[Ophthalmological traits in older adult and risk of Alzheimer's disease: results from a French geriatric cohort]. / Caractéristiques ophtalmologiques chez la personne âgée et risque de maladie d'Alzheimer : résultats d'une cohorte française gériatrique.

ABSTRACT

Ophthalmological changes have been reported in Alzheimer's patients. Our objectives were to determine whether i) GCC (ganglion cell complex) and RNFL (retinal nerve fibre layer) thickness were associated with different stages of AD (i.e., no AD, prodromal AD, dementia-stage AD), and ii) GCC and RNFL thickness predicted disease progression in older non-demented patients with subjective memory complaints followed for four years. Ninety-one French older community-dwellers with memory complaint and without open-angle glaucoma or age-related macular degeneration (mean, 71.60 ± 4,73 years; 44% women) from the GAIT study underwent examination with HD-OCT, measuring the thickness of the macula, the macular GCC and the RNFL. They also had a complete cognitive diagnosis (i.e., cognitively healthy, prodromal AD, or dementia AD), and a cognitive follow-up 4 years later looking for a possible conversion. Age, sex, body mass index (BMI), number of comorbidities, and Instrumental activities of daily living (IADL) score were considered as potential confounders. At baseline, 37 (40.7%) patients were diagnosed as cognitively healthy, 47 (51.6%) as MCI, and 7 (7.7%) as AD. Mean GCC thickness was higher in cognitively healthy patients than in MCI patients (79.23 vs. 76.27 µm, p = 0.023), particularly in the inferior and nasal fields (p = 0.023 and p = 0.005, respectively). This difference was also found between cognitively healthy patients and others (MCI and AD) in the superior, inferior and nasal fields (p = 0.030, p = 0.014 and p = 0.002, respectively). There was no difference in RNFL thickness between the different cognitive statuses. After 4 years of follow-up, 12 patients (70.6%) of the 17 followed had not changed their cognitive status, while 5 (29.4%) had converted to a more advanced stage of AD. There were no significant differences between the two groups in either GCC thickness (p = 0.429) or RNFL thickness (p = 0.286). We found decreased CGG thicknesses in Alzheimer's patients at prodromal and dementia stages, compared with cognitively healthy participants. There was no association between RNFL thickness and cognitive status, nor between CCG or RNFL thicknesses and the risk of progressing to AD stages after 4 years of follow-up.