Search for a command to run...
Background Sensory impairment is a strong modifiable risk factor for dementia; however, whether vitamin D deficiency (VDD) was associated with increased risk in this high-risk group remains unclear. Individuals with vision or hearing impairment are prone to VDD due to reduced sunlight exposure, making this question clinically relevant. Methods From the TriNetX Global Collaborative Network, this retrospective cohort study identified adults aged ≥50 years with documented vision and/or hearing impairment and serum 25-hydroxyvitamin D [25(OH)D] measurement. Patients were classified into VDD group (<20 ng/mL) and control group (≥30 ng/mL). The index date was defined as the date of the first 25(OH)D measurement that met the cohort-specific threshold and satisfied all eligibility criteria. The primary outcome was incident dementia over a 10-year follow-up. Secondary outcomes included dementia subtypes (i.e., vascular dementia and Alzheimer’s disease), cognitive impairment, osteoporotic fracture (positive control), healthcare visits (detection bias assessment), and recurrent VDD. Results After propensity score matching, 158,382 patients were included in each cohort. Compared to the control group, the VDD group was associated with a significantly higher risk of incident dementia [hazard ratio (HR), 1.55; p < 0.001], vascular dementia (HR, 1.70; p < 0.001), Alzheimer’s disease (HR, 1.48; p < 0.001), cognitive impairment (HR, 1.40; p < 0.001), and subsequent VDD (HR, 4.73; p < 0.001). The risk of osteoporotic fracture was significantly associated with VDD (HR, 1.34; p < 0.001), whereas healthcare visits were slightly lower in the VDD group (HR, 0.91; p < 0.001), arguing against detection bias. Vitamin D insufficiency was associated with a significant but attenuated association with dementia (HR, 1.39; p < 0.001). Conclusion In this cohort study, VDD was associated with an increased risk of incident dementia in adults with vision and/or hearing impairment, with exploratory findings supportive of a graded pattern. Although residual confounding cannot be excluded, these findings raise the possibility that low vitamin D status may be a potentially modifiable contributor to dementia risk in sensory-impaired populations.