Endocrine Abstracts (2019) 65 P203 | DOI: 10.1530/endoabs.65.P203

The track-pump study

Agnes Johnson, Nick Oliver, Ian Godsland & Neil Hill


Imperial College Healthcare NHS Trust, London, UK


Background: ‘Glycaemic tracking’ describes the stability of HbA1c over time commonly observed in individuals with type 1 diabetes (T1DM). Preliminary studies have noted the relationship between early metabolic control and long-term outcomes, while recent evidence has been presented for glycaemic tracking in a large number of newly diagnosed individuals with T1DM. The benefits of tight glycaemic control, in terms of reducing microvascular complications and lowering HbA1c, are well reported. Insulin pump therapy may be used to achieve this strict glycaemic control more effectively.

Aims: To investigate if the phenomenon of glycaemic tracking occurs in a cohort of people with T1DM. To analyse the effect of insulin pump therapy on tracked HbA1c levels.

Methods: HbA1c readings were collected retrospectively for 160 patients on insulin pumps at Charing Cross Hospital. Statistical analyses were undertaken to compare yearly median HbA1c for the ten years surrounding inception of pump use. Pre- and post-pump median HbA1c were compared for paired data sets of varying time frames. HbA1c was plotted against time in 90-day intervals, and linear regression performed to model this relationship.

Results: The only significant (P<0.0001) difference in annual median HbA1c was found between the values from one year either side of the pump date. For all time frames analysed, there was a significant reduction in median HbA1c and IQR/median ratio following pump inception. Regression modelling of pre- and post-pump HbA1c values demonstrated a 4.4 mmol/mol lower Y-intercept at pump inception and a significantly more stable gradient (P=0.0004) post-pump.

Discussion: This study provides supportive evidence for glycaemic tracking in T1DM. Our results suggest that insulin pump therapy results in an incremental improvement in glycaemic control and a reduction in HbA1c variability. Larger-scale, prospective studies are needed to test this hypothesis further, and analyse the relationship between the timing of intervention and HbA1c trajectory.

Article tools

My recent searches

No recent searches.

My recently viewed abstracts