AD: Middle-aged fat, thin in old age, dementia risk soared 288%!

At present, obesity has become the number one killer affecting human health
.
Almost all chronic cardiovascular and cerebrovascular diseases are inextricably linked to obesity, including dementia that seriously affects the healthy life expectancy of the elderly
.

However, as the saying goes, "There are no absolutes in the world"
.
With the progress of scientific research, the proposal of the "obesity paradox" seems to open up a way
of redemption for obesity, the "root of all evil".

The so-called "obesity paradox" means that for certain chronic diseases, people with a higher BMI index survive longer
.
Although numerous studies have pointed to a positive correlation between obesity and dementia risk, some studies published in recent years have shown that the "obesity paradox" also exists in the onset of dementia – older people with lower BMIs increase the risk of dementia [1-3].

In addition, the researchers observed that patients with late-life dementia showed significant weight loss ten years before diagnosis (or even earlier) compared to healthy controls [4,5].

You may have noticed that the phenomenon of "obesity paradox" in the occurrence of dementia is mostly observed
in the elderly population.
That is, there is a possibility that individuals with a decrease in BMI in midlife are more likely to develop dementia if BMI rises in middle life
.
To demonstrate this hypothesis, a recent large longitudinal population-based cohort study explored the association
between long-term BMI trajectories and dementia risk.

After nearly 40 years of follow-up with 2,405 middle-aged participants (30-50 years old), Boston University researchers found a statistically significant association between a first-increase and then a decreasing BMI trajectory and the onset of dementia (HR: 3.
88, 95% CI: 1.
38-10.
9).

The study was published in Alzheimer's & Dementia, the leading journal on neurological disorders
.

Screenshot of the first page of the paper

The study included 2,405 eligible participants from the Framingham Study (baseline age 30-50 years, BMI measurements, complete follow-up), and a total of 15,992 repeat BMI measurements were recorded after 39 years (1979-2018) of follow-up
.
The diagnosis of dementia was systematically assessed using an MMSE scale
score for all participants with cognitive decline.
Sources of information on cognitive decline include regular cohort follow-up, neurological and neuropsychological assessments, data on medical visits, and household surveys [4].

In the statistical method, the researchers first used a mixed effect model to fit the BMI repeat measurement data of individuals, and divided the test population into two groups according to the overall trend of BMI, namely BMI decline group (slope < 0) and non-BMI decline group (slope ≥ 0).

Next, for people with declining BMI, the researchers used trajectory analysis [6] to identify potential BMI trajectories and further grouped
the population according to the different BMI trajectories identified.

In the subjects of this study, the investigators identified three different BMI decline trajectories (Figure 1): early decline, which accounted for 11.
1% of the population; late decline, accounting for 76.
8% of the population; Increase and decline, accounting for 12.
1%
of the population.
Together with the non-BMI decline group, the overall population was divided into four different BMI change groups
.

Next, the researchers set dumb variables on four different BMI change groups and brought them into the Cox regression model, using the non-BMI decline group as the reference group to observe the association
between other BMI decline groups and the risk of dementia.
The model adjusted for age
, sex, education level, baseline overweight, and lifestyle factors.

Figure 1.
Different BMI descent trajectories

From the results of the Cox regression model (Figure 2), it is not difficult to find that there is no significant statistical association
between people with early BMI decline and late BMI decline and the occurrence of dementia.
However, the point is that compared to the non-BMI decline group, the risk of dementia increased by 288% in the BMI trajectory group that rose first and then fell!

Finally, the researchers performed sensitivity analysis to verify the stability
of the results.
Further adjustments for other comorbidities, baseline BMI levels, or consideration for the risk of death competition did not change the significance of the results
.

Figure 2.
Association of decreased BMI groups with the risk of developing dementia

The findings confirm our initial hypothesis that the pattern of change between rising BMI in middle life and decreasing BMI in old age is associated with
an increased risk of dementia.
The authors do not offer a discussion
of the underlying mechanisms of such a result.

In fact, there is still great controversy
in the scientific community about whether the "obesity paradox" really exists.
The so-called "magic is one foot higher, the road is one foot high", more and more large-scale epidemiological evidence has confirmed that the cardiovascular morbidity and mortality of obese people have increased significantly, thus pointing to the "obesity paradox"
.

Returning to our study, the authors also acknowledge that the "obesity paradox" phenomenon in the occurrence of dementia is likely due to the reversal of the observed causal relationship caused by the weight loss that occurs in patients with dementia during the asymptomatic preclinical period!

If we look closely at the three BMI decline trajectories in Figure 1, combined with the Cox regression results, it is not difficult to find that there is no significant association
between the two red and blue BMI trajectories that continue to decline and the onset of dementia.
In other words, if we don't have a BMI rise in midlife due to weight control failure, even if our weight continues to decline, it doesn't increase the risk
of dementia.

In addition, by observing the green trajectory, we can even draw this inference: for a middle-aged BMI that rises significantly, even if you try to reduce your BMI in later days, you still increase the risk
of dementia.
It's too late!

In short, the results of this study suggest that for the prevention of dementia, it is not enough to focus on the measurement of BMI at one time, but also to pay attention to the impact of long-term changes in BMI on the incidence of dementia
.
Considering the insidious occurrence of dementia and the serious health harm caused by dementia to the elderly population, should we take the long-term change trajectory of BMI as a new screening indicator for the prevention of dementia in the elderly?

References:

1.
Albanese E, Launer LJ, Egger M, et al.
Body mass index in midlife and dementia: systematic review andmeta-regression analysis of 589,649 men and women followed in longitudinal studies.
Alzheimers Dement (Amst).
2017; 8:165-178.
https://doi.
org/10.
1016/j.
dadm.
2017.
05.
007

2.
Fitzpatrick AL, Kuller LH, Lopez OL, et al.
Midlife and late-life obesity and the risk of dementia: cardiovascular health study.
Arch Neurol.
2009; 66(3):336-342.
https://doi.
org/10.
1001/archneurol.
2008.
582

3.
Li J, Ogrodnik M, Devine S, Auerbach S, Wolf PA, Au R.
Practical risk score for 5-, 10-, and 20-year prediction of dementia in elderly persons: Framingham Heart Study.
Alzheimers Dement.
2018; 14(1):35-42.
https://doi.
org/10.
1016/j.
jalz.
2017.
04.
013

4.
Knopman DS, Edland SD, Cha RH, Petersen RC, Rocca WA.
Incident dementia in women is preceded by weight loss by at least a decade.
Neurology.
2007; 69(8):739-746.
https://doi.
org/10.
1212/01.
wnl.
0000267661.
65586.
33

5.
Singh-Manoux A, Dugravot A, Shipley M, et al.
Obesity trajectories and risk of dementia: 28 years of follow-up in the Whitehall II Study.
Alzheimers Dement.
2018; 14(2):178-186.
https://doi.
org/10.
1016/j.
jalz.
2017.
06.
2637

6.
  Wills A, A Pérez, Wang J, et al.
Association between change in body mass index, unifiedParkinson’s disease rating scale scores, and survival among persons with Parkinson disease.
Jama Neurol.
Published online 2016:1.
https://doi.
org/10.
1001/jamaneurol.
2015.
4265