Fat Metabolism in Formerly Obese Women: Part I Resting Metabolic Rate

Wow!  Many recent events and blog posts about the web have hit a lot of my deja vu "memory nerves".  This has had me going to the draft bin looking for what it is I'm reminded of.  This one was sparked by being linked to a series discussing hunger and satiety over at gnolls.org.  I'm going to link you to Part IV as that is where the paper I'm discussing in this post/series is cited, but J. Stanton  has links to other installments.  

So, most of this current post has been in my hopper since February!  But this article/study has actually been "on my plate" longer still, as the PDF was downloaded to my HD back in June 2010 after reading James Krieger's  piece on the ease of regain.   I'm going to break this up into three parts:  

• Part I dealing just with resting energy expenditure and its contribution to our overall "metabolic rates".  
• Part II will deal with the resting metabolism in terms of contributions of fat and carbohydrate
• Part III will address expenditure and substrate usage during and after exercise.  

Fat metabolism in formerly obese women
Ranneries, et.al.  AJP-Endo, 1998.

Overview:

An impaired fat oxidation has been implicated to play a role in the etiology of obesity, but it is unclear to what extent impaired fat mobilization from adipose tissue or oxidation of fat is responsible. The present study aimed to examine fat mobilization from adipose tissue and whole body fat oxidation stimulated by exercise in seven formerly obese women (FO) and eight matched controls (C).

Measured:

• Lipolysis in the periumbilical subcutaneous adipose tissue
• Whole body energy expenditure (EE)
• Substrate oxidation rates (glucose, fatty acids)

... fat mobilization both at rest and during exercise is intact in FO, whereas fat oxidation is subnormal despite higher circulation NEFA levels. The lower resting EE and the failure to use fat as fuel contribute to a positive fat balance and weight gain in FO subjects.

Subjects:

FO*  n=7:  Body weights in excess of 120% normal weight when obese.  Body weights reduced by conventional CRD to 110% normal weight, weight stable for at least 2 months.eight.  Weight losses were in excess of 10% bw and ranged from 15-20kg (33-44 lbs).

C  n=8:  Never obese subjects who were matched in terms of various parameters such as weight, BMI, body composition, etc.   These were well matched as can be seen in Table 1.

*Note:  Average weight of all subjects was around 140 lbs.  Which means that when obese, the FO's weighed in the 170-185 lb range.  When I put the average height and these weights into a standard BMI calculator using the average weight as baseline and adding 15 or 20 kg, the FO were not technically obese on average (BMI <30), or borderline.  I'm trying to figure a way that each of the 7 individuals could be technically obese (e.g. each with individual BMI of at least 30) given this information.  But I do think it is worth noting we weren't talking morbid obesity here and likely most of them were sub-200 lbs to start.  

All subjects were fasted overnight, and had not engaged in strenuous activity for three days prior to the study.  They also consumed a 15% protein, 30% fat, 55% carb standard diet (provided to them free of charge) during those three days.  They "sat quietly" for 15 minutes prior to the baseline measurement which is the only data I'll be discussing in this installment.

Differences in Resting Energy Expenditure, REE:

Note:  REE is often referred to as basal metabolic rate, BMR, or resting metabolic rate, RMR

The results:  

Control:  4.88 ± 0.74   (n=8)               Formerly Obese:  3.77 ± 1.07  (n=7)

The average REE of the formerly obese group was thus 23% less than that of the controls.   Let's do the calculation for basal energy expenditure per day.

 

Wow!  That's a pretty whopping (and depressing) difference of almost 400 cal/day.  This is perfect fodder for the HAES/Dr. Linda Bacon's of the world to demonstrate just how futile an effort losing weight is.  And don't be fooled, low carbers out there, you are not immune to this metabolic adaptation either.  Because if you've lost weight you've done it by establishing a sustained caloric deficit, even if you did so without conscious effort.  

The REE also seems to be rather high for the control group in light of average intakes of around 1550 cal/day for American women in the 1970's.  But I would remind readers that the controls were not lean, they were matched to the FO's who were still at 110% of normal weight.  

I would also note that while REE is generally ~60% of TDEE, this is perhaps a book-keeping error in that the energy we expend in activity -- e.g. walking for an hour -- includes that we would have expended at rest.  I note that the caloric levels of the diet provided to all subjects for the three days prior to the study was:

... computed from equations in which FFM and FM are used to predict 24-h energy expenditure (EE) based on previous measurements in our respiration chambers and multiplied by 1.12 to account for a higher free-living activity.

In other words, my REE calculations for calories per day are essentially equivalent to the TDEE (total daily energy expenditure) for a very sedentary person.  I would note (as bolded), that the factor used for "accounting for free-living activity" -- e.g. not being confined to a metabolic chamber -- amounts to only 12% increase.  Using that factor, we can surmise the TDEE's are 1879 and 1453 cal/day for C and FO respectively.  So a never obese woman can consume almost 1900 cal/day to weigh on average just under 140 lbs, while a formerly obese woman must consume under 1500 cal/day to maintain that same weight.  Still not comforting, but we -- especially us women -- have GOT to get away from formulaic expenditure estimates based on just total weight, and this notion that 1500 cal/day is some sort of starvation ration.  It is not!

Continuing with a look at the data, I note that the variability is also considerable.  Those "±" are for the standard deviation.  It is unfortunate that the individual values were not presented in such a small study, or at least the range of values (min,max) wasn't presented.  With such small samples trying to derive this sort of thing is an exercise in futility.    However one thing that did jump out to me was the considerable difference in the variability of the data with SD of 1.07 for FO (n=7) vs. only 0.74 for C (n=8).  

Why does she keep pointing out the n's here?  Because here is the formula for standard deviation for a sample.  The ∑ term is the "sum of the squares of the deviations".  To manually calculate the SD, as those of my age in science/engineering programs actually had to do innumerable times during the course of our studies, you first calculate the mean (that's the x with the bar over the top).  Then each individual deviation is the individual value (x) - mean.  Some of these will be positive and some negative, however when we square them all values become positive. Then we add them up, hence "sum of the squares of the deviations", let's use "d" for individual deviation.

  I note SD-squared is a statistical value itself, called the variance.

We obtain:   C:  3.83     FO:  6.87

And let's standardize that per data point: Giving:    C:  0.358        FO:  0.981

Now, let me be the first to point out that this isn't normal statistical practice because the investigators had the individual data points and could have provided them in tabular or graphic form.  But since we don't have that to work with, I'm just manipulating things backwards to give you an idea of the variability of the data sets.  Clearly we have much higher variability for the FO than the C.  This is somewhat evident in the differing SD's of 1.07 and 0.64 for FO & C respectively, but I believe expressed as an "average" of the 0.981 and 0.358 for FO & C respectively, it is more representative of the variation in REE amongst FO vs. never obese.  I'll address this in some thoughts at the end of this post.

There are some caveats mentioned by the authors that are easily overlooked when one focuses on tabulated data but are also important to the interpretation of the data.  I'll break it out in bullet point form with comments interspersed:

• REE did not correlate significantly with FFM (r = 0.54, P = 0.06), which could be attributed to the low variability in FFM among the subjects (REE = 4%). 

Because there is a well established correlation between fat free mass (FFM, aka lean body mass - LBM) and REE/RMR/BMR,  we would expect to see one in this study.  However the fact that we do not is easily attributed to the study design where subjects of similar weight and body weight distribution were included.

• A positive correlation existed between REE and VO_{2 max}  (r = 0.62, P = 0.02).
• Most of the difference in REE between FO and C could be accounted for by differences in VO_{2 max} 

The controls had higher VO_{2 max}  than the formerly obese.  The study said nothing about the exercise status of the groups, however it was stated that the weight loss was achieved through conventional energy restriction.  Therefore we can presume no formal exercise.   

• After adjustment for differences in both FFM and VO_{2 max} the group difference was attenuated

FO: 4.07 ± 0.65      C: 4.64 ± 0.81 kJ       (P = 0.19)

I often leave statistical notations like P-values out of posts to keep them more readable to the general audience.  However I'm including them here because this is probably the most important piece of information as regards metabolic adaptation in this study.  The term "adjusted for" means essentially controlling for those parameters.  When the expenditure was essentially "standardized" to some baseline FFM and VO_{2 max} , the difference between groups was more than just "attenuated" as the authors state, it is essentially statistically obliterated.  Common P-values for demonstrating statistical significance are 0.05, 0.02, 0.01, 0.001.  These correspond to the probabilities (multiply P-value times 100) that the difference could have been produced by random selection.   The smaller the P-value, the more sure the researcher is that the results they are reporting are due to the variable being investigated.  Here, a P-value of 0.19 is statistically untenable.  In other words, REE/FFM/VO_{2 max} was statistically the same in FO v. C although were this the presentation, there may have been some wording to the effect that they were similar but trended towards being lower in the formerly obese.

I would love to see the pooled data of the 15 subjects here.  It sounds like if we treated these women as individuals, the predictor of REE would be a combination of the long-established FFM correlation and VO_2max.

Implications for the Formerly Obese Individual and Long Term Maintenance ... Some Thoughts

Many who read here also read at many of the blogs out there that I do.  There's no need to name names here, but I think we can all agree that the comments of late have been swamped a bit more than usual with long n=1 accounts.  I'm certainly among those who are trying to reconcile the results of all of these types of studies with our personal experiences.  One tendency, that is all-too-human, but should be resisted at all costs, is to tend to believe the validity of those studies that agree with our personal experiences and write-off as flawed any of those that are in conflict with same.  I'm going to be doing a post on some realistic ways to view these studies to assist in your personal journey or dietary/lifestyle strategies and what may work better for you.  

Clearly here we see that the variability in the FO group (however it is looked at, but let's look at mean and SD of 3.77 ± 1.07) we see that as a percentage, the SD is 28% of the mean.  This is often called the CV = coefficient of variation.  We also revisit the change in mean REE.  That is if we take the C = 4.88 value as some sort of baseline energy expenditure, the 3.77 for the FO is 23% lower.  See where I'm going here?  I would not be surprised to find that one FO subject had a higher REE than all of the controls.  Note, I'm not speculating that this occurred, just that mathematically it is entirely possible and would not require some fluke. So if someone has lost a lot of weight and they are able to maintain eating 2500 cal/day, I say hooray for you and don't look a gift horse in the mouth, but your personal experience is not in conflict with the results here.  Also even if you've never been obese and you have to eat only 1500 cal/day or you start gaining, I say "I'm sorry", but you also don't render these results moot.  This is also part of the reason for my tedious somewhat unorthodox backwards stats in the main part of the post.  To show the variability in a different way that perhaps gets it across.

So what does the REE data from this study tell us, and how can we use it to our advantage?  Well for starters, it is yet another one in likely thousands of studies that demonstrate metabolic adaptation -- that sustained caloric restriction (intentional or otherwise) and substantial fat loss leads to reduced metabolic rate.  Not in everyone to the same extent, but almost invariably.  There are obesity clinics and the like where you can get your REE measured by a 10 min CO2 test.  If you are obese and can't lose, it would behoove you to have that measured so you know the starting point.  If you've hit a long plateau, you might not want to know the answer ... but I would also recommend such a test if you're frustrated by trying all sorts of things and not seeing more results.  I say "you may not want to know", because you may well be one of those folks for whom the REE drops more than average.  In any case, that "is what it is" and knowing what you're working with is better than flailing around in the dark.  

Studies like this also stress how you can't go back.  No no no.  You'll never ever be able to eat like you used to and maintain your more slender self.  And it may not be fair, but if, like most of us, your weight loss has slowed in later stages and plateaued out, you will have to keep doing what you're doing to maintain.  Your body has achieved its new energy balance point!  And, you may not ever be able to eat like you could have had you never gotten obese.  Lots of things in life are unfair .....

What if you're not happy with that?   Well first, before you let panic set in, I think we need to keep in mind a few things about this study.  Before we get too depressed about it's implications for the "damaged metabolism" let's consider the time frames here.  These women were only weight stable after significant losses for at least 2 months.  If they lost weight by conventional energy restriction we're talking a minimum of 4 months (2 lb/wk) to as long as 11 (1 lb/wk) ... and that's for steady losses with 100% compliance.  A mere minimum 8 weeks after "reaching goal" seems insufficient.   Also, again here the individual data would be so much more helpful  even though the numbers are too small to draw any meaningful inferences.  I would love to find a study following reduced obese over a few years measuring REE with maintenance.  I don't do accurate enough record keeping to really say, but I'm pretty sure my metabolism has rebounded a bit over the past years.  I would like to see the energy balance phase to be at least as long as the weight loss phase to compare REE with weight matched controls.  I suspect that the results would differ from what we see here.  

Still ... what have we always known is correlated with metabolic rate?  Lean mass.  What did we learn from this study that was somewhat hidden by the focus on FO vs. C?  That REE is highly correlated with VO_{2 max.} What is VO_{2 max}?  Well, for sure when one does a Google search on this the overwhelming links direct you to sports and fitness sites.  Most simply it is a measure of maximum oxygen usage and related to endurance.   How do you increase it?  Training -- increasing duration of (that old chronic cardio) and/or intensity of activity.  If anyone wants to continue their pipe dreaming that switching to some sort of "fat burning" metabolism will improve their mitochondrial function in spectacular fashion irrespective of activity, I'll not stop you.  But for the rest of you, exercising to increase VO_{2 max} seems like a more rational and, more importantly, effective approach.  Forget the calories burned in the exercise itself, calorie restriction seems to have an impact on VO_{2 max} that manifests itself as a reduced REE.   

I've discussed here before several studies analyzing weight reduction by exercise only, exercise plus diet, diet only, different types of exercise, etc. etc.  It seems clear to me that whoever you are, if you've become obese -- the more obese the worse the problem -- then following prolonged energy deficit accompanying weight loss your REE will be lower.  One tool in that toolbox to prevent regain?  Move more ... period.  And perhaps a strategy to maximize increasing VO_{2 max}  has a whole lot more merit to it than certain gurus would like you to believe.

In Part II  I'll revisit this exercise angle in the context of substrate partitioning in the resting state.