Nicholas Brendon has died.

Nicholas Brendon, dead

Brendon, who played Xander Harris in Buffy the Vampire Slayer from 1997 to 2003, passed away at age 54. His family’s statement said he “passed in his sleep of natural causes.”

His death is the second of a major Buffy cast member in 13 months.

Michelle Trachtenberg, who joined Buffy as Dawn Summers, was 39 when she died in February 2025 from complications of diabetes mellitus.

Nicholas Brendon’s is the sixth death of a major WB cast member since 2024.

James Van Der Beek, who played the titular character in Dawson’s Creek, died in February 2026 age 48, after a battle with colorectal cancer.

Shannen Doherty, who played lead character Prue Halliwell on Charmed, died from breast cancer in July 2024, at 53.

Julian McMahon, who played major character Cole Turner on Charmed, died at age 56 in July 2025, from head and neck metastatic cancer.

Eric Dane, who played recurring character Jason Dean on Charmed, died in February 2026 at age 53 from respiratory failure resulting from ALS.

Counting family of major cast members, Buffy lead Anthony Stewart Head announced the sudden death of his longtime partner Sarah Fisher December 2025 at the age of 61.

The responses on social media mourned these unexpected deaths.

“I’m very concerned why are so many people passing away?!!”

“We lost 2 90’s stars and Mc Steamy in one month! wtf”

“Damn, this has been a brutal year so far for actors from late 90s/early 2000s shows on The WB. James Van Der Beek, Eric Dane, and now Nicholas Brendon” @Adaminhtowntx

“Jesus Christ… how my of my friends are going to die this year?!! Fffuuuccckkk” @sirjeremylondon

The hollowing of the WB generation

These deaths come when there is also a measurable increase in mortality in the age cohort that made up much of Buffy’s original audience, and the WB era generally.

The WB was the predecessor network to The CW, and in the late 90s and early 2000s aired popular teen and young adult dramas during the primetime block of 8 p.m. to 10 p.m. Three genre-defining shows from that era were Buffy the Vampire Slayer, Charmed, and Dawson’s Creek. The audiences for these shows were primarily younger Generation X and Elder Millennials, born between 1972-1988.

People who were teenagers and young adults in the late 1990s and early 2000, the era when these WB series were on the air, are now in their late 30s through early 50s. Since 2020 their mortality rates have been increasing. A 2025 JAMA Network Open study examining U.S. mortality among adults aged 25 to 44 from 1999 through 2023 found that mortality in 2023 was dramatically higher than expected based on pre-2011 trends. It quantified that gap as 71,124 unexpected deaths in 2023 alone.

Research led by the American Cancer Society and published in The Lancet Public Health reported that cancer incidence and mortality has increased in younger generations in 17 of 34 cancer types. For some groups, the gap was stark, with the 1990 birth cohort showing roughly two-to-three times higher incidence than the 1955 cohort for pancreatic, kidney and gastrointestinal cancers. Moreover, mortality increased for colorectal cancers, as well as uterine, gallbladder, testicular, and liver cancer.

Moreover, a March 2026 PNAS analysis reports that people born between 1970 and 1985 are experiencing worse mortality patterns than their predecessors. The trend covers major causes, including cardiovascular diseases and cancer. Particular concern was noted for cancers that have historically been less common for this age group, including colon cancer.

More than 20 million excess deaths since 2020

The increase in mortality for younger Gen X and Millennials is corroborated by an exponential rise in unexpected deaths across all age cohorts.

In demography, the term for unexpected deaths on a population scale is “excess mortality” or “excess deaths.” This measures how many more people have died than would be expected based on prior trends.

Since 2020, the COVID pandemic has officially contributed approximately 7.1 million deaths, according to the World Health Organization (WHO). However, it is becoming increasingly clear that this figure is a significant undercount. Scientific American has just published a piece about how, in the United States alone, COVID killed at least 150,000 more people in its first two years than are reflected in official figures.

And yet, zooming out to global unexpected deaths since 2020, The Economist counts between 20-35 million excess deaths from COVID.

For scale, the First World War killed 17 million people.

If more than 20 million people have died unexpectedly, and only 7 million are accounted for in the official COVID death toll, how have those 13+ million people died?

Part of the answer may be a limited understanding of COVID’s mechanisms and secondary effects, not only by individuals, but by healthcare professionals and institutions.

What SARS-CoV-2 does to the body

When people hear “COVID deaths,” many think of a respiratory virus that kills people in the acute phase of infection.

But SARS-CoV-2 is not a respiratory virus. COVID is a vascular disease.

COVID can have respiratory symptoms because the airway is a common transmission site, but the underlying disease is systemic. The virus infects the body’s blood vessels, specifically the lining which are called endothelium. The vascular system is not just the heart and major blood vessels, but spans the whole body. SARS-CoV-2 uses the endothelium as a superhighway to the body’s systems. Endothelial infection makes vessels form micro-clots, which reduce oxygen delivery, as well as create micro-injuries, inflammation and dysregulated immune responses across multiple organs. The downstream effects can create complications “from head to toe.”

People are perhaps most familiar with the long-term respiratory problems which can result from COVID infections. In 2022, Buffy‘s lead actor Sarah Michelle Gellar reported her first known COVID infection.

“I realize I’ve been really quiet on here. After two and a half years COVID finally got me. Thankfully I’m vaccinated and boosted,” she wrote on an Instagram Story. “But to those out there that say ‘it’s just a cold’ …maybe for some lucky people it is. But for this (relatively) young fit person, who has struggled with asthma and lung issues her entire life, that is not my experience.”

“Even with therapeutics and all my protocols it’s been tough. I know I’m on the road to recovery, but it’s certainly not been an easy road. I’ll be back soon (hopefully with super antibodies…even if just for a bit),” Gellar continued. “To quote a friend of mine – ‘I will wear a mask in my shower if that means I don’t get this again.’”

Another WB star, Alyssa Milano, who played Phoebe Halliwell in Charmed, went public with her Long COVID diagnosis in 2021. Her acute infection in April 2020 began with stomach issues and fatigue. The infection produced sequelae associated with Long COVID, a condition with 200+ possible symptoms.

“I have always had every single symptom imaginable, so every symptom that they list whether it be from acute COVID or long COVID, I have had. Shortness of breath, heart palpitations, brain fog, exhaustion at 4 o’clock in the evening, tingling in my hands and feet and just forgetfulness,”

The number of people experiencing Long COVID which meets diagnostic criteria is estimated to be between 5%-20% of all people. The World Health Organization (WHO) estimates that 1 in 20 people worldwide had Long COVID as of 2023. The CDC estimates that 1 in 5 people develop post-acute COVID sequelae.

The downstream effects of COVID infections can also lead to death.

How COVID causes premature deaths

Population studies have disaggregated excess deaths since 2020 by causes and age groups. A Lancet Regional Health—Europe paper examining post-2020 excess mortality reported that, in middle-aged adults (50–64), deaths involving cardiovascular diseases were 33% higher than expected.

Nicholas Brendon falls into this cohort.

The same study reported that, across all ages, excess deaths from all causes were higher than baseline, including cardiovascular diseases (12%), heart failure (20%), ischaemic heart disease (15%), acute respiratory infections (14%) and diabetes (13%).

The mechanisms underlying acute and post-acute COVID symptoms shed light on how infections can later lead to premature death.

For cardiovascular diseases, SARS-CoV-2 infection injures and inflames the endothelium, leaving arteries less able to deliver oxygen efficiently to heart muscle. In parallel, systemic inflammation can destabilize atherosclerotic plaques making the fibrous “cap” more likely to rupture. When a plaque ruptures, the body’s clotting system can rapidly form a thrombus that blocks a coronary artery and triggers a myocardial infarction (i.e. heart attack).

Nicholas Brendon also had clinical vulnerabilities, including a congenital heart defect and an addiction history. Repeated SARS-CoV-2 infection worsens the trajectory of those vulnerabilities. A Nature Medicine study found that people who survived the acute phase of COVID had a higher one-year risk of a wide range of cardiovascular diseases. With addiction history, a 2023 study reported that patients with alcohol use disorder who had a previous COVID infection had a significantly higher risk of incident cardiovascular diseases within 12 months than AUD patients without a COVID history. COVID presents an additive risk on existing vulnerabilities.

People are perhaps less aware of how COVID can lead to new-onset and aggressive cancers. James Van Der Beek, Shannen Doherty and Julian McMahon all died from cancers between the ages of 48-56.

Under normal conditions, “immune surveillance” means T cells and NK cells constantly identify and remove cells that look abnormal before they can grow into tumors. SARS-CoV-2 can create sustained immune dysregulation which reduces the efficiency of tumor surveillance while simultaneously creating inflammation. This is why cancers might present at a more advanced stage or behave more aggressively.

Michelle Trachtenberg died at the age of 39 from complications from diabetes.

SARS-CoV-2 can push people managing diabetes towards premature death by triggering inflammatory cytokines. These raise blood glucose and make the body’s tissues more insulin-resistant, so the same amount of insulin moves less glucose out of the bloodstream. In parallel, SARS-CoV-2 can also have direct and indirect effects on the pancreas itself, including β-cell dysfunction, which can reduce insulin secretion right when the body needs more. This creates a vicious cycle of higher glucose driving more inflammation and vascular stress.

Finally, Eric Dane died on February 19, 2026, at 53, with his official cause of death reported as respiratory failure and ALS listed as the underlying cause.

COVID has common and well-documented neurological sequelae. The underlying mechanism is that inflammatory signals can activate microglia (i.e. the brain’s immune cells) and disrupt the blood–brain barrier and the gliovascular unit. This increases neuroinflammation and makes neural tissue more vulnerable to secondary injury.

In ALS, progression is shaped by neuroinflammation, oxidative-stress and failures of protein homeostasis. A 2025 study focused on SARS-CoV-2 and TDP-43 found that inflammatory and oxidative signaling following a COVID infection, could push systems already near a threshold, such as vulnerable motor neurons and their supporting glia, toward accelerated degeneration.

Clinicians have documented rapid functional decline after SARS-CoV-2 infection in ALS patients who had previously been slowly progressive. Moreover, National ALS Registry mortality data have reported that motor neuron disease deaths were higher since the beginning of the pandemic than in the preceding years.

The end of COVID-19 reporting

The question is, if the weight of evidence points to an exponential increase in excess deaths since 2020, and the recent deaths of multiple WB stars are consistent with this shifted baseline, how are people not noticing? Or, if they are noticing, why are they shifting to the next topic?

One possible reason is a worldwide failure of government agencies to resource an effective long-term public health response to the ongoing COVID-19 pandemic.

The United States’ federal COVID-19 public health emergency ended on May 11, 2023. Globally, the World Health Organization ended COVID-19’s status as a Public Health Emergency of International Concern on May 5, 2023. And yet the WHO also emphasized that even though COVID no longer met the formal criteria for that emergency category, this did not mean it was no longer a public health threat.

That end of COVID as a public health emergency led to the scaling back of comprehensive testing and reporting. After May 2023, the U.S. moved from comprehensive case and lab reporting to fragmented surveillance systems like wastewater monitoring. In January 2026, a proposed funding cut would reduce CDC support for the national wastewater surveillance system from approximately $125 million to $25 million.

PCR testing has also been reduced, so many infections are not detected in the first place. At-home antigen tests aren’t reported into public systems, and those tests have high false negatives. Caltech found that at-home COVID tests had between 30% to 60% accuracy.

In this environment where monitoring for COVID has been substantially scaled back, one could be forgiven for assuming that COVID no longer presents a significant health risk.

How cognitive biases co-sign institutional silence

The systemic failure to address rising unexpected deaths from the secondary effects of COVID infections can be ratified on the individual level.

When public health agencies fail to address the ongoing COVID pandemic, authority bias can lead people to implicitly treat this silence as evidence that the danger has passed.

Even when increasing deaths are perceived, normalization bias can then turn this elevated harm level into the new normal. In conditions of sustained danger, humans adapt quickly, yet find it difficult to maintain chronic vigilance. For this reason, elevated harm can be perceived as normal because this frame is psychologically stabilizing but physically dangerous.

Once the harm is perceived as normal, motivated reasoning maintains this belief by selectively incorporating information that supports this worldview. People then downgrade, or wholly discount, contrary information to preserve the status quo.

Finally, system justification creates the belief that institutions are basically competent. Accepting that COVID has killed tens of millions more people than reported implies a magnitude of institutional failure that is psychologically costly to internalize. Even when people admit to a legitimacy crisis in other areas of governance, the belief that the system will alert people to health risks often remains.

The effect is circular reasoning. Because public health authorities do not communicate that there is an ongoing crisis of COVID deaths, this means that no crisis can possibly exist.

Let’s talk about how these cognitive biases may be working right now.

COVID’s bereavement crisis

Despite the official excess death figures and high-quality research on COVID’s mechanisms cited in this article, many readers may already have formulated reasons to discount the information.

One likely reason is causation. Because the epidemiological data does not prove that COVID directly caused Nicholas Brendon’s death in particular, the statistics don’t matter. Therefore, the 20+ million unexpected deaths and the settled knowledge about SARS-CoV-2’a multi-system damage can be wholly discounted as relevant to one’s own life.

Proving direct causation in a specific individual is an impossible standard.

The way deaths are recorded in the U.S. is not designed to trace a chain from infection months ago to vascular/inflammatory damage to a cardiovascular event today.” Most deaths are certified through clinical judgment on a death certificate, and what gets recorded is typically the immediate cause (for example, myocardial infarction, stroke, respiratory failure, etc.).

What‘s recorded and counted as a COVID death is a narrow range of cases where a clinician has a documented recent infection and a clinical picture that makes COVID feel obviously relevant to the immediate cause of death. This skews toward respiratory deaths during the acute stage of illness, rather than the deaths that occur later as heart attacks, organ failure, strokes, even pneumonias secondary to COVID infections.

Public health surveillance reinforces this, because they often operationalize COVID deaths within a time window after a confirmed positive test. The Council of State and Territorial Epidemiologists’ guidance, for example, includes COVID deaths among cases where death occurs within 30 days of the specimen collection used to define the case. This means a COVID death is only counted when there is a confirmed positive test and only within 30 days.

Mis-coding according to apparent cause of death led to undercounting in previous pandemics. A CDC study of early AIDS deaths found that in 1983-1986, before HIV/AIDS coding procedures were implemented, it was listed as the underlying cause in only 46% of deaths among people with AIDS. Many others were recorded as pneumonia or other infections.

Finally, even obtaining postmortem forensic proof is rare. A National Vital Statistics Report found the U.S. autopsy rate was only 7.4% in 2020. Even when a death triggers an autopsy, the standard aim is to identify the proximate event, like an arrhythmia, pulmonary embolus, or overdose, not reconstructing the underlying mechanisms. This would entail examining where a past SARS-CoV-2 infection contributed through microscopic endothelial injury, microthrombi, or inflammation.

Another possible reason for discounting the hard data and clinical evidence is clinical vulnerability. According to this logic, Nicholas Brendon had a heart defect and addiction history, so dying at 54 is expected. Even though, as discussed, research points to COVID as an additive risk and probable accelerant for existing vulnerabilities. Clinical vulnerabilities are also more common than many assume. The CDC’s estimates the prevalence at six in ten U.S. adults living with at least one chronic disease. Four in ten live with two or more. So Nicholas Brendon’s clinical vulnerabilities are not rare exceptions, but place him within 40-50% of all people.

Moreover, excess mortality, by definition, counts deaths from all causes minus the deaths expected based on prior trends. So it includes people with and without clinical vulnerabilities and it is not epidemiologically valid to discount deaths because the people who died were not perfectly healthy. Those vulnerabilities are part of the expected death baseline that the model already assumes.

Some may find no easy opening to discount the logic of the data, so instead will dismiss the credibility of whoever presents it. This may take the form of thinking that it’s inappropriate to speculate on a high-profile person’s tragic death, even if reported as epidemiological contextualization and not personal health information. Therefore, because the messenger is perceived as flawed, the hard data can be safely discounted.

First, this should go without saying, but treating discomfort as a proxy for illegitimacy of argument is a moral contamination fallacy. Epidemiological data do not become less true or less relevant because someone finds the conversation unpleasant.

It also helps to separate two things which are often conflated: private medical information and epidemiological context. Publishing a private person’s test results or medical records would be personal health information and therefore both ethically fraught and usually unverifiable. Contextualizing a publicly reported premature death of a public figure within documented population data queries whether that kind of death is becoming more probable in the population and in that age cohort, given what excess mortality and research are showing.

Moreover, the public narrative around a high-profile death is often shaped by what representatives choose to release. If the only acceptable public language is whatever passes through a PR filter, premature deaths that plausibly align with settled knowledge about excess death probabilities will continue to be described as normal.

Finally, and perhaps most insidiously, many people may simply no longer have the capacity or desire to internalize the implications of mass deaths from COVID. Even if it ultimately means their own premature deaths, those of everyone they love, and for that matter the whole cast of the WB’s peak era.

The direct and indirect effects of premature deaths has likely compromised the collective capacity to assimilate new information. There is simply a psychological limit to what people can internalize when loss becomes repeated. Acute grief can cause a sense of unreality, or dissociation, even when people appear to be functioning outwardly. When the loss is sudden, or when multiple losses accumulate, there is risk of prolonged grief disorder.

The conservative estimate of 20 million excess deaths since 2020 means that many more people are experiencing bereavement. A demography study published in PNAS estimated that each COVID-19 death leaves approximately nine close family members, including parents, children, siblings, grandparents. That means 180 million close kin bereavements. If you widen the lens from kin to close relationships, including close friends, Dunbar’s social network model estimates an inner circle of 15 people. This means 300 million people losing someone in their closest circle, or approximately the populations of the U.S.

The next category is the people supporting the bereaved. If those 300 million bereaved people each have an inner support circle of 15, that is 4.5 billion people. Which is about half of the world’s population experiencing some form of caregiving stress.

This may explain why, at population scale, repeated premature deaths can produce a blunting form of disengagement that makes the implications of ongoing mass mortality difficult to integrate.

You may be able to think of someone right now who tragically and unexpectedly died since 2020–and yet feel unsure of how to incorporate this new information.

Seizing the means of vividness bias

Vividness bias is the tendency for a single concrete, emotionally legible example to outweigh data that are more relevant, but difficult to picture.

Vividness bias is often misused by reactionaries to bypass logic, as the Buffy episode “Gingerbread” dramatizes. Buffy’s mom Joyce Summers takes news of the alleged death of two children to attempt to burn her own daughter at the stake.

Vividness bias can also be used intentionally to break through numbness and spur oneself to action.

So let’s imagine structural denial about excess mortality as the unspoken agreement of the adults in the town of Sunnydale, California circa the late-1990s to ignore all the vampires.

Though it was never a monster-of-the-week in Buffy the Vampire Slayer, motivated reasoning arguably enabled far more deaths than any season’s “big bad.”

In “Angel,” after the vampire Darla bites Buffy’s mom Joyce Summers, Joyce returns from the hospital with no memory of the incident saying, “The doctor said it looked like a barbecue fork. We don’t have a barbecue fork.” In “School Hard,” after a literal vampire attack at Sunnydale High, the police chief asks Principal Snyder “So, do you want the usual story? Gang-related? PCP?”

The adults of Sunnydale have practical reasons for their motivated reasoning. They have jobs, mortgages, and social networks tied to the assumption that the town is a normal California suburb—and not a hellmouth where their children are daily being preyed upon by vampires.

There is also a deeper epistemic stake in motivated reasoning, where people are invested not just in their practical interests, but in their worldview. The psychological costs of admitting that their town is overrun by vampires are so intolerable that, rather than shifting their assumptions, many people downgrade the evidence. Or attack the messenger.

Perhaps the most extreme example is in the episode “Normal Again”, when Buffy is first called to be a slayer and her parents involuntarily institutionalize her in inpatient psychiatric. She explains how she got out by recanting: “I was only there a couple of weeks. I stopped talking about it, and they let me go. Eventually… my parents just… forgot.”

As teenagers, we may have been puzzled as to why the adults of Sunnydale were in denial, so why do we find ourselves in a similar position?

Resisting motivated reasoning

If you notice yourself agreeing with the evidence about excess deaths and COVID’s multi-system effects in the abstract, but finding it hard to internalize, Season 3 of Buffy gives two images which can make the data emotionally vivid.

In “The Wish” (Episode 9), we experience an alternate Sunnydale in which Buffy never arrived. The city is overrun with vampires and the people who are left are trying to keep daily life going inside a town that‘s been hollowed out by mass death. The season ends with a counter-model. In “Graduation Day, Part Two” (Episode 22), the Mayor is planning his ascension, where he will become a demon and massacre student body.

At the climax, the students stop finally admit that there is something wrong in the town and unite to defeat the Mayor. As the entire graduating class reveals the weapons hidden under their graduation attire, Xander himself takes command:

“First wave!”

“Bowmen!”

“Everyone! Hand to hand!”

Buffy the COVID Slayer

In September 2025, Sarah Michelle Gellar posted behind-the-scenes images from the Buffy reboot. One photo showed her wearing a high-filtration mask.

This took courage.

Mask wearing has become so politicized that even a wealthy celebrity risks serious backlash. In the media industry, mask wearing can be professionally damaging. Formal COVID workplace rules have been rolled back and ongoing infection risk has become a liability issue. The joint Hollywood “Return to Work” COVID safety agreement, negotiated by SAG-AFTRA and other unions, expired on May 11, 2023, aligning with the end of the U.S. federal public health emergency.

As a result, the damage from COVID exposure on sets is currently being litigated. The family of Paul Woodward, a driver who was exposed to COVID while working on American Horror Story, sued for wrongful death. Additionally, actor Blake Lively’s lawsuit against director Justin Baldoni included allegations about on-set COVID exposure affecting her and her infant son. In that context, a high-profile actor wearing a mask on set is a public acknowledgement that COVID risk persists in media, even when the corporations that produce it are failing to protect its employees and resisting liability when harm inevitably occurs.

Nicholas Brendon died amid millions of premature deaths

Perhaps we can find it heartening that someone like Sarah Michelle Gellar is living the values of Buffy the Vampire Slayer, even in this small action of taking care of herself and others and having the courage to visibly refuse to collaborate with mass harm. Maybe the rest of us can take this as an impetus, not to conclude that Nicholas Brendon’s death was sad but unavoidable—as was Michelle Trachtenberg, Shannen Doherty, James Van Der Beek, Julian McMahon and Eric Dane—but instead recognizing that it happened in the context of tens of millions of premature deaths.

If you don’t know where to start with this knowledge, consider following Sarah Michelle Gellar’s lead.

Have the courage and care to wear an N95 mask.

Featured image and additional images via the Canary