Jump to content
Sign in to follow this  

A Case Study - BMI baseline in CR

Recommended Posts

Michael's new sticky on establishing a CR baseline got me curious. I have some old data, and am interested in perhaps this thread can serve an an example of his points illustrated as I have some decent data available. Though it is tempting to just read the bolded part, I think the full narrative is helpful in arriving at the best conclusion and I think readers will find those details quite interesting too.

Age: almost 43

Height: Currently about 5 foot 8.5 inches. In college a little more but still under 5 foot 9 inches. I use 5 foot 8.5 inches, but for individuals with a larger discrepancy I presume original ( as opposed to acquired with aging) height for BMI?

Build - medium - neither tall/lanky nor super-stocky

Historical weight -
Here's where things get interesting:

Skinny kid: dug up my "physical growth NCHS percentiles chart" which had datapoints from birth to age 14 and ranged from 5th to 30th percentile.

Weight starting / early college: 141 pounds (1994). However I was working out very aggressively ( I tend to do everything 100%) with significant muscle mass.

Mid-college: You could say I've always been fond of optimizing health of sorts (read: health nut). In ~1995 I read the book "stop aging now" by Jean Carper.

Incidentally, on the topic of accelerated biologic aging, years at the age of 41 as I was establishing a health baseline as I was introducing intermittent fasting, I tested my telomeres here: https://cellsciencesystems.com/patients/telomere/test-guide/and got back a "normal" telomere length of 94. I thought that might be a little low thinking initially 100 is the reference, but the one standard deviation reference range the green corresponding to one standard deviation extends below than above so perhaps I still managed to do better than average ( attached)? I am thinking of using to monitor rate of aging but am not sure on its reliability. This is despite an RMR that also looks like it is a bit elevated but not by much ( agree?)... I weight 5-10 pounds more now than I did then.



IAC, around this time I transition from intense weight lifting to moderate swimming and focusing on my studies with accompanying loss in muscle mass and weight. I did not follow a formal CR program but I do recall deliberately limiting how many bagels I had a day ( gotta love those early 90s low fat "health foods"! - lots of low fat 12 foot turkey subs, low-fat yogurt, skim and soy milk, boca burgers, orange juice, breakfast cereals and mega-salads), and sure enough the weight dropped and dropped to an all time college low of 118 pounds

End of College: at the advice of my doctor I worked to gain weight - It was always easy for me to lose weight and hard to gain weight ( takes many calories - see my cronometer post with 3500+ calroies a day), so I started eating really big, even for me, but especially going back to weight lifting and stabilized mostly in the 125-130 range sustained by increasingly intense workouts for most of the rest of college and followed by two higher readings at the very end of college as "149 and 155 pounds "with shoes" associated with the most aggressive weight training to build muscle mass in my life ( it was so noticeable someone based on seeing the chanes referred a friend to me to ask for advice how I did it ( intense methodical 1-2 hour weight resistance training using metric goals plus a very high protein in the diet - at the time thought healthiful - lots of 1 foot long turkey subs, nonfat yogurt, boca burgers & soymilk). These were taken in 1997, at the end of college.

After college: because my educational training was very intense, for years, and health was no longer on my radar. I was a memorization machine nonstop, and my interest in health pretty much was neglected and forgotten for many years. In the years that followed my weights was in the ?130s (looking up - will follow-up) and pretty much minimized weights again. I was no longer going to regular checkups and the next weight after 1997 I could find was in May 2005 at the age of 30: Weight = 139 pounds 9 ounces. No special diet, but still doing significant weight lifting (hence heavier).

Follow-up pre CR society weights ( around ages 36-39) with no restriction eating at all, but naturally losing weight relative to peak in college and early graduate years, as I was not aggressively working out through my lengthy post-college educational period as I had in the past:
2011: 131 pounds
2012: 125 pounds
2013: 128 pounds

Would not trust any numbers after 2013: rekindled interest in optimal health, reading on caloric restriction, etc. and though had never intended for formally practice CR do see it as a potential route to optimal health. Motivation has been reducing cancer and DM risk after some risk factors surfaced, hence start of intermittent fasting, etc.

Somewhere by around 2013-2015, with no small measure of influence also by an elevated fructokinase and digging up again an early cancer in my family with a history of aggressive malignancy (in the context of serial [5+] high CT-brain exposure in youth but that's another story) the interest in health was re-cultivated. So after over 20 years of suppressing my passion for optimal health, after completing rigorous training and then getting settled a few years practice, in my mid-life I have come full circle back to the health optimizer I always was :)
On 11/13/15 started my 8-hour window intermittent fasting.

The only other datapoint I have that may be useful is at the same time I got tested for RMR ( I weigh 5-10 pounds more now than the 115 pounds I weighed then), I also had the DEXA study showing a low % fat. It is lower than the figures I have gotten back via bioimpedence ( which had me as high as 10.4% fat using the Tantia Body Composition Analyzer at my gym recorded within a week of the DEXA value below, but I believe the DEXA is more accurate?
Here are the DEXA results:

Percent Body Fat 4.6 %Fat (Very Lean)
Total Fat Mass 5.5 lbs. of fat tissue
Non-Bone Lean Mass 107.5 lbs. non-bone lean
(water, organs, muscle)
Bone Mineral Content 5.3 lbs. bone mineral
Total Lean Mass 112.8 lbs total lean mass

Imagine what my % body fat was when I weighed >10 pounds lighter (though was not doing weight training at the time)? The lowest weight I have ever gone was was was a weight of 101 with a BMI just over 15 in April 2016 - you might guess I was experimenting with my diet, and reading on the forum :)

For the case study above, what would you deem the closest approximation to the baseline weight? Any thoughts on the teleomere/RMR also appreciated! :)

[ minor edits to improve accuracy on reviewing scanned old health records following post]

Edited by Mechanism

Share this post

Link to post
Share on other sites

The detailed post above illustrates some of the challenges of establishing a true baseline.  As you can see, My college weight was highly variable from lower weight due to some extent of caloric restriction through some periods, and higher weights associated with extensive muscle mass development from a very intense weight lifting regimen ( never had any period in my life with appreciable fat despite my medium build).  Other nuances such as DEXA and RMR results ( in the context of always requiring a ton of calories to sustain my weight, even during my underweight college years when moderating this a bit was a goal) are there as well.


On the high caloric intake of 3500 calories while having a low BMI - Interestingly enough I was found to have and rare larva of Strongyloides stercoralis on my January 2016 baseline that was not noted on the same test a month earlier - bizarre, though it did follow a trip to Asia where Ms. Mechanism & I waded in a traditional Japanese bathhouse.  This is a red herring however as voracious appetites + skinny body habitus run in my family including in the family member who passed away from an aggressive metastatic prostate cancer.  


Part but not all of that genetic tendency towards staying slim with large appetite is likely to be my fairly uncommon beneficial PPARG Pro12Ala variant  mutation - the basic science of it is fascinating - see https://blog.23andme.com/news/mouse-research-indicates-effect-of-a-snp-in-humans-might-depend-on-diet/- .  IAC, after treating the infection still need an exuberance number of calories to maintain a given BMI with signs of Crohns or steatorrhea on GIEffects testing , which is not surprising as I have always been skinny and/or lean with a large appetite.


IAC, the above vignette presents for me a quandary applying Michael's summary sticky note synopsis on how to arrive at a baseline CR weight:


1) I assume use traditional height rather than current height?... they are close together so I am leaning towards using my present height of 5 foot 8.5 as I was always just under 5 foot 9 inches in college ( though it says 5'9" on my license rounding up)


2) baseline to use prior to reducing it by 10-15% (I call this "baseline 1") 


3) after reducing "baseline 1" by 10-15% , I arrive at a new baseline = "baseline 2" which should be easy enough ( though for this special case using "baseline 1" would be interested whether you would reduce it by 10% or 15% to get "baseline 2)


4) Last step should be easy enough, I would presumably consume as many calories as required to keep my weight at X% lower than "baseline 2."  This should be easy enough to determine empirically and in my case would require more calories than most.


With all the idiosyncrasies noted above, I am much more interested in a quality comprehensive high confidence incorporating relevant nuances of the case, rather that a quick one, as I know this is a lot of data which was included to optimize accuracy for this higher weight variance perhaps somewhat unusual case.  


The response may also serve as a great real-life application operationalizing theory for tough real-life cases with all their subtle nuances and complexity.  I also happen to have more data than most.


The perspective and suggestion of weight value to use hearing from experienced CR adherents would be greatly appreciated!  

Edited by Mechanism

Share this post

Link to post
Share on other sites

Mechanism,  in  the 2015 topic "Do 'Naturally Thin' People Benefit Less from CR [etc.]?",  Dean P. provides another "case study"  (the more complicated formula involved was suggested by Shwet Shyamal--see thread for details):



Dean P. :  .... I'm not sure how actionable your formula is however. Take my case for example. Upon graduation from high school I weighed 128 lbs (BMI = 19.2). The follow year I gained 40 lbs of mostly muscle via weight lifting and swim team in college, so my BMI went to 25.2. I was healthy at both weights so which would you consider my "healthy early adult BMI", since it determines B, which in turn determines my optimal BMI to target, B'.


In the first case (healthy early adult BMI = 19.2), using 0.8 as the weighting factor as you suggest, according to your equation:


Let your healthy early adult BMI = B

=> Individual optimal BMI, B' = L^x * B^(1-x), where 0 < x < 1 depends upon what CR%, c you consider safe for humans on average               ... (1)


my optimal BMI would be:


B' = 17.5^0.8 * 19.2^(0.2) = 17.8  (= 119 lbs at my height  of 5' 8.5") 


In the second case (healthy early adult BMI = 25.2), according to your equation my optimal BMI would be:


B' = 17.5^0.8 * 25.2^(0.2) = 18.8 (= 126 lbs at my height)


My current weight (117lbs) is a bit below the first alternative, and substantially below the second. It's hard for me to see a sound basis for adjudicating between them.





It seems to me more reasonable to take Dean's weight prior to the 40lb gain as his "biological setpoint" weight.


Some guesswork seems unavoidable, and on top of that the 15% reduction from setpoint weight is, in M.R.'s definition of CR,  only a bare minimum , not a fixed optimal target.  In any case, a large number of other considerations and calculations would come into play during the course of CR practice   after the original deterimination of  one's theorectical "setpoint", arguably making the whole adventure as much  art as science.

Edited by Sibiriak

Share this post

Link to post
Share on other sites

Thank you Sibiriak for the link [ https://www.crsociety.org/topic/11271-do-naturally-thin-people-benefit-less-from-cr-good-non-technical-cr-science-overview/?do=findComment&comment=13052 ] to an interesting strategy for arriving at CR BMI.  Shwet’s formula has the advantage of ameliorating risk of too aggressive CR in individuals who are naturally thin.  OTOH it has the disadvantage of potentially underestimating the degree of CR required for optimal response.  Some of the assumptions, as he acknowledges and Dean reiterated are subjective, such as assuming the thinnest naturally thin have a natural BMI of 18.5, that these individuals should not restrict calories more than 5%, that 95% of the healthy population should have a natural ad libitum BMI between 18.5 and 24.5, that c=15% caloric restriction is the %CR human beings should perform on average.  Use of a geometric rather than arithmetic mean between minimum safe BMI and adult BMI is another.  He does however provide conceptual arguments for his approach and back-of-the envelope estimates, however, and the arbitrary nature of these estimates are both a weakness and strength.  A potential weakness in the subjectivity described is also a strength, in that none of these assumptions or values are set in stone, but rather and you can adjust the assumptions and equation accordingly. 


Michael’s sticky note discussion at https://www.crsociety.org/topic/12481-how-many-calories-should-i-eat-whats-my-goal-weight-whats-my-setpoint-what-cr-am-i/  is less subjective, mostly requiring assessing what %CR you think is optimal, and as with the Shwet approach determining your healthy early college BMI.  His approach does not explicitly limit %CR for the naturally thin, though obviously you can still do this by adjusting the %CR goal with his methodology.  A major advantage of his approach is it far more closely approximates the protocol followed by the trials on whose empirical data the operational definition of CR and its potential benefits (and liabilities) was arrived at.  If you are naturally thin, his approach can always be used with a lower %CR goal yielding similar results.

I still have the dilemma of which BMI to consider my optimal.  I think 118 was probably too low because I was aware of CR and probably was calorically restricted at that time.  141 ( or 149 or 155 when I graduated) was of course much to high as I was building (for me) heavy muscle. 

Interestingly enough, though I stabilized in the 125-130 range mostly, I also had a fair number of BMIs below that, so if I am slightly more conservative ( or aggressive) and use 123 pounds as my healthy young adult ad libitum weight, this corresponds to a BMI of 18.4 or so, and going by Shwet’s assumption a healthy weight is in the 18.5-24.5 range we can round up a tiny bit to 18.5


For Shwet’s formula, this corresponds to a CR BMI of

B' = 17.575^0.8 * 18.5^(0.2) = 9.90641 * 1.79240 = BMI = 17.756 or around 118 pounds


Had I used Shwet’s formula for a slightly higher BMI of 19.2 ( which corresponds to a weight of 128 pounds, which might be more fair since it looks like my weight stabilized in the 130s after losing muscle ), I would have arrived at a goal BMI of 17.8 corresponding to a weight of 119 pounds.  How do I know without pushing the numbers?  Funny enough Dean ( from that thread) is also 5 foot 8.5 inches and used this weight for his own calculation! ( though, did I miss something? Shwet used L = 17.575, Dean apparently used L = 17.5 ? ].


With Michael’s approach, since I was already skinny, my premise is that I do not need to “cut back about 10-15%” to arrive at the real healthy weight baseline as they had in the mouse or rat studies since I was not anywhere near the overweight category but rather was at the demarcation for underweight/ideal weight.  If I use BMI of 18.5, and assume only 5% CR that gets me to Shwet’s computation for L=17.575 with a weight of just under 118 pounds.  CR of 10% would correspond to a weight as low around 111 pounds, or around 115 pounds or so if my true baseline was 128 pounds as discussed in scenario 2 used for shwet’s formula above.


I’m curious Michael, what numbers would you use for the scenario above.  And would the % fat on DEXA or other variables in the original narrative influence the choice in any way?

Share this post

Link to post
Share on other sites

Hi Mechanism!


It's clear that you have a large RMR -- that is an inefficient metabolism; so you can eat a lot of food without gaining much weight.  In other words, you are "naturally skinny", like several of the best known members of the CR Society, such as Dean and Michael.


I'm the opposite -- I have a low RMR -- that is, an efficient metabolism.  A little extra food is rapidly translated into weight gain.


As Michael points out, BMI is not a good measure of degree of calorie restriction (if there is any good measure of that).


IMO, the best procedure is to reduce your calorie intake as much as you can, while maintaining adequate nutrition -- and keeping protein intake on the low side.  As time goes on, you may be able to tolerate further lowering of your caloric intake.



Being "naturally thin", severe calorie reduction does look difficult in your case -- not that it will mean that you're consuming too little food -- but that it might give you an anorexic look, even though you would possibly still eating more than enough food.


(It's a problem that I don't have.)


It would be interesting to see some studies on the effects of CR on genetically engineered rodents -- one genetically identical group with a high RMR, and other with a low RMR -- is there any difference in their response to CR?   


I don't know -- but my guess is "No".


  --  Saul

Share this post

Link to post
Share on other sites

That's one issue which discourages me in taking up a proper CR regimen. As soon as I cut the calories (especially the carbs), my weight drops, almost instantaneously. Past personal and dangerous experiments in starvation, which resulted in illness, have left me reluctant to allow a more than 5% bodyweight fluctuation from my ideal weight. 


Another fundamental issue, which MR did not address: is a calorie always a calorie? From the other thread on nuts and Atwater factors, it turns out it may be not. I have also my personal suspect that an abundance of fibers takes down the Atwater coefficients. Even though the modified USDA method is applied and other errors may supersede the factors errors.


Bottom line is that a calorie maybe a calorie, but there is some uncertainty in it, often unkown.


Nuts and fibery foods may yield less calories than the theoretical value from cronometer.

Edited by mccoy

Share this post

Link to post
Share on other sites

By the way, and I don't know if at this point a new thread on the atwater factors should be initiated, but I just verified that the NCCDB database in cronometer, food=walnuts, uses the traditional factors: 9,4,3.75. Whereas the USDA voice on food=black walnut yields a slightly lower calories value than using the traditional factors

Edited by mccoy

Share this post

Link to post
Share on other sites

To be correct, according to MR's guidelines which I just reviewed, absolute calories are not important, whereas relative calories are of interest, that is, it is enough to use a consistent database with a consistent method, since the reference is the weight, not the dietary energy itself.




Michael Rae:
If you are lucky enough to have had a clear, healthy 'setpoint' in your youth — a weight to which you tended to gravitate when you were in your early twenties, and that was within the healthy BMI range — take that as your baseline, and restrict Calories down to a level that keeps you at least 15% below that.


In my case, I remember that, in my early twenties, barring the periods where I ate too little or too much, my reference weight was about the same as now (56 years) = about 66 kg, with similar muscle mass and body composition.


Now, 66-15% means that, to be in CR, I should weigh no more than  56 kg. At least 10 kg less than now, preferably even less.


I'm not sure I'd feel good with that weight. Is it just mental reluctance and bias? 

Edited by mccoy

Share this post

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this