Exploring Orexin Neuron Loss as an Early Driver of Sleep Disturbances in Alzheimer Disease

Lea Grinberg, MD, PhD

While much of the traditional Alzheimer disease (AD) research has focused on amyloid accumulation, recent insights point to tau pathology as a stronger driver of clinical decline. Importantly, tau appears in different brain regions than amyloid and may begin accumulating in subcortical nuclei that regulate wakefulness even before it appears in the entorhinal cortex.

At the 2025 SLEEP Annual Meeting, held June 8-11 in Seattle, Washington, dementia expert Lea Grinberg, MD, PhD, presented on the role of orexin/hypocretin neuron loss in aging and early AD, and how these early changes may connect to sleep dysfunction. Above all, her work highlighted the potential for sleep-focused biomarkers and existing neuromodulatory drugs to reshape both diagnosis and treatment strategies in the years ahead.

During the meeting, NeurologyLive^® caught up with Grinberg to discuss her research linking tau accumulation in sleep-wake regulatory nuclei to early sleep disturbances in AD. Grinberg, a neuropathologist working at Mayo Clinic Florida, explained how therapies targeting tau or orexin could influence sleep and potentially slow disease progression, while also addressing clinical considerations and challenges in applying these approaches. Furthermore, she touched on the promise of EEG-based biomarkers, the need to integrate sleep measures in AD trials, and why daytime monitoring could provide a more sensitive window into early disease biology.

NeurologyLive: Discuss your presentation in detail, why was this a topic of interest for you?

When we think about Alzheimer disease, most people immediately think of amyloid and tau accumulation in the brain. For many years, the focus was on amyloid spread. But more recently, it has become clear that the best correlate of clinical decline is actually tau accumulation. Something that even some specialists may not realize—unless they are neuropathologists like me—is that the regions of the brain where amyloid and tau begin to accumulate are completely different. They move in opposite directions, so at very early stages of the disease they are disconnected in terms of topography. This gave us the opportunity to better understand what symptoms appear earliest.

One of the key contributions from my lab over the last two decades has been showing that tau accumulation doesn’t just begin in the entorhinal cortex, as is commonly thought. It actually starts earlier in subcortical regions. And when we began connecting the dots, we noticed that many of these subcortical nuclei are those responsible for regulating wakefulness. Since sleep and wake are tightly interconnected, we asked whether tau in these areas could explain some of the early sleep problems seen in Alzheimer disease.

There is a lot of epidemiological evidence showing that individuals who later develop Alzheimer disease often have a history of sleep disturbances, even worse than the general population. This inspired us to investigate further. And yes, we found that certain areas involved in wake regulation lose neurons very early in the disease—sometimes before other regions show pathology. These losses correlate directly with EEG signatures of disrupted sleep and wake cycles, which may be detectable earlier than cognitive decline.

What we’ve now discovered is that orexinergic neurons, part of the wake-promoting system, are the first to die in Alzheimer disease—even before the locus coeruleus. The locus coeruleus does lose volume early, but our work shows that this shrinkage is secondary to the loss of orexinergic input, not due to direct neuronal loss at that stage. So, our research is now mapping the progression of orexinergic neuron loss in both structural and molecular terms, and combining it with data from other affected systems. Importantly, we’re also studying receptor-level changes, because many drugs already exist that target these neuromodulatory systems. Although they were not developed for Alzheimer disease, their safety is established, and they may be repurposed to alleviate sleep dysfunction and possibly slow disease progression.

We’ve got approved therapies that focus on amyloid right now, and several in development for tau. Could tau-targeting therapies have an impact on sleep and broader Alzheimer risk?

We’ll have to wait for the studies, but if I had to make a prediction, I would say yes—tau therapies are much more likely than amyloid therapies to help with sleep. That’s because the nuclei we’re studying are not significantly affected by amyloid until very late, if at all. But tau affects them at what we call Braak stage 0—the very earliest stage of Alzheimer pathology. So it makes a lot of sense.

I also think we’ll eventually be using these drugs not just for treating diagnosed Alzheimer disease, but for people in midlife who present with persistent sleep problems. Once our biomarkers become sensitive enough to detect the biology earlier, these therapies may be applied to improve sleep and hopefully delay disease progression.

What are some of the challenges with developing or implementing an orexin-targeted therapy?

There are already drugs on the market, and several recent papers—many in animal models, but some small human studies—suggest that if used correctly, they could influence Alzheimer pathology. But my clinical colleagues are not yet using them. Even when drugs are FDA approved, clinicians remain hesitant because of concerns about side effects and drug half-life. That said, newer agents with shorter half-lives may offer safer profiles for this population.

Another challenge is that many clinicians don’t yet fully understand the underlying biology. Even some Alzheimer specialists are surprised when they hear about these orexinergic findings. Awareness is growing, but slowly. With more published data and more clinical trials, I expect greater comfort in prescribing these drugs.

One complexity we’re noticing in our receptor mapping work is that hypocretin receptors 1 and 2 change differently in Alzheimer disease. Some drugs target both, but if their effects cancel each other out, that may blunt efficacy. So it’s important to carefully evaluate which receptor subtype is being modulated. Still, the growing number of orexin-targeted drugs, along with better safety and delivery profiles, could open the door for meaningful interventions.

What biomarkers should we be developing that bridge sleep disorders and Alzheimer disease?

I love this question. Right now, our typical biomarkers—amyloid PET, tau PET, plasma phosphorylated tau—don’t really detect the disease at the earliest stages. They tend to pick it up only at moderate or severe stages. People often think they detect the disease “before” cognitive decline, but we need to remember that cognitive decline itself is already a late symptom. By the time memory problems appear, the brain has been under pathology for years.

So, what can we use earlier? I think EEG-based sleep–wake signatures are very promising. For example, daytime EEG is rarely used but could be highly informative. Even a one-hour, eyes-closed EEG may capture changes. There are already papers supporting this, but it hasn’t gone mainstream yet. With the development of home-based EEG devices, we could have an inexpensive, noninvasive way to detect early changes and to monitor drug effects.

Should sleep measures be included more systematically in Alzheimer clinical trials? Is this feasible on a large scale?

Yes, they absolutely should be included—and I think it’s very feasible. Clinical trials require very specific outcomes, and of course, sleep profiles vary based on comorbidities like sleep apnea. Some groundwork is needed to standardize what we’re measuring. But in terms of cost and feasibility, it’s much easier than PET scans, less invasive, and highly scalable.

We already send patients home with devices for studies. Data can be transmitted online, devices don’t need to be returned, and the process is relatively inexpensive. If implemented properly, sleep measures could become a sensitive and cost-effective biomarker for early Alzheimer disease, and a valuable tool for monitoring treatment response.

Transcript edited for clarity.