabetes (T2D). Our study explored associations between MM condition count and all-cause mortality, glycaemia (HbA1c), and glycaemic variability ((GV), an emerging measure of glycaemia defined as fluctuations in blood glucose levels) in a population of people with T2D in Taiwan.

Approach

We examined data from a longitudinal cohort of 34,229 people with T2D enrolled in the Taiwan National Diabetes Care Management Program between 2002 and 2004, followed until 2011. Primary independent variable was the number of multimorbid conditions in addition to T2D. Outcomes of interest were: all-cause mortality; HbA1c and GV represented by coefficient of variation (CV) of HbA1c (HbA1c-CV) and fasting plasma glucose (FPG-CV). We examined associations between independent and dependent variables using Cox proportional hazard regression (MM and all-cause mortality) and multivariable linear regression (MM and HbA1c and GV).

Findings

We found a statistically significant association between increasing MM and mortality. Compared to those with T2D with no additional chronic conditions, having 1, 2, 3 and 4+ conditions in addition to T2D, the hazard ratio (HR)(95%CI) for all-cause mortality were 1.18 (0.99-1.19), 1.30 (1.18-1.42), 1.67 (1.52-1.84), and 2.23 (2.03-2.46), respectively. Both concordant (related to T2D) and discordant (unrelated to T2D) conditions were significantly associated with mortality. For 1, 2, 3 and 4+ concordant conditions in addition to T2D, the HRs (95%CI) were 1.13 (1.06-1.20), 1.47 (1.36-1.58), 2.12 (1.92-2.35), and 2.87 (2.48-3.32), respectively and for discordant conditions 1.28 (1.20-1.37), 1.28 (1.20-1.36), 1.42 (1.32-1.54), and 1.44 (1.30-1.59), respectively.There was an association between having MM (1, 2, 3 and 4+ additional conditions) and HbA1c, however it was not related to condition count (β(95%CI) -0.30 (-0.35, -0.25), -0.33 (-0.39, -0.28), -0.32 (-0.38, -0.26), and -0.31 (-0.38, -0.24) respectively). For GV, 4+ additional conditions to T2D showed a significant association with increased FPG-CV with β(95%CI): 1.96 (1.33, 2.59) while 1, 2 and 3 conditions in addition to T2D was associated with decreased HbA1c-CV (β(95%CI) -1.20 (-1.47, -0.93), -0.98 (-1.26, -0.70), and -0.37 (-0.69, -0.05), respectively) and 4+ conditions associated with increased HbA1c-CV (β(95%CI): 0.56 (0.21,0.91)).

Consequences

Our study demonstrates that increasing MM is a significant predictor of all-cause mortality. The association between MM and mortality is significant for total, concordant and discordant condition count, but was strongest for concordant. The association between having MM and HbA1c is not related to increasing MM condition count. The effect of MM on glycaemic variability is mixed. Our findings emphasis the need for clinical guidelines to consider MM in people with T2D, both concordant and discordant.

Submitted by:

Jason Chiang

Funding acknowledgement:

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.