Is a Single Gene the "Culprit" Behind Alzheimer's? The True Meaning of "90%" Indicated by APOE

"90% of Alzheimer's Related to 'Single Gene'"—Is This Figure a Sign of Hope or a Source of Misunderstanding?

"Over 90% of Alzheimer's disease is linked to just one gene."

Such headlines have spread overseas and quickly went viral on social media. The focus of the discussion is the APOE (Apolipoprotein E) gene, which is involved in lipid metabolism. According to an analysis led by UCL (University College London), without the common types of APOE (ε3 and ε4), many cases of Alzheimer's disease might not have occurred.

However, this "90%" does not mean "the onset is determined solely by the gene." Rather, the intensity of the number makes misinterpretation more likely. Here, we will explain the "real impact" of this news by organizing the content of the research and the reactions (expectations, concerns, critiques) that spread on social media.

---

First, what is APOE: Three types of ε2, ε3, and ε4

APOE primarily has types (alleles) known as ε2, ε3, and ε4, and since people inherit one from each parent, there are six possible combinations (e.g., ε3/ε4). Traditionally, ε4 is considered to increase risk, ε2 is protective, and **ε3 is often treated as "neutral."**

However, this analysis challenges the understanding that "ε3 is neutral."
The research team re-established the ε2/ε2, considered the lowest risk, as the "baseline (low-risk baseline)" and estimated how much ε3 and ε4 contribute to the onset when viewed from that perspective.

---

What did the research do: Estimating "Population Attributable Fraction (PAF)" using four large datasets

The paper (npj Dementia, published January 9, 2026) estimates using the following four resources.

• UK Biobank: Approximately 170,000 people aged 60 and over, understanding AD/dementia through medical records such as diagnostic information

• FinnGen: Approximately 289,000 people aged 60 and over, similarly understood through medical records

• A4 Study: Approximately 4,415 asymptomatic individuals with amyloid PET (brain amyloid positive)

• ADGC: Approximately 5,007 AD cases and controls confirmed by autopsy

The central indicator is the Population Attributable Fraction (PAF).
In simple terms, it represents "how many cases could be estimated not to have occurred if the population were in a low-risk state."

---

Results: "72-93%" and "about 45%"—but there's a range

The conclusion is provocative.

• PAF of Alzheimer's Disease (AD): 71.5% to 92.7% (range depending on the dataset)

• Brain Amyloid Positive (A4): 85.4%

• All Dementia (UKB and FinnGen): 44.4% and 45.6%

UCL's explanation also summarizes that "72-93% of Alzheimer's disease and about 45% of all dementia are unlikely to occur without the influence of ε3/ε4."

The important point here is that "90% explained by a single gene ≠ 90% confirmed to be caused by the gene".

---

Why does misunderstanding occur? "Population Attributable Fraction (PAF)" is not the "proportion of cause"

This was one of the most common reactions on social media.

• "So, can genetic testing almost predict the future?"

• "Does this mean lifestyle doesn't matter?"

• "If my parents have Alzheimer's, am I doomed?"

However, experts caution that the expression in the reports is too strong. Comments gathered by the UK's Science Media Centre pointed out the risk of reading PAF as "causation itself." For example, the confusion between "susceptibility" and "cause", the framing of ε3 as "risk" making the numbers appear larger, and the fact that environmental factors (lifestyle) remain important "co-factors" were among the points raised.

In fact, UCL's own release also emphasizes that "it's not determined by APOE alone." Even with the highest risk ε4/ε4, the lifetime risk is estimated to be less than 70%, and many people with strong genetic predispositions do not develop the disease.

---

Why the push for "targeting APOE": The limits of amyloid drugs and the next target

The background to the attention on this study is the state of Alzheimer's treatment in recent years.

While drugs that remove amyloid have emerged, raising hopes for "disease modification," discussions continue over the magnitude of their effects, side effects, and cost-effectiveness. In the UK, there has been a context of cautious decision-making, and the Guardian reports on researchers' claims that "APOE should be targeted head-on as the next step" along with the challenges of implementation.

The paper also takes the position that, given the "limited effectiveness" of recent anti-amyloid treatments, diversifying targets is necessary.

---

However, the reality is challenging: The "almost everyone is a target" problem and the physiological function of APOE

APOE plays an important role in transporting lipids (such as cholesterol) in the brain and body.

Therefore, it's not as simple as "just eliminate it because it's harmful." The Guardian also points out the possibility of other side effects if APOE is completely inhibited, and the fact that over 99% of people have ε3/ε4 makes the sheer number of potential targets a challenge for preventive strategies.

In short, it's not just about creating an "effective drug," but also about designing
who, when, to what extent, and how to intervene (whether through gene editing, pathway inhibition, or risk stratification).

---

Reactions on social media: Hope, confusion, and the "how to read the numbers" debate

This topic also divided reactions on social media into three typical layers.

1) The "hopeful" layer: Research groups and medical influencers spreading the news

Accounts of Alzheimer's research support groups introduced the study with the message that "APOE3/4 could be involved in at least 70% of cases," emphasizing the significance of the research.
Additionally, medical and scientific communicators quoted and spread the paper's statement that "without ε3/ε4, many AD and dementia cases are less likely to occur."

2) The "fearful and personal" layer: Concerns about genetic testing and family history

The fear that "there are patients in the family" or "I might have APOE4" is more easily amplified the larger the numbers. The important point here is that **APOE is not a "definitive diagnosis" but a "factor that shifts probabilities."** The researchers also repeatedly state that "many people do not develop the disease even if they have the risk type."

3) The "hold on a minute" layer: Critiques on statistical phrasing

Expert comments gathered by the Science Media Centre raised concerns about the structure where calling ε3 a "risk" inflates the PAF and expressions that are easily misunderstood as "causes."

On LinkedIn, posts carefully checked assumptions, such as quoting reports and noting that "a world where everyone has ε2/ε2 is assumed."

---

"If genes are strong, does lifestyle mean nothing?"—The answer is the opposite

This is the biggest point of misunderstanding, but the conclusion is quite the opposite.
The stronger the genetic influence, the more the probability of onset shifts through the "multiplication" with environmental factors (lifestyle and social factors). UCL's release also touches on modifiable factors such as social isolation, high cholesterol, and smoking, stating that "there are multiple ways to reduce complex diseases."

Alzheimer's is not a binary choice of "100% genetic or 100% environmental."
On the "terrain" of genes, factors like sleep, exercise, blood pressure, glucose metabolism, smoking, and isolation—"weather"—combine to result in onset. This study shows that APOE is larger than imagined among that terrain, which is a reasonable interpretation.

---

Future focus: Targeting APOE is not just about "drugs"

The path this research opens is not just about "increasing new drug candidates."

• Designing clinical trials: Stratifying by APOE type to more easily find groups where effects appear

• Prioritizing preventive interventions: Focusing lifestyle interventions on high-risk groups (though ethical and discrimination discussions are also needed)

• Exploring biological pathways: Following how APOE is involved in amyloid processing, inflammation, and lipid metabolism with high resolution

On the other hand, powerful interventions like gene editing, given that the potential targets could be "almost everyone," inevitably bring discussions about the reality, cost, and risk of such medical approaches. It can be said that we have entered a phase where grounded design is as necessary as hope.

---

Conclusion: Don't be swayed by numbers; read it as a "map for the next research investment"

The number "90%" is strong. That's why on social media, both hope and fear amplify simultaneously.
However, what this study shows is not a narrative of "fate determined by genes," but a map for rewiring where to target for prevention and treatment.

Whether approaches targeting APOE truly become the "next breakthrough" depends on future research and clinical trials. At least for now, it can be said that because genetic factors are significant, interventions in lifestyle factors remain important—and will continue to be so.