Cross-sectional and longitudinal associations between sleep disturbances and psycho-affective symptoms in older adults: influence of amyloid positivity

BackgroundPsycho-affective symptoms and repetitive negative thinking (RNT) may increase Alzheimers disease (AD) risk. As sleep plays a key role in emotional regulation, we investigated both cross-sectional and longitudinal associations between sleep disturbances and psycho-affective health, according to amyloid-beta (A{beta}) status. MethodsOne hundred and thirty-four cognitively unimpaired older adults from the Age-Well interventional study (mean age = 68.8 {+/-} 3.8 years; 82 women; 37 A{beta}+ individuals) completed psycho-affective (Geriatric Depression Scale, State-Trait Anxiety Inventory), RNT (Rumination Response Scale-brooding, Penn State Worry Questionnaire) and sleep questionnaires (Pittsburgh Sleep Quality Index, Insomnia Severity Index). Wrist actigraphy (n=131; mean recording duration = 7.7 {+/-} 0.5 nights) provided objective measures of sleep fragmentation (sleep fragmentation index, wake time after sleep onset) and their night-to-night variability. Cross-sectional multiple linear regressions and longitudinal linear mixed-effects models (mean follow-up = 3.98 {+/-} 1.15 years) examined associations between sleep and psycho-affective symptoms, adjusting for age, sex, continuous positive airway pressure use, and the non-pharmacological intervention group (for longitudinal analyses). ResultsIn the whole cohort and in A{beta}- individuals, higher self-reported sleep difficulties and insomnia symptoms were cross-sectionally associated with greater anxiety, depression and RNT (all pFDR-corr<0.05). In A{beta}+ individuals, objective sleep fragmentation and instability were cross-sectionally associated with higher anxiety, depression and RNT. Longitudinally, higher baseline insomnia severity, mean sleep fragmentation, and night-to-night sleep instability predicted brooding worsening in A{beta}+ individuals only, while poorer perceived sleep quality predicted depressive symptoms worsening (all pFDR-corr<0.05). ConclusionsIn cognitively unimpaired older adults, associations between sleep disturbances and psycho-affective symptoms differ according to A{beta} status. In A{beta}+ individuals specifically, objective sleep fragmentation and instability are linked to psycho-affective vulnerability and brooding worsening over time. Sleep disturbances could thus represent early modifiable targets to mitigate psycho-affective symptomatology in older adults.