The effect of level of injury on diabetes incidence and mortality after spinal cord injury – a longitudinal cohort study

Abstract

AbstractObjectivePersons with spinal cord injury (SCI) are at increased risk to develop diabetes mellitus (DM) compared to their able-bodied counterparts, likely due to body composition alterations and autonomic nervous system dysfunction. These factors are more pronounced in persons with tetraplegia (TP) versus paraplegia (PP), however, the effect of level of injury (LOI) on DM incidence is largely unknown. Therefore, the objective is to examine the effect of LOI on DM incidence in persons with SCI.DesignRetrospective longitudinal cohort study of veterans with SCI.MethodsWe obtained electronic record data on age, sex, race/ethnicity, LOI and HbA1c concentration from January 1st2001 through December 31st2021. Cox proportional hazard regression analyses were used to assess the association between LOI, DM and all-cause mortality.ResultsAmong 728 non-diabetic veterans with SCI (350 TP/ 378 PP, 52±15 years, 690 male/38 female) 243 developed DM, of which 116 with TP and 127 with PP. Despite chronological variations between TP and PP, DM risk over the entire follow-up did not differ among the groups (hazard ratio (HR): 1.06, 95% CI: 0.82 - 1.38). Mortality was higher in TP versus PP (HR: 1.40, 95% CI: 1.09 – 1.78). However, developing DM did not increase the risk of death, regardless of LOI (HR: 1.07, 95% CI: 0.83 – 1.37).ConclusionsIn this cohort of veterans with SCI, the level of injury had minimal effect on long-term DM development but increased mortality as previously reported.Significance statementPersons with spinal cord injury are at increased risk for developing diabetes mellitus (DM); however, the effect of level of injury is unclear. In the current study using an electronic health record system we found that the incidence of DM is similar between persons with a high (i.e., tetraplegia) versus low (i.e., paraplegia) lesion level, and that developing DM does not affect the risk for mortality. Additionally, our findings suggest that baseline glycemic level (such as HbA1c), unlike baseline body mass index, is a strong predictor of DM development in this population. Further studies are warranted to explore pathophysiological factors responsible for DM development among patients with higher and lower LOI to develop targeted preventive and therapeutic strategies.