CALERIE Study: "Mild CR Slows Biological Aging"

[Saul]

Hi!

 

The link below is an article claiming that "CR reduces the rate of biological aging":

 

http://www.empr.com/news/aging-caloric-restriction-geroprotective-therapies/article/664118/

 

However, the study is the Calorie Study, which as we know really compares a group of obese individuals with a normal group; so it's basically a study showing the advantages of not being obese.

 

  --  Saul 

[kpfleger]

This new paper based on CALERIE was buried in Al's updates but I thought it worth calling out specifically:

Change in the Rate of Biological Aging in Response to Caloric Restriction: CALERIE Biobank Analysis
2017 Belsky et al. J Gerontol A Biol Sci Med Sci. DOI: https://doi.org/10.1093/gerona/glx096

(Full text PDF. Also lay-readable press blurb.]

 

For those not following the explosion of super-important work in the area of estimating "biological age" based on biomarkers, there has been quite a lot in recent years: some simple statistical models based on common blood tests, some fancy deep learning models based on the same (including the demo site aging.ai), and a lot of exciting work based on DNA methylation. It is already possible to predict mortality more accurately than by simply using statistics based on chronological age. In mice studies have progressed to testing biological age before and after known lifespan interventions (CR, rapamycin, etc.), finding that they move the biological age in the expected younger direction (or less quickly in the older direction at least). [bTW, just in case anyone is not already aware of it, a great blog to follow this biomarker/bio-age stuff and other aging research is fightaging.org.]

 

This new paper is the first I've noticed that applies biological age models to CALERIE data, and the first time I've seen the "bio-age test before and after a known lifespan intervention" paradigm applied to humans. The results are not surprising in the context of how people here already know CR affects the typical blood markers, but the size of the effect perhaps is, especially given that the CR achieved by CALERIE was only mild (mean 12%). Controls bio-aged 0.71yrs every 1yr, but CR group only 0.11yrs (which was not statistically different from zero). Attached is the key figure. The error bars are not tiny, but this still seems like quite a difference.

 

Plus, it's nice to see coverage for lay-readership that flat-out says CR slows aging.

 

-Karl

 

[Saul]

Yes; I think that I posted about this as well.

 

BUT, as Michael has noted, the Calorie study was not a study of CR versus ad lib -- it was a study of normal vs. obese.

 

So the results merely note the advantages of obesity avoidance.

 

  --  Saul

[Sibiriak]

 Dean Pomerleau argued:

 

In short, for middle-aged, healthy adults, mild CR sufficient to push one's BMI from ~25 to around ~22.5 was sufficient to reduce key markers of systemic inflammation, without compromising people's ability to mount an immune system response to a viral challenge.

 

Yet another result supporting my contention that the potential (but unproven) benefits of CR in people can be achieved, and most of the risks of CR avoided, via a healthy, obesity-avoiding diet, without serious calorie restriction.

 

 

More here:

 

https://www.crsociety.org/topic/11235-main-calerie-ii-results/?do=findComment&comment=17631

Edited May 29, 2017 by Sibiriak

[Michael R]

Karl, may I merge this thread in with Saul's?

[kpfleger]

Michael, sure merge away. I had thought that theory & science was the better forums but put it wherever you think is best.

 

Sibiriak, I read the prior thread before posting but thought this new paper was still a worthwhile and noteworthy addition to the set of papers about the CALERIE data, for the reasons I highlighted (which the prior thread did not really discuss).

 

Saul, great minds. :-) But the control group was not obese. Obese is BMI above 30. The paper says that CALERIE participants were specifically selected to not be obese. Overweight is the label for 25-30, but I wouldn't even label the control group as a whole overweight if the mean BMI was 25 (the exact threshold separating it from "normal"). Furthermore, while I'm not up-to-date on the exact distribution of BMIs across the population, the CALERIE participants were specifically selected to be healthy at the start of the study, and the paper notes that this caused the starting bio-age estimate to be a bit younger than chronological age attributing this to the fact that the CALERIE group were selected with a healthy filter in contrast to the group (NHANES) used to fit the bio-age model. Therefore the CALERIE group was healthier than the average population (and presumably it is a good bet that their mean BMI was lower than the mean BMI of the overall population. That all seems to indicate that it is reasonable to treat it as a study of mild CR at the BMI 22.5 mean level specifically with a nutrient dense diet (based on description I read of what the intervention group went through) vs. a reasonably average healthy person.

 

Also, the end of the paper discussed the dose-response in which they repeated their analysis splitting the CR group into the half with more severe CR (though after the fact and not randomly) and found that they showed a bigger effect, though not statistically significantly bigger than the weaker CR group. They don't say what the average % of CR achieved by more restrictive half was though.

[Sibiriak]

@kpfleger   I posted Dean P.'s remark to clarify the BMI range involved,  not to suggest that discussion of this new study would not be valuable.

 

Do you agree with Dean's previous general assessment, which I find straightforward and useful?

 

for middle-aged, healthy adults, mild CR sufficient to push one's BMI from ~25 to around ~22.5 was sufficient to reduce key markers of systemic inflammation, without compromising people's ability to mount an immune system response to a viral challenge. 

 

From the recent study:

 

The biomarkers from the CALERIE database were: serum albumin, alkaline phosphatase, C-reactive protein, total cholesterol, creatinine, glycated hemoglobin (estimated from serum glucose), systolic blood pressure, urea nitrogen, uric acid, and white blood cell count.

 

At first glance, it seems to me  to be a stretch to  describe that particular limited set of biomarkers  as a measure of "biological aging"--  but  I am not too familiar with the details of the "biological age" models.

Edited May 30, 2017 by Sibiriak

[Michael R]

@kpfleger   I posted Dean P.'s remark to clarify the BMI range involved,  not to suggest that discussion of this new study would not be valuable.

 

Do you agree with Dean's previous general assessment, which I find straightforward and useful?

 

for middle-aged, healthy adults, mild CR sufficient to push one's BMI from ~25 to around ~22.5 was sufficient to reduce key markers of systemic inflammation, without compromising people's ability to mount an immune system response to a viral challenge.

 

I would certainly agree with his general assessment that "mild CR sufficient to push one's BMI from ~25 to around ~22.5 was sufficient to reduce key markers of systemic inflammation" and diseases of aging— but, as I've just posted, they are quite substantially more reduced by more severe CR as exemplified in the CR Society cohort:

 

 

Most CR Society Cohort data from PMID 15096581; most CALERIE 2 data from PMID 26187233. Additional data from 6 mo to one-year CALERIE results in PMID 21841020*; from separate 2-year CALERIE 2 data in PMID 27410480^†; from Biosphere 2 in PMID 12023257^§; from separate CR Society Cohort report in PMID 16720655^¶.

 

From the recent study:

 

The biomarkers from the CALERIE database were: serum albumin, alkaline phosphatase, C-reactive protein, total cholesterol, creatinine, glycated hemoglobin (estimated from serum glucose), systolic blood pressure, urea nitrogen, uric acid, and white blood cell count.

 

At first glance, it seems to me  to be a stretch to  describe that particular limited set of biomarkers  as a measure of "biological aging"--  but  I am not too familiar with the details of the "biological age" models.

 

By instinct and reasoning, I tend to agree with you, although there is some rather good supporting empirical data behind it, being a better predictor of mortality and disability than chronological age (more on this later). Of note, several of the components of the score are improved more by more severe CR (see above), and probably more: I don't know this, but I assume that major disease risk factors that they don't report on in the main study results were probably not significantly moved by year 2 of CALERIE, and many of them were in the more rigorous CR Society Cohort. This is a good sign for more rigorous CR, whether or not this is truly a valid metric of biological age.

[Michael R]

Following up: unlike Aging.AI, InnerAge, Terry Grossman's nonsense, the various blood cell methylation tests, etc, previous versions of the Biological Age Score used in this study have some real value. It has been shown to be a better predictor of mortality than chronological age across a wide range of ages in a large cohort; is mechanistically tied to declining integrity of multiple organ systems (tho’ the blood-and-BP only version used here doesn’t have a good marker for pulmonary function) and is associated with being “less physically able, show[ing] cognitive decline and brain aging, self-reported worse health, and look[ing] older” (PMID 26150497)

 

That said, I'm far from convinced that this Biological Age Score really is a measure of "biological aging:" it's too closely linked to things already known to associate with "mere" disease risk and metabolic contributors thereto, and is rooted in a lot of things that can be moved quite readily in response to lifestyle and drugs. But its strengths as laid out above make it informative and useful, even if we think of it as a score of "accelerated secondary aging" or some such. Ie, it seems to have good criterion validity, but I'm skeptical of its construct validity and probably its content validity.

[InquilineKea]

So what is Klemera–Doubal method (KDM) Biological Age and why is no one more than 10 years younger?