Across a large body of brain scans from more than ten thousand adults, a quiet pattern has come into view: the longer and heavier a person's history of smoking, the less brain tissue tends to remain. Researchers writing in Npj Dementia found that smokers carried smaller gray and white matter volumes than non-smokers, and that some of this loss seems to travel alongside body weight, though not all of it. The finding places tobacco and obesity among the familiar shadows cast over later life, yet it stops short of proving that one causes the other, or that the shrinkage will become dementia. What
Heavier Smoking Tracks With Brain Shrinkage in Alzheimer's-Linked Regions, and BMI May Play a Role
The more someone had smoked, the less brain tissue showed up on the scan.
Mark
What is the headline result, in plain terms?
Mimi
Across 10,134 adults, people with a smoking history had less gray and white matter than non-smokers, and more pack-years went with more shrinkage.
Mark
Why did the researchers bring BMI in?
Mimi
Smokers in the group also had higher BMI, and BMI rose modestly with pack-years. So they ran a second model that adjusted for it.
Luke
And what did that do to the result?
Mimi
The link weakened across 11 brain regions, in both significance and effect size. But the hippocampus, posterior cingulate and precuneus still showed significant atrophy.
Luke
So can we say BMI explains part of the damage?
Mimi
Not firmly. The authors say a mediating effect is possible but not definitively proven, and they call it exploratory.
Mark
Why can't the study say smoking causes the shrinkage?
Luke
It's cross-sectional, a single snapshot. It can't show what came first, and it lacks the timing needed to test mediation properly.
Mark
Does this mean the smokers had dementia?
Luke
No. There was no cognitive testing and no amyloid or tau data. Shrinkage in Alzheimer's-relevant regions is suggestive, not a diagnosis.
Mimi
The smokers were also older and had more diabetes and hypertension, so the adjustments matter. Only longitudinal studies can settle the BMI question.
O Pulso
- A study of 10,134 MRI scans has tied smoking pack-years to measurable shrinkage across much of the brain.
- Smokers showed significantly lower gray and white matter volumes than non-smokers, sharpening concern about tobacco's hidden toll.
- Adjusting for body mass index weakened the link in 11 regions, suggesting weight may be carrying part of the damage.
- Hippocampus, posterior cingulate and precuneus loss persisted, the areas most associated with Alzheimer's, keeping the alarm live.
- The design is cross-sectional with no cognitive or amyloid/tau data, so the next move is longitudinal tracking before any dementia verdict.
Across a large body of brain scans from more than ten thousand adults, a quiet pattern has come into view: the longer and heavier a person's history of smoking, the less brain tissue tends to remain. Researchers writing in Npj Dementia found that smokers carried smaller gray and white matter volumes than non-smokers, and that some of this loss seems to travel alongside body weight, though not all of it. The finding places tobacco and obesity among the familiar shadows cast over later life, yet it stops short of proving that one causes the other, or that the shrinkage will become dementia. What it offers is a clearer map of where to look next, and a reminder that the brain keeps a record of habits we often treat as private.
A research team working with Npj Dementia brought together MRI scans from 10,134 people and their smoking histories, and the picture that emerged was steady: the more someone had smoked, the less brain tissue appeared on the scan. Smokers had significantly lower gray and white matter volumes than those who had never smoked. The participants, aged 18 to 97 and drawn from four study sites, had each filled out a questionnaire about how many packs they smoked a day and for how many years. Those answers became pack-years, a standard measure of lifetime exposure.
The groups differed in more than tobacco. Smokers were more often women and Caucasian, older, heavier, and more likely to have type 2 diabetes or hypertension, so the analysis accounted for age, sex, study site, and then body mass index. Brain volumes were measured with FastSurfer, a well-validated deep-learning tool. When BMI entered the model, the smoking-related effects weakened across 11 regions, pointing to weight as a possible middle step in the chain. The authors were careful not to claim it had been proven.
Some loss remained even after that adjustment. The hippocampus, posterior cingulate and precuneus still showed atrophy in smokers, regions bound up with Alzheimer's disease. The stakes are real: dementia affects an estimated 47 million people worldwide, and smoking is credited with up to 14% of cases. The authors name smoking and obesity as targets for prevention, while acknowledging that a single snapshot cannot establish cause, and that without cognitive testing or amyloid and tau markers, shrinkage cannot yet be called dementia. Their next step is to follow people over time and to study white matter hyperintensities alongside pack-years.
A research team in Npj Dementia has put MRI scans from 10,134 people side by side with their smoking histories, and the pattern they found is consistent: the more someone had smoked, the less brain tissue showed up on the scan. Smokers had significantly lower gray matter and white matter volumes than non-smokers.
The participants came from four study sites and ranged in age from 18 to 97. Each had a non-contrast whole-body MRI. Beforehand, each filled out a questionnaire covering demographics, medical history and smoking, including how many packs they smoked per day and for how many years. Those answers were converted into pack-years, a standard measure of lifetime tobacco exposure. Anyone with a non-zero value went into the smoking group, which had 3,292 people. The 6,842 with zero pack-years formed the non-smoking group. The smoking group's mean was 11.93 pack-years.
The two groups differed in more than smoking. Compared with non-smokers, the smokers were more often women and Caucasian, were older, had higher BMI, and had higher rates of type 2 diabetes and hypertension. The analysis therefore adjusted for confounders. To measure brain volume, the team ran 3D T1 scans through FastSurfer, a widely validated deep-learning pipeline, and used a separate deep-learning model to segment intracranial volume. They then ran regressions in the smokers on pack-years and regional brain volumes in two models. Model 1 adjusted for age, sex and study site. Model 2 added BMI.
The BMI step is where the findings get more complicated. A Pearson correlation showed a modest positive relationship between higher BMI and more pack-years. When BMI entered the model, statistical significance and effect sizes weakened across 11 brain regions. The authors read this as a possible mediating role for BMI, but not a definitively proven one: part of the link between smoking and shrinkage may travel through body weight.
BMI did not explain everything. Even after adjustment, smokers still showed significant atrophy in several regions, including areas relevant to Alzheimer's disease: the hippocampus, the posterior cingulate and the precuneus.
The backdrop is a growing dementia burden. One recent estimate puts the number of people diagnosed worldwide at about 47 million, with roughly 10 million new cases projected each year. Smoking is credited with up to 14% of dementia cases globally, and toxins in cigarette smoke may drive neuroinflammation, a mechanism closely tied to Alzheimer's. Earlier meta-analyses linked smoking to higher dementia risk, but few large studies had connected smoking history and intensity directly to MRI-measured atrophy, a recognized biomarker of neurodegeneration.
The authors say their results point to smoking and obesity as two risk factors that could be targeted in dementia prevention. They also list limits. The study is cross-sectional, so it cannot establish causation, and it lacks the temporal resolution needed for reliable mediation or moderation testing. It included no cognitive testing and no Alzheimer's biomarkers such as amyloid or tau, so the shrinkage cannot be tied directly to dementia. The BMI finding is described as preliminary and exploratory.
What comes next is longitudinal work that follows people over time to test BMI's role. The authors also call for research on white matter hyperintensity volume and atrophy in relation to smoking history and pack-years.
Citações Notáveis
BMI may play a possible, but not definitively proven, mediating role in the link between smoking and reduced brain volume; the study's cross-sectional design cannot establish causation.— Study authors, Npj Dementia (as reported by News-Medical.net)
Quer a matéria completa?
Leia o original em News-Medical.net ↗
Science & Space