OBJECTIVES : Spontaneous online conversations by patients and caregivers across the clinical severity stages of Alzheimer’s Disease (AD) were used to inform how to tailor patient-centric diagnostic, disease and lifestyle management approaches over the course of disease.

METHODS : We used Natural Language Processing (NLP) to analyze Social, Physical, Emotional, Cognitive, and Role Activity (SPEC-R) issues extracted from patient and caregiver narratives over 84 public social media sites to evaluate patient burden in Mild Cognitive Impairment due to AD (MCI-AD), Early-AD (E-AD) and Moderate-Severe AD (MS-AD).

RESULTS : 112,464 narratives from 692 patients and 10,174 caregivers were qualified into the final analytic sample with 333 patients clinically diagnosed for MCI-AD and 359 for E-AD. No individuals with MS-AD were captured over the social media forums. Of the caregivers, 971 were caring for patients with MCI-AD, 3,776 for patients with E-AD, and 5,427 for patients with MS-AD. Among patients with MCI-AD and their caregivers, 95.5% patients and 83.0% caregivers discussed cognitive issues, e.g. memory impairments. 61.1% patients and 51.4% caregivers shared emotional issues, e.g. anxiety and depression. Among patients with E-AD and their caregivers, emotional issues were the most frequently discussed by both patients (93.6%) and caregivers (53.5%), e.g. anxiety, depression and other emotional experiences. Cognitive issues followed, and shared by 86.9% patients and 39.6% caregivers. Among caregivers for patients with MS-AD, 42.7% reported on emotional issues and 20.7% on physical issues, e.g. fatigue and agitation.

CONCLUSIONS : Individuals with MCI-AD or E-AD were active on social media to share their experiences and challenges while those with MS-AD appeared to no long participate in online communications relating to their illness. Caregivers were active on social media across disease severity stages. However, frequencies of caregiver reporting on prominent patient challenges differed from that of patient self-reporting.