AnUser
#26
I think it’s bad whether it is muscle or fat to have high BMI, but there are waay worse things than to have a lot of muscle mass, it’s in fact probably very healthy and there is no issue at all.
To have a high BMI from muscle mass early in life is probably very healthy and wise, as there is an exponential decrease in muscle mass over time, which means when you start with high muscle mass you end up with high in older age as well.
It is also assosciated with a decreased risk of hip fractures to have higher BMI.
1 Like
AnUser
#27
True but most people who get down to BMI 18.5 will probably be skinny fat.
scta123
#28
Might be. But from my personal experience my lowest BMI was 19 just before surgery on my knee some 20+ years ago and my lean body mass was exactly the same as months later when my BMI was 25. And in my case it probably was the opposite.
1 Like
The masters paper controls for the effect of BMI independent of body shape factors. You can look at my discussion of the principal components in the original post. That means the conclusions on the hazards of BMI are independent of body shape (at least with the measurements in the NHANEs database, which are extensive). However, this means the principal components give a good indicator on how body shape effects long-term health.
So the two conclusions of the paper are
- BMI of 18.5-20 is possibly optimal, but BMI from <18.5 all the way to 25 are within error bars. So I think it’s fine, really. Also, maintaining a stable weight is good; time spent at a healthy BMI matters more than attaining that BMI at a particular test, and weight-loss due to disease states is not healthy, obviously.
- Modifiable factors of body shape can significantly affect within-BMI risk. Keep your body fat % low, muscle high (especially in your legs), and get your waist circumference as low as possible without sacrificing your health.
9 Likes
Alex
#30
With respect, you look fantastic. I cannot imagine you are unhealthy as you do not look overly thin at all.
2 Likes
Alex
#31
Another problem I have had with losing a lot of pounds is that sitting on hard surfaces is not nearly so comfortable!
But as with insulation, it is a small price to pay.
2 Likes
Ulf
#32
Waist circumference as low as possible for me means BMI at 20.
1 Like
blsm
#33
Thank you so much sir. It can be tricky to be objective about ourselves so I find honest input from others useful. I’m excited to do a dexa scan soon because although my BMI is technically borderline (fluctuates between 18.1-18.6) the mirror tells a different story and I really think my body composition is perfectly fine/healthy for me.
6 Likes
What do you all think about this conclusion?
“For this purpose, we derived an optimal value of BMI of 24 for males (26 for females) and optimal value of WHtR of 0.5 for males (0.46 for females).” from this study?
2 Likes
Apologies, here is the right study:
4 Likes
Ulf
#36
The analysis by Amelia1917 of the BMI paper is superb. I just big to differ in one detail. From her conclusion: “BMI of 18.5-20 is possibly optimal, but BMI from <18.5 all the way to 25 are within error bars”.
I would replace “possibly optimal” with “probably significantly superior”. It it true that the differences between the BMI ranges all the way to 25 are within error bars, but only barely. Estimating crudely from the graph in figure 5, I get an HR hazard ratio of all-cause mortality of 0.95 to 1.4 of the BMI 20 to 25 compared to 18.5 to 20, with HR of 1.2 falling in the middle.
A statistical curve will show a fat belly and thin tails. It should only be a 5 to max 10 % chance that HR falls in the 0.95 / 1.05 range i.e. BMI 20 - 25 being roughly equal in health outcome to 18.5 - 20. The bulk of the statistical probability will be in the HR 1.1 to 1.3 range i.e. 10 - 30 % higher all cause mortality for the 20 to 25 range compared to 18.5 - 20 being likely. . For me that is a significant difference.
This is a very crude calculation without any mathematial equation! My course in statistics from college days is rusty to say the least. Is there anyone who can put the bars of the right-hand part of figure 5 into an equation?
Amelia1917, might you have access to the underlying data? E.g. having the exact midpoint figure of HR that I guess to be 1.2 would help. As would knowing the exact figures of the upper and lower extremes of each bar.
1 Like
This is not surprising. This is a result of the body slowing down its metabolism to conserve energy. This is seen in animals on CR and humans that cut calories very low. It has less to do with body fat being insulating and more to do with thyroid hormones dropping due to low leptin levels. A good example of this is that if you fast for a few days. You will most likely notice that you feel colder one day into the fast even though you obviously can’t lose much body fat in that time. However leptin levels can temporarily drop almost 50% with short term fasting and that will lower heat production.
3 Likes
Here in Hong Kong, when it goes below 70 degrees the locals break out the winter coats. I still go around in short sleeved shirts. Now when it breaks below 60, I wear a light jacket and the locals dress as if it were going to snow. The local population has a very very low BMI average.
4 Likes
Asians need a lower bmi than Caucasians on average.
2 Likes
Neo
#40
Causality of fat mass index (FMI) on longevity seem to be linear and not U or J curve shaped in a person who avoids risk factors:
Findings
Linear MR analyses indicated a positive association between genetically predicted fat mass and all-cause mortality (HR 1.10, 95 % CI 1.08–1.12, P < 0.001). The association between FMI and all-cause mortality was manifested as J-shaped (HRs across FMI categories: 1.04, 1.00, 1.07, 1.21, 1.54), which was significantly modified by the number of low-risk lifestyle factors (P for interaction<0.001). When evaluating individual lifestyle factors, we observed a nonlinear relationship between FMI and all-cause mortality among participants who had high-risk lifestyle factors, while a linear relationship was observed among participants who had low-risk lifestyle factors, especially for those with adequate physical activity (HRs across FMI categories: 0.95, 1.00, 1.05, 1.17, 1.44) and who never smoked (0.96, 1.00, 1.03, 1.14, 1.51).
Interpretation
Genetically determined fat mass is causally and linearly associated with mortality. The J-shape association between anthropometric FMI and mortality is caused by high-risk lifestyle factors.
Association between fat mass and mortality: analysis of Mendelian randomization and lifestyle modification - ScienceDirect
@amelia1917 @AlexKChen has this paper been discussed in the CR(ON) community?
1 Like
Alex
#41
I am a BMI of 25.7. Given I have a 12% body fat percentage, I am fine with that.
It seems to me that waist size and a couple of other measures are better in aggregate than BMI as a risk measure in many circumstances. 5 Alternatives To Body Mass Index (BMI)
1 Like
Neo
#42
Just making sure it was clear that my last point above was exactly in the spirit of brining in complementary/better measures to/than BMI.
The paper above is a Mendelian randomization study and is based on FATT mass index (and not BMI).
2 Likes
I mean, how tf DO you control for confounders in BMI, given that fat-mass and muscle mass do opposite things and completely destroy the signal. It’s only that so much of the Western world is so unhealthy that you even have these associations. If there was only a way to go back into all the data and magically parcel out, but no, data is too messy.
1 Like
You control for it by using a mirror. If you are shredded, no need to worry about BMI.
3 Likes
Not to be a simp for Ryan Masters, but a lot of these questions are answered in the paper at the top of the thread.
Fat mass and muscle mass affect body shape differently, the NHANES data included an impressive amount of body shape measurements for its participants, and he controlled for them. The size of your waist, your thighs, your hips. Skin-fold measurements (a good marker for adiposity) in various parts of your arms and legs.
When you control for that, there is still a BMI effect on mortality. That is, even if you are shredded, it is worthwhile to think about BMI.
As for the paper @Neo shared, I don’t find the headline result particularly surprising (that genetically-determined adiposity is bad for you), but I do find that the J shaped curve in the general American populace that reduces to a linear curve when you control for lifestyle adds credence to the notion that being skinny is only good if you’re taking care of yourself! If you’re skinny and not engaging in healthy behaviors… perhaps a bad omen.
As for it being complimentary to/better than BMI, one thing the Masters paper didn’t really talk about was how much of the variance in risk can be determined by BMI, once you’ve controlled for everything else. He reported hazard ratios and error bars for different BMI ranges, which is good, but he had all the data he needed to similarly report the effects of adiposity. If you go through the crosstabs like I did at the top of the thread, it’s clear that the ABSI index is high-yield and that fits the existing literature but yea.
I think it’s clear that it’s less about being shredded, it’s more about being within a healthy weight range (pick the hazard ratio in the Masters paper you’re comfortable with; 20% is like… fine imo) and maintaining a low level of adiposity, with a preference for a smaller waist and stronger legs.
3 Likes
