Insomnia symptoms may influence the relationship between self-reported and objective cognition scores in older adults, according to a study published in the Journal of Sleep Research.
It is well known that poor subjective evaluation of cognition and sleep is related to cognitive decline among older adults. Recognizing that poor subjective assessment of cognition and sleep is linked to cognitive decline among older adults, researchers sought to evaluate the interactive associations of specific objective cognitive domains (ie, processing speed, attention/processing speed, verbal episodic memory, inductive reasoning) and subjective sleep parameters (ie, total sleep time [TST], sleep onset latency [SOL], wake after sleep onset [WASO], sleep efficiency [SE]) with self-reported cognition in older adults with insomnia.
An additional aim of the study was to comprehend the specific cognitive domains that are associated with self-reported cognition.
The researchers hypothesized that worse sleep would interact with objective cognitive performance in the associations with self-reported cognitive ratings, based on prior findings in middle-aged and older adults without insomnia. In particular, they hypothesized that objective cognitive performance would be inversely associated with self-reported ratings of cognitive performance, and that this association would be stronger among those who experienced worse sleep (ie, longer SOL, shorter or longer TST, longer WASO, and lower SE).
The study was conducted in a group of individuals who were recruited as part of a larger randomized controlled trial, Intraindividual Variability in Sleep and Cognitive Performance in Older Adults (REST; ClinicalTrials.gov Identifier: NCT02967185). The trial, REST, examined the effects of behavioral therapy for insomnia in older adult patients.
A total of 51 older adults with insomnia completed 14 days of self-reported cognition ratings (0=poor, 100=very good); sleep (ie, TST, SOL, WASO, SE) ratings; and daily cognitive tasks (ie, letter series [reasoning], word list delayed recall [verbal memory], Symbol Digit Modalities Test [SDMT, attention/processing speed], and number copy [processing speed]). For each cognitive task, multiple regressions established whether average objective cognition or sleep was independently/interactively associated with average self-reported cognition, adjusting for age, education, and depression. The mean age for a patient age was 69.19±7.95 years and 70.6% were women.
The researchers found the interaction between SDMT performance and TST was associated with self-reported cognition. In particular, the link between scores and self-reported cognition was congruent in those with the longest and average WASO, as well as with the shortest and average SE. All full regression models, including TST, SOL, WASO, and SE, were statistically significant (P =.001, P =.01, P <.0001, and P <.001, respectively).
The interaction between attention/processing speed and TST was associated with self-reported cognition (P =.007). Lower SDMT scores (ie, worse attention/processing speed) were associated with lower self-reported cognition ratings at the shortest (ie,
approximately 282 minutes; P =.0002) and average (ie, approximately 362 minutes; P =.0001) TST, but not the longest (approximately 442 minutes; P =.48) TST.
Additionally, the interaction between attention/processing speed and WASO was significantly associated with self-reported cognition (P =.0001). Lower SDMT scores (ie, worse attention/processing speed) were significantly associated with lower self-reported cognition ratings at longer (ie, approximately 96 minutes; P <.0001) and average (ie, approximately 62 minutes; P <.0001) WASO, but not at shorter (ie, approximately 28 minutes; P =.55) WASO.
Further, the interaction between attention/processing speed and SE was associated with self-reported cognition (P =.002). Lower SDMT scores (ie, worse attention/ processing speed) were associated with lower self-reported cognition ratings at the lowest (ie, approximately 60%; P <.0001) and average (ie, approximately 76%; P <.0001) SE, but not the highest (ie, approximately 85%; P =.87) SE.
These findings imply that in an older adult population with insomnia, a compatible association exists between attention/processing speed and self-reported cognition in those who describe worse subjective sleep (ie, shorter TST, longer WASO, and lower SE).
Several limitations of the present analysis should be noted. First, because the study used a nonstandardized measure of self-reported cognition, the generalizability of the findings may be limited. Further, participants completed the objective cognition tests at home, rather than in the laboratory. The current study utilized paper and pencil type of cognitive tasks, instead of online testing, which may provide a greater level of control.
According to the researchers, “Understanding how sleep problems play into the relationship of objective cognition and self-reported cognition will better help.”
Costa AN, Curtis AF, Musich M, Guandique AA, McCrae CS. Self-reported cognition in older adults with insomnia: associations with sleep and domain specific cognition. J Sleep Res. Published online October 11, 2022. doi:10.1111/jsr.13751