Abstract
METHODS
Patients with primary CRC (1998-2011) were selected from the Eindhoven Cancer Registry and linked to the PHARMO Database Network including outpatient pharmacy and clinical laboratory data. Patients with more than 2 years of GLDs use prior to cancer diagnosis were included. Linear mixed-effects models were conducted to evaluate changes in HbA1c for colon cancer (CC) and rectal cancer (RC) patients in the 4 years around CRC diagnosis.
CONCLUSIONS
Among users of GLDs, HbA1c decreased with 0.12-0.18 % (1-2 mmol/mol) per year before CRC diagnosis. Only among CC patients, HbA1c increased after diagnosis (0.12 % per year; 1.3 mmol/mol). Modest changes in HbA1c before CRC diagnosis may reflect the effects of an undiagnosed cancer, such as weight loss, anaemia, or the use of anti-anaemic preparations.
RESULTS
Of all CRC patients (n = 4714), 294 (6 %) GLDs users with CC and 144 (3 %) with RC were selected. In the crude model, mean HbA1c at cancer diagnosis was 6.9 % (51.6 mmol/mol) among CC patients and 7.1 % (53.5 mmol/mol) among RC patients. Among CC patients, HbA1c decreased with 0.12 % per year (p = 0.0002) before cancer diagnosis in the adjusted model, and after diagnosis, it increased with 0.12 % per year (p = 0.02). In subgroup analyses, effects on HbA1c were more pronounced in users of anti-anaemic preparations. Among RC patients, HbA1c decreased before diagnosis with 0.18 % per year (p = 0.0006), whereas after diagnosis it changed non-significantly.
AIMS
This study aims to evaluate the impact of cancer and its treatment on HbA1c values among individuals with colorectal cancer (CRC) using glucose-lowering drugs (GLDs).