Associations of antibodies against several infections with Alzheimer disease neuropathology: a prospective cohort study analysis

This prospective cohort study of the 1946 British birth cohort found no significant associations between serological measures of common infections and Alzheimer's disease neuropathology (plasma p-tau217 and A{beta}-PET status), suggesting that exposure to these pathogens does not directly drive core AD pathology, though potential interactions with APOE {varepsilon}4 carriage and education warrant further investigation.

The Gist

Imagine your brain is a bustling city. For decades, scientists have suspected that Alzheimer's disease is like a slow-motion traffic jam caused by two specific types of "trash" piling up in the streets: Amyloid plaques (sticky gum on the sidewalks) and Tau tangles (twisted wires in the power grid).

For a long time, there's been a theory that germs (bacteria and viruses) might be the ones throwing this trash around. The idea is that when your brain gets invaded by an infection, it tries to fight back by building a wall of Amyloid, but this wall accidentally traps the Tau wires, causing the city to gridlock.

The Big Question: Do common infections we all carry (like cold sores, stomach bugs, or chickenpox) actually cause this trash to pile up?

The Study: A 70-Year-Old Detective Story

Researchers in the UK decided to play detective. They looked at a group of people born in the same week in 1946. By the time the study started, these people were in their late 60s.

They did two main things:

1. The "Infection Report Card": They took blood samples to check for antibodies against 14 different germs. Think of antibodies as "wanted posters" your body keeps. If you have the poster, it means you've met that germ at some point in your life. They checked for:
   • Herpes viruses (cold sores, chickenpox, shingles).
   • Stomach bacteria (H. pylori).
   • Toxoplasmosis (the "cat litter" bug).
   • And others.
2. The "Brain Scan": About 7 years later, they checked the "trash levels" in the brain using two methods:
   • Blood Test: Measuring a specific protein (p-tau217) that acts like a smoke alarm for brain damage.
   • PET Scan: A special camera that takes a picture of the Amyloid "gum" in the brain.

What They Found: The "Ghost" in the Machine

The researchers expected to find that people with more "wanted posters" (more past infections) had more "trash" in their brains. They thought, "More germs = More brain damage."

But the city was surprisingly quiet.

• The Main Result: For almost every germ they checked, there was no connection to the brain trash. Having a history of cold sores, stomach bugs, or chickenpox didn't seem to make the Amyloid or Tau levels higher. It was like checking if people who had visited a specific park were more likely to have litter in their living rooms, and finding no link at all.
• The One Oddity: There was a tiny, strange signal regarding the Herpes Simplex 2 virus (a different type of cold sore virus). People who had it actually had slightly lower levels of the brain trash. The researchers admitted this was likely a fluke or a weird statistical glitch, not a real protective effect.
• The "Special Groups" Clue: The study looked at whether certain people were more vulnerable.
  • The Genetic Key: People with the APOE ε4 gene (a genetic risk factor for Alzheimer's) who also had the CMV virus (a common herpes virus) or H. pylori (stomach bug) showed slightly higher trash levels. It's as if the genetic key unlocked a door that let the germ do more damage, but only for that specific group.
  • The Education Factor: People with less education who had the stomach bug (H. pylori) showed a different pattern than those with more education. This suggests that life circumstances might change how our bodies react to germs.

The Verdict: No Smoking Gun (Yet)

The authors concluded that common infections probably aren't the main culprits in causing the early stages of Alzheimer's brain damage in the general population.

Why didn't they find a link?

• Maybe the germs aren't the trash collectors: The theory that germs trigger the trash pile-up might be wrong, or at least not the main driver.
• Maybe we missed the crime scene: The study looked at people around age 70. Maybe the germs do their damage much later, or maybe they only cause trouble if the infection is severe (like a bad fever), not just a quiet, lifelong presence.
• Maybe it's a team effort: Perhaps it takes a specific combination of three or four different germs working together to cause the problem, and the study didn't have enough data to catch that complex team-up.

The Takeaway

Think of Alzheimer's as a house fire. For years, we suspected that smoking (infections) was the cause. This study checked thousands of houses and found that, for most people, having a history of smoking didn't seem to correlate with the fire starting.

However, the study did find that for a specific type of house (those with a genetic weakness) and in specific neighborhoods (lower education), smoking might still be a risk factor.

The Bottom Line: Don't panic about your past colds or stomach bugs. This study suggests that simply having been exposed to these common germs doesn't mean you are destined to get Alzheimer's. The mystery of what causes the brain's "trash" to pile up remains unsolved, but we've ruled out a few suspects.

Technical Summary

1. Problem Statement and Objectives

There is growing interest in the potential link between common infections and the risk of Alzheimer's Disease (AD). Hypothesized mechanisms include the "antimicrobial protection hypothesis" (where Aβ aggregation serves as a defense against pathogens) and systemic inflammation triggering neuroinflammation. While some studies suggest interactions between specific pathogens (e.g., HSV1, CMV, VZV) and genetic risk factors (APOE ε4) or socioeconomic factors (education), evidence linking these infections to core subclinical AD neuropathology (amyloid-beta and tau deposition) remains scarce.

Objective: To investigate prospective associations between serological evidence of exposure to 14 common pathogens and AD neuropathology markers (plasma p-tau217 and Aβ-PET status) in a population-based birth cohort. The study also examined whether these associations are modified by APOE ε4 carriage and educational attainment.

2. Methodology

Study Design and Population:

• Cohort: The MRC National Survey of Health and Development (NSHD), a UK-based birth cohort born in 1946.
• Sample:
  • Exposure: Multiplex serology data from plasma collected at ages 60–64 years (N=1,793 with valid data).
  • Outcomes:
    • Primary: Plasma phosphorylated tau at threonine 217 (p-tau217) measured at ages 69–71 (N=1,356).
    • Secondary: Cerebral amyloid-beta status via PET imaging (Aβ-PET) using 18F-Florbetapir, measured at ages 69–71 (N=424).
• Follow-up: Approximately 7 years between exposure assessment and outcome measurement.

Exposure Assessment:

• Multiplex Serology: Quantified IgG antibody levels against 18 pathogens.
• Exposure Metrics:
  1. Serostatus: Binary (positive/negative) indicating lifetime exposure.
  2. Pathogen Burden Indices (PBIs): Cumulative counts of seropositive pathogens (Total and Neurotropic-specific).
  3. Seroreactivity Tertiles: Continuous antibody concentrations divided into tertiles to reflect recent infection/reactivation.
• Pathogens Included: 14 pathogens (e.g., HSV1, HSV2, CMV, VZV, EBV, H. pylori, T. gondii, etc.). HPV and KSHV were excluded due to low seroprevalence.

Statistical Analysis:

• Primary Outcome (p-tau217): Multivariable quantile regression at the median and 75th quantile (to capture tail effects in skewed distributions).
• Secondary Outcome (Aβ-PET): Multivariable logistic regression for amyloid positivity (Centiloid >10).
• Models:
  • Model 1: Adjusted for sex, age at serology, and age at outcome.
  • Model 2: + APOE ε4 carriage.
  • Model 3: + Educational attainment.
• Interactions: Tested for effect modification by APOE ε4, education, and specific pathogen interactions (HSV1 × CMV/VZV).
• Sensitivity: Analyses adjusted for BMI and eGFR; unweighted analyses; exclusion of imputed PET data.

3. Key Results

General Findings:

• No Strong Main Associations: The study found no notable or consistent associations between seropositivity for most common pathogens and higher levels of plasma p-tau217 or Aβ-PET positivity.
• Magnitude Comparison: The effect sizes of pathogen exposures were negligible compared to APOE ε4 carriage. For instance, APOE ε4 carriers had ~0.19 pg/mL higher p-tau217 at the 75th quantile, whereas the largest pathogen effect (CMV) was only ~0.06 pg/mL and attenuated after adjustment.

Specific Observations:

• HSV2: An unexpected finding showed an association between HSV2 seropositivity and lower p-tau217 at the 75th quantile, though this was inconsistent across models and sensitivity analyses.
• Aβ-PET: There were imprecise suggestions of higher odds of amyloid positivity for EBV (OR 2.84) and Toxoplasma gondii (OR 1.69), but confidence intervals were wide, and these did not correlate with p-tau217 findings.
• Pathogen Burden: No associations were found between total or neurotropic pathogen burden indices and either outcome.

Interaction Effects (Exploratory):

• APOE ε4 Carriage: Suggestive evidence that associations differed by genotype.
  • Among APOE ε4 carriers only, CMV seropositivity and higher anti-CMV pp28 antibodies were associated with higher median p-tau217.
  • H. pylori seropositivity was also associated with higher p-tau217 in APOE ε4 carriers.
  • Conversely, total pathogen burden was associated with lower p-tau217 in non-carriers.
• Education: H. pylori serostatus showed opposing associations with p-tau217 based on education level (positive association in lower education, negative in higher education).
• HSV1 Interactions: No evidence was found that VZV or CMV exposure modified the association between HSV1 and p-tau217.

4. Key Contributions

1. Biomarker Precision: This is one of the first large-scale studies to link broad serological infection profiles with preclinical AD biomarkers (p-tau217 and Aβ-PET) rather than clinical dementia diagnosis, reducing reverse causation bias.
2. Comprehensive Exposure: Utilized a multiplex serology panel to assess 14 pathogens simultaneously, including individual agents, cumulative burden, and seroreactivity intensity.
3. Effect Modification Analysis: Rigorously tested the hypothesis that infection effects are modified by APOE ε4 and education, providing nuanced data on potential subgroups where infection might play a role.
4. Null Evidence: Provides robust evidence against the hypothesis that common, latent infections (measured by IgG) are major drivers of core AD neuropathology aggregation by age 70 in the general population.

5. Significance and Limitations

Significance:

• The findings cast doubt on the theory that common, chronic infections are primary drivers of early AD neuropathology (amyloid and tau aggregation) in the general population.
• The lack of association with p-tau217 (a highly specific marker for AD) suggests that if infections contribute to AD, they may act through mechanisms other than direct amyloid/tau aggregation, or at later stages of the disease trajectory.
• The study highlights the need to focus on specific host-pathogen interactions (e.g., in APOE ε4 carriers) rather than general infection burden.

Limitations:

• Cross-sectional Outcome: While prospective, the outcome was measured at a single time point (~70 years), potentially missing later-stage disease progression.
• Exposure Measurement: Serology (IgG) indicates lifetime exposure but does not capture infection severity, timing, or reactivation frequency, which might be critical for pathogenicity.
• Selection Bias: The cohort is healthier and more educated than the general UK population, potentially underestimating associations if infections disproportionately affect those with lower socioeconomic status.
• Statistical Power: The sample size for Aβ-PET (N=424) was limited, reducing power to detect interactions or smaller effect sizes.
• Generalizability: The cohort consists of White British individuals, limiting applicability to other ethnic groups.

Conclusion:
The study concludes that exposures to many common infections, measured by circulating antibodies, are not strongly associated with the aggregation of core AD neuropathology measures at age 70. While some suggestive interactions with APOE ε4 and education warrant further investigation, the data do not support a broad causal role for these infections in the early pathogenesis of AD neuropathology.