Research Signal – Source date 2023-05-09
For people already living with cognitive impairment, managing blood pressure and repeated breathing interruptions during sleep may be more than a matter of general health. A study published on 9 May 2023 examined whether treatment status for arterial hypertension and obstructive sleep apnea was associated with the subsequent course of cognition in people with common neurocognitive disorders.
The findings point toward a possible benefit from addressing these conditions, but they do not prove that treatment caused cognition to decline more slowly. The study was a retrospective analysis of 126 people from the German RIFADE cohort. It followed cognitive measurements over time and compared patterns across people with Alzheimer’s disease, vascular neurocognitive disorder, mixed disease, and a smaller unspecified group.
The study followed cognition after impairment was present
The participants were not healthy adults being studied for dementia prevention. They had already entered care with a neurocognitive disorder. The cohort included people diagnosed with Alzheimer’s disease, vascular neurocognitive disorder, or a combination of the two. Some participants had an unspecified diagnosis. People living in nursing homes and people with several other neurological conditions, including severe Parkinson’s disease, frontotemporal degeneration, or Lewy body disease, were excluded.
Researchers screened the cohort for arterial hypertension and obstructive sleep apnea. Ninety participants, or 71.4%, had hypertension. Forty, or 31.7%, had sleep apnea. The research question was not simply whether either condition appeared in the medical record. The investigators also examined whether it was treated during the periods in which cognitive change was measured.
Cognition was assessed repeatedly with the DemTect, the study’s primary outcome, and with a clock-drawing test as a secondary outcome. A DemTect score at a visit is one kind of measurement. The change between two neighboring measurements is another. The difference matters because a factor can be associated with the level of a score without showing the same pattern in the amount of change between visits.
Treatment status was measured across observation intervals
The researchers divided each participant’s follow-up into periods between successive cognitive measurements. For each period, hypertension and sleep apnea were classified according to whether the condition was absent, treated, or untreated. A condition counted as corrected for that interval when it met the study’s correction criteria for more than half of the period. This interval-based classification was different from saying that treatment had been maintained continuously across the entire study.
The analysis used mixed-effects repeated-measurement models. These models accounted for the fact that the same person contributed multiple observations and adjusted for age, education, baseline DemTect score, and timing variables such as time since inclusion or time to measurement. The approach helped compare cognitive patterns while allowing for differences between individuals, but it could not recreate the balance of a randomized trial.
The distinction between a risk-factor state and a treatment effect is important throughout the paper. A statistically significant term involving absent, treated, and untreated states indicates that the groups differed in the modeled outcome. It is not automatically a direct estimate of how much treatment helped, nor does it show that every treated group differed significantly from every untreated group.
The main results changed with the model
In the two-factor analysis combining hypertension and sleep apnea, hypertension status was associated with change in DemTect scores. The hypertension-by-sleep-apnea interaction was not significant, and the sleep-apnea term was not significant for that particular change outcome. In other words, the joint model did not establish a statistically reliable synergistic effect in which treating both conditions produced a distinct combined benefit.
When diagnosis was added to the model, the picture became more specific and more complicated. The three-way interaction among hypertension, sleep apnea, and diagnosis was significant for DemTect scores. Sleep-apnea status also had a significant main effect on those scores. In a separate model examining sleep apnea and diagnosis, the sleep-apnea-by-diagnosis interaction for DemTect change was not significant. These are different statistical questions, so the results should not be compressed into the claim that sleep-apnea treatment improved cognition in every diagnostic group.
The study’s estimated patterns were nevertheless notable. Across the Alzheimer’s, vascular, and mixed groups, the modeled change in DemTect scores was generally better when sleep apnea was treated than when it was untreated. In Alzheimer’s and vascular neurocognitive disorder, the most favorable pattern appeared when both sleep apnea and hypertension were treated. In mixed neurocognitive disorder, the most favorable combination in the reported estimates involved treated sleep apnea with absent hypertension.
The least favorable patterns also differed by diagnosis. Untreated sleep apnea appeared in the worst combinations for the Alzheimer’s and vascular groups. In the mixed group, the least favorable reported pattern involved absent sleep apnea with untreated hypertension. These diagnosis-related comparisons are useful for generating questions, but they are not treatment-selection rules. Some of the relevant cells were small, and untreated-hypertension comparisons were not available in the Alzheimer’s and vascular groups. The apparent differences therefore cannot be treated as complete evidence for how treatment should be chosen by diagnosis.
Why the comparisons cannot prove causation
Treatment was observed rather than assigned. People who received treatment could differ from those who did not in ways that were only partly captured by the available measurements. The treated groups differed in age, education, and starting cognitive score, and the authors adjusted for those factors in the models. Adjustment can reduce the influence of measured differences, but it does not make observational groups equivalent to randomized groups.
The timing of treatment may also matter. A person able to begin or maintain therapy may have had different access to care, symptom recognition, clinical follow-up, tolerance, or overall health than someone who remained untreated. The study design cannot determine which of these factors contributed to the observed cognitive patterns. Nor can it establish that controlling hypertension or sleep apnea reverses neurocognitive disease, guarantees slower decline, or produces a clinically meaningful improvement in daily independence.
The cohort itself also limits generalization. It came from one German clinical setting, included 126 people, and excluded several diagnostic groups and nursing-home residents. Exclusions and small diagnostic cells make the reported patterns important to investigate, while leaving uncertainty about how consistently they would appear in other populations.
Sustained treatment was much less common for sleep apnea
The treatment counts show a striking difference between the two conditions. Of the 90 participants with hypertension, 79 had permanently corrected hypertension during the observation. Among the 40 people with sleep apnea, only 13 had permanent treatment. Eleven of those 13 used continuous positive airway pressure, while one used a mandibular advancement device and one used position therapy.
The authors discuss the practical challenge of treating sleep apnea in people with neurocognitive disorders, including the fact that many patients do not tolerate continuous positive airway pressure. That point should not be turned into a documented explanation for every untreated participant in this cohort. The study did not establish a single reason why each person remained untreated, and some participants with untreated sleep apnea still had favorable cognitive outcomes.
The uneven treatment exposure changes how the results should be read. The paper raises the possibility that sleep-apnea treatment could be underused in a population already experiencing cognitive impairment, but it does not show that one device or treatment approach is best for every patient. These findings do not support individualized recommendations for blood-pressure targets, medication choices, or device settings.
What this could change
If the observed associations are reproduced in stronger studies and sustained treatment can be made feasible, coordinated management of blood pressure and sleep apnea could become one part of helping people with existing neurocognitive disorders preserve cognitive abilities for longer. That would offer patients and families a practical area of care to discuss alongside the broader management of cognitive disease.
The possible benefit remains conditional. These findings do not establish an exact number of years of preserved cognition, a percentage reduction in dementia risk, improved independence, or reduced caregiver burden. Its contribution is narrower but still meaningful: it suggests that treatment status may be relevant to the cognitive course after impairment is already present, while showing why diagnosis, measurement method, and treatment feasibility all need to be considered when interpreting this observational evidence.
Key Sources
- Primary source: Treatment of hypertension and obstructive sleep apnea counteracts cognitive decline in common neurocognitive disorders in diagnosis-related patterns.
DOI or Journal Page · Peer-Reviewed Paper · 2023-05-09
Keep exploring