Abstract
Abstract
Introduction
Insomnia is a common sleep disorder that affects about half of people with type 2 diabetes (T2D). Recently, it has been found that sleep disturbance (i.e., sleep continuity disturbance, short sleep, and/or circadian dysrhythmia) is associated with the dysregulation of insulin secretion and insulin response at molecular sites. Therefore, in order to assess the specific association of insomnia with A1C and glucose levels, an archival analysis was undertaken evaluating A1C and glucose levels in patients with T2D with and without insomnia.
Methods
A retrospective electronic medical charts review was conducted at an academic medical system. Two groups from 01/01/2012 to 08/14/2017 were identified. The first group included patients who had diagnoses codes of insomnia and T2D (insomnia+T2D), and the second group included patients who only had a diagnosis code of T2D (T2D-only). Charts with at least two records of HbA1c and glucose level were included. A mixed model analysis compared the difference in A1C variability between groups. Then, Likelihood ratio test compared the within-patients’ mean and standard deviation between groups.
Results
105 insomnia+T2D patients (50% females, mean age= 66.67 ± 13.13 years) and 854 T2D-only patients (49.82% females, mean age=66.99 ± 12.82) were identified. There was no significant between-group difference in age, gender, race, and BMI. Patients in the insomnia+T2D group had higher A1C variability (σ2=1.99) compared to T2D-only patients (σ2=1.53), p<0.001. This significant difference was present after using the Likelihood ratio test, p<0.001. The A1C between-group difference was 0.46 which is clinically meaningful. Glucose level was higher in the insomnia+T2D group (176.7 ± 99.29) compared to the T2D-only group (162.42 ± 73.16), p=.04.
Conclusion
Patients with insomnia and T2D had higher A1C variability and glucose levels compared to people with only T2D. A1C variability contributes to long term complications. Limitation in this study include not accounting for medication effects as this may change the prevalence of insomnia diagnoses. Therefore, prospective study is needed to confirm our findings. Furthermore, future research might consider treating insomnia in people with T2D to improve diabetes outcomes.
Support (If Any)
None.