Obesity in mid life linked to dementia

“Obesity in middle age increases the risk of dementia,” reported The Guardian. The newspaper reported that compared to people of normal weight during middle age, the risk of dementia later in life was around 80% higher for those who had been overweight and around four times greater for those who had been obese.

This news story was based on a Swedish study of twins aged over 65 that had recorded their height and weight when they were around 40. Researchers looked at how the participants’ midlife weight related to their risk of various types of dementia at their current age (average 74 years).

The study found a 71% increased risk of dementia associated with being overweight during mid life and an almost four-fold increased risk associated with being obese at this time. However, when looking at how using twins had affected their results, the researchers also conclude that genetics as well as lifestyle may contribute to this effect.

Although this study will need further follow-up to fully understand the association, this study does support existing advice that maintaining a healthy weight in middle age – as through all life stages – may help reduce the risk of a variety of diseases.

Where did the story come from?

The study was carried out by researchers from the Karolinska Institute in Sweden and funded by the US National Institute on Aging, The Swedish Research Councils and Swedish Brain Power.
The study was published in the peer-reviewed medical journal Neurology.

The newspapers generally reported the research well, but there is some discrepancy between the risk figures reported in the papers and in the research article itself. This may reflect the papers rounding the figures up or down. The odds of a person who was overweight in mid life having Alzheimer’s in older age was 91% higher than a person of normal weight in mid life, as opposed to the 80% reported in the papers. There was a 71% increased risk of dementia of any type in people who were overweight at middle age compared to those who were of normal weight.

What kind of research was this?

This was a cohort study of twins that looked at whether there was an association between weight in middle age and the risk of developing dementia in older age. It also looked at how midlife weight was associated the risk of by Alzheimer’s disease, a specific form of dementia.

A cohort study is the best way to find factors that are associated with a health condition, but it cannot say whether those factors are a cause or consequence of the condition. Because this study compared twins, it was able to control for some genetic factors that may have influenced the participants’ risk of developing Alzheimer’s or other forms of dementia. The study also controls for the early life environment, which it assumed the twins would have shared.

What did the research involve?

The study recruited 8,534 participants from a nationwide twin registry in Sweden (the Swedish Twin Registry). The participants were twins, born in 1935 or earlier, and were aged over 65 years at the time of the current assessment (average 74 years). The study included both identical and non-identical pairs of twins.

The participants took part in a phone interview that screened for the most common diseases and included a brief assessment of their cognition. They were asked about their current height and weight, education, demographic factors, health status and behaviour, current and past diseases, and whether they used medications. As members of the twin registry they had also reported their height and weight when they were middle aged (when the participants were on average 43 years old), which provided the BMI data which the researchers used.

Participants who had scored poorly on the cognitive test during the initial screening interview were invited, along with their twin, to come in for a full clinical work-up. During these sessions, participants underwent validated diagnostic tests to assess whether they had Alzheimer’s disease or any other type of dementia.

The study also integrated information about the participants from the Inpatient Discharge Registry, which contained information about their history of health conditions such as diabetes, high blood pressure, heart disease and stroke.

The researchers analysed the results using a type of statistical technique similar to logistic regression. This technique looks at how much various factors influence the likelihood of a condition.

What were the basic results?

Out of the 8,534 participants, 350 (4.1%) had a dementia of some kind. Of these, 232 had Alzheimer’s disease and 74 had vascular dementia. Another 114 participants were diagnosed with ‘questionable dementia’.

Overall, 6% of the female participants and 4.6% of the male participants had dementia or questionable dementia. The researchers found that, compared to participants without dementia, twins with dementia were older, had a lower level of education and a lower current body mass index (BMI), but then had a higher BMI when they were middle aged. The people with dementia were more likely to have diabetes, heart disease and a past stroke. They found that 2,541 twins (29.8%) reported that they were overweight or obese when they were middle aged (a BMI of between 25 and 30 is considered to be overweight, with a BMI of over 30 considered to be obese).

The researchers calculated how much age, gender, education, high blood pressure, stroke and heart disease influenced the risk of developing dementia (or probable dementia) of any kind. They also separately calculated how these factors affected the risk of developing Alzheimer’s disease. They then adjusted their analysis for these factors so they could calculate the odds of the participants developing dementia if they were overweight or obese in middle age.

The researchers found that, compared to people of normal weight people (at middle age), overweight people (at middle age) were 71% more likely to have any kind of dementia in their current older age (odds ratio [OR], 1.71, 95% confidence interval [CI], 1.30 to 2.25). People who were obese during their middle age were almost four times more likely to have dementia of any type (OR 3.88, 95% CI 2.12 to 7.11).

For Alzheimer’s disease specifically, people who were overweight during middle age were 91% more likely to develop the condition than their counterparts with normal weight during middle age. People who were obese during middle age were 343% more likely to have Alzheimer’s disease in their current older age compared to people who were of normal weight during this period (OR 1.91, 95% CI 1.30 to 2.80, and 3.43, 95% CI 1.49 to 7.90, respectively).

While the study looked at twins in order to isolate the influence of lifestyle factors, the researchers performed a second analysis that examined whether the risk of twins developing dementia may have been linked because of their shared genetic influences on their weight and other factors that may affect their risk of dementia.

This analysis assessed the risk of dementia associated with mid-life BMI using data from twins where one twin had developed dementia and the other twin had not. They found that the calculation of risk was different compared to their calculation based on the study population as a whole. From this they conclude that genetic and family environmental factors may contribute to the association they had seen between midlife BMI and dementia.

How did the researchers interpret the results?

The researchers said that in their nationwide Swedish twin study, being overweight and obese at midlife increased the risk of dementia due to Alzheimer’s disease, vascular dementia or any other cause. This relationship was independent of diabetes over the course of life and diseases affecting the heart and circulation.

The researchers say their twin analysis suggests that familial factors such as genetics and early life environments contribute to the association between midlife weight and dementia in late life.


This cohort study has shown an association between increased weight during middle age and increased risk of having dementia, including Alzheimer’s disease, when the participants were, on average, 74 years of age. However, this association may not just be a result of lifestyle as genetic factors may also play a role. This study was large and well conducted but there are inherent limitations, some of which the researchers highlight.

  • The researchers compared people with dementia to people without the condition in a cohort of people over 65 years of age. It is possible that there is a difference in life expectancy in people with and without dementia, or with conditions that are linked to dementia (for example. cardiovascular disease). Therefore, by looking at only people who were alive at the average age of 74, they may not fully account for the effect of midlife weight on dementia as a proportion of people who may have had, or would have gone on to have dementia, may have already died prior to this study. This is especially important considering midlife obesity is associated with a lower life expectancy.
  • Out of a nationwide cohort of twins (including all twins in Sweden) participants volunteered to participate in the study. The researchers noted that those cohort members who participated were more likely to be older, less educated and female. This may affect how well this study population reflects what would be seen in the whole population.
  • The study used self-reported estimates of height and weight given by the participants during middle age. As with any self-reported measure there is likely to be some level of inaccuracy in their estimates.
  • The researchers point out that they used BMI as a measure of how much fat the participants carried, but said that BMI alone may not be an ideal representation of body composition. Measurements such as waist circumference would have been a useful addition.
  • They say that both obesity and Alzheimer’s disease are genetically influenced disorders, and that by comparing a twin case (person with dementia) with a twin control (person without dementia), their results may be distorted by the cases and controls being ‘overmatched’. Additionally, they grouped identical twins with non-identical twins meaning that in the non-identical twins genetic effects were not perfectly accounted for.

This study does provide support for the possible role of greater weight at mid life and the development of dementia. Although further follow-up prospective cohort studies are needed to fully understand this association, this study points towards the importance of maintaining a healthy weight during middle age – as in all life stages - to try and reduce the risk of a number of conditions, including dementia.

NHS Attribution