Tortoise Weight Formula- tBMI

Status
Not open for further replies.

Madkins007

Well-Known Member
Moderator
10 Year Member!
Joined
Feb 15, 2008
Messages
5,393
Location (City and/or State)
Nebraska
UPDATED/EDITED 1-5-2012

Susan Donoghue, DVM wrote an article about tortoise weights in 1997- "Nutritional status of tortoises using morphometrics to assess body condition".Vivarium magazine, Volume 8 Number 2, in which she offers is the helpful formula "SCLcm^3 x 0.191 = tWTgr", where SCLcm is 'Straight-line carapace length in centimeters' and tWTgr is 'Target weight in grams'. (See the end of the article for advice on getting a good SCL).

The data is based on 76 tortoises and box turtles representing 11 species, and works fairly well even for young tortoises.

We are going to make this formula more 'user friendly' by using it to calculate the tortoise's Body Mass Index, the BMI that all dieters are familiar with. Because this is tailored for torts, we are going to call it the tBMI.

tBMI= cWTcm/tWTgr
(or SCLcm^3 x 0.191)[/b]

EXAMPLE:
A tortoise with a straight-line carapace length (SCL) of 12.3 centimeters and a cWTgr (current weight in grams) of 349.
- The tWTgr is 355.425597 (12.3^3x0.191)
- 349/355.4= 0.982
- tBMI= 0.982

(Note- Google do all the work for you if you type in "[current weight in grams]/([SCL in centimeters]^3x0.191)=". The answer you get will be the tBMI.)

RESULTS:
You can determine if your tortoise is normal, dehydrated, or obese based on these results. Note- the figures used below are extrapolated from several sources and should be used as guidelines only!

- 0.66 or lower: There is less than a 2.5% chance that this tortoise is healthy at this weight. It is probably very dehydrated and/or underfed and should be seen by a vet.
- 0.66 to 0.83: There is only about a 15% chance that this tortoise is healthy at this weight. It is probably dehydrated and/or underfed and needs appropriate care.
- 0.83 to 1.00: This is a normal range, but statistically a little underweight, which may signal mild dehydration and/or being underfed. Review and correct cares and diet as appropriate.
- 1.00 to 1.16: This is a normal range, but statistically a little overweight, which may signal mild overfeeding. Review and correct cares and diet as appropriate.
-1.16 to 1.33: There is only about a 15% chance that this tortoise is healthy at this weight. It is probably at least somewhat obese and needs appropriate care.
1.33 or over: There is less than a 2.5% chance that this tortoise is healthy at this weight. It is probably obese and needs appropriate care.

RESOURCES:
- http://www.tortoiselibrary.com/ for articles on hydration/dehydration, a bibliography, and more
- BARTHEL, Tom “The Hydration Equation” Reptiles, July 2007
- DONOGHUE, Susan,1997. "Nutritional status of tortoises using morphometrics to assess body condition".Vivarium magazine, Volume 8 Number 2
- MADER, Douglas R., DVM, ed. "Reptile Medicine and Surgery". Saunders Elsevier, 2nd Edition 2006.

APPENDIX: Getting a good SCL reading.
The easiest way to get a good SCL without tools like calipers is to...
1. Place a metric ruler on the floor, '0' against the wall.
2. Set the tortoise on the ruler, head towards the wall, and position it so the shell touches the wall.
3. Place a block, box, etc. behind the tortoise, on the ruler, and move it so it touches the back of the tortoise's shell.
4. Move the tortoise and read the ruler at the block.

(Note- if the ruler has a space beside the '0' mark, you should either cut the ruler down so '0' is on the end, or remember to subtract the size of the gap from the final measurement... or cut the end off so '0' in on the end!)
 

Baoh

Well-Known Member
10 Year Member!
Joined
Nov 18, 2007
Messages
1,827
Location (City and/or State)
Florida
Since that does not determine or take into account body composition, I do not see how this is a reliable indicator of obesity or hydration status. An animal may have more total tissue mass without having a greater percentage of adipose tissue. Tissue densities would play heavily into the math of such and could easily place it into a portion of the continuum arbitrarily deemed less ideal. Livers of most, if not all, animals which have them are often subject to massive changes in mass in response to caloric status and nutritional composition. Additionally, it does not seem to account for significant differences related to shell shape morphometrics, with examples of these being yniphora in one and impressa in another. Width may be another parameter needing address. For females, there are matters of egg bearing. For some animals, we have found it wise to add mass prior to breeding seasons instead of trying to regain mass derived from the female's standard carrying weight due to concerns of potentially greater susceptibility to frailty or other illnesses which can be affected by compromised body mass post-laying. A breeding female sacrificing her resources for the sake of egg production may be adequately hydrated soon after laying and yet is still in need of a variety of nutrients (protein, vitamins, minerals, and lipids) that soaks would do little or nothing to augment. Any seasonal changes would also not be accounted for. If there are seasonal fluctuations that are a healthy part of a tortoises life, its values may achieve out-of-range status at the apex or nadir for any particular individual.

From a mathematical quality perspective, the scale itself is also skewed. I suspect that asymmetry could be given a pass, though, as it seems to point to a better-safe-than-sorry sort of thing than anything else.

I am not opposed to what it is trying to do, but I also do not see it achieving what it purports to do. BMI in humans, for example, is only relevant due to the sedentary nature of many modern (typically Western) populations. Athletes and people of more significant physical activity levels, for one subpopulation example, are deemed poor/inappropriate matches for it as an evaluation tool. When a male sulcata patrols the yard for a large enough part of the day or spends a good chunk of time modifying his burrow, he is hardly comparable to an equivalently sized female sitting on her laurels in the shade of a hibiscus and enjoying a snooze.
 

Madkins007

Well-Known Member
Moderator
10 Year Member!
Joined
Feb 15, 2008
Messages
5,393
Location (City and/or State)
Nebraska
I do not know enough about this to debate you on it, but it is based on clinical studies by Susan Donoghue, DVM, DACVN who is enough of a reptile nutritionist to have done the nutrition section of Dr. Mader's book "Reptile Medicine and Surgery". The study presented in the citation in the OP. I suspect that most of the issues raised in your first paragraph are why the basic formula is a bit 'wonky'- Target WTgr= SCLcm^3 x 0.191.

My only contribution to her research was to try to present it in a way that more of the members might be able to work with since so many do not seem to be interested in doing the math and conversions. The scale is mine, based on extrapolations from articles on dehydration and obesity, mostly cited except for one article on obesity in Sulcata feed a 'free access' Mazuri diet that I cannot seem to relocate. That is why there are so many 'weasel words' in the scale- "Use as guidelines only!"

While I admit the scale is not symmetrical, the main goal was simplicity. There is almost certainly a 'cushion' around 1.0 that is still healthy, but I don't know what it is and, as you suggest, I was trying to be cautious.

I used BMI as a rough model because I figured most people were at least somewhat familiar with it, and because it accurately describes what the formula is supped to do.

If you have a better method, idea, or suggestion, I would love to hear it! This is a frequently asked question and it would be wonderful to have a simpler to use tool.
 

Baoh

Well-Known Member
10 Year Member!
Joined
Nov 18, 2007
Messages
1,827
Location (City and/or State)
Florida
I see the proposed as being not especially different from the Jackson Ratio in terms of what it sets out to do. It would require the study of many individuals of each species in order to find the ranges appropriate to each species. Even then, that would just be a matter of mass per length and not tissue proportion (which is as important if not more important within a value set).

No method proposed would cover composition and that would be required to determine genuine obesity (rather than mere higher body weight). Perhaps DXA/DEXA or some form of densitometry such as hydrostatic weighing would fit that bill. It would still require examination of several animals from each species in order to establish normative values within a species, but the hydrostatic weighing would have the benefit of being able to be performed at home for tortoises of more reasonable size.

With enough volunteers, this would be an easy and inexpensive way for a researcher to make a name for himself/herself.
 

Zamric

Well-Known Member
10 Year Member!
Joined
Oct 29, 2011
Messages
3,301
Location (City and/or State)
The Crystal Unicorn
Hmmmm according to this formula, WalkingRock is 0.786 ... oh my .... I think he's dead... nope, I see movement! If this is accurate then WalkingRock must be in very serious trouble! He shows no signs of distress or I'm not as good at this as I thought! I thought 75 lbs was a good size but according to this chart, he should be over 95 lbs.
 

Madkins007

Well-Known Member
Moderator
10 Year Member!
Joined
Feb 15, 2008
Messages
5,393
Location (City and/or State)
Nebraska
Baoh said:
I see the proposed as being not especially different from the Jackson Ratio in terms of what it sets out to do. It would require the study of many individuals of each species in order to find the ranges appropriate to each species. Even then, that would just be a matter of mass per length and not tissue proportion (which is as important if not more important within a value set).

No method proposed would cover composition and that would be required to determine genuine obesity (rather than mere higher body weight). Perhaps DXA/DEXA or some form of densitometry such as hydrostatic weighing would fit that bill. It would still require examination of several animals from each species in order to establish normative values within a species, but the hydrostatic weighing would have the benefit of being able to be performed at home for tortoises of more reasonable size.

With enough volunteers, this would be an easy and inexpensive way for a researcher to make a name for himself/herself.

You may want to get and read the original study before continuing to discuss it. I have a copy of it and I see that the data originally used 42 animals from 9 species, and was able to expand to 76 animals representing 11 species (including Pancake Tortoises and Ornate Box Turtles). All were examined and weighed by a well-qualified veterinarian. I also notice that the study mentions Dr. Jackson's 1980 studies on Med. species and builds in part on it.

However, thanks to your comments, I see that I erred in the healthy weight range and need to refigure that. (She mentions 'standard deviations' quite a bit without putting a value on it. While I know what a standard deviation is, I am not as comfortable calculating it.)

I used a 10% variation for the 'markers', when I should have used a 30%. I'll correct the original post. Thank you for making me re-evaluate! Please let me know if this sounds better.
 

Baoh

Well-Known Member
10 Year Member!
Joined
Nov 18, 2007
Messages
1,827
Location (City and/or State)
Florida
Madkins007 said:
Baoh said:
I see the proposed as being not especially different from the Jackson Ratio in terms of what it sets out to do. It would require the study of many individuals of each species in order to find the ranges appropriate to each species. Even then, that would just be a matter of mass per length and not tissue proportion (which is as important if not more important within a value set).

No method proposed would cover composition and that would be required to determine genuine obesity (rather than mere higher body weight). Perhaps DXA/DEXA or some form of densitometry such as hydrostatic weighing would fit that bill. It would still require examination of several animals from each species in order to establish normative values within a species, but the hydrostatic weighing would have the benefit of being able to be performed at home for tortoises of more reasonable size.

With enough volunteers, this would be an easy and inexpensive way for a researcher to make a name for himself/herself.

You may want to get and read the original study before continuing to discuss it. I have a copy of it and I see that the data originally used 42 animals from 9 species, and was able to expand to 76 animals representing 11 species (including Pancake Tortoises and Ornate Box Turtles). All were examined and weighed by a well-qualified veterinarian. I also notice that the study mentions Dr. Jackson's 1980 studies on Med. species and builds in part on it.

However, thanks to your comments, I see that I erred in the healthy weight range and need to refigure that. (She mentions 'standard deviations' quite a bit without putting a value on it. While I know what a standard deviation is, I am not as comfortable calculating it.)

I used a 10% variation for the 'markers', when I should have used a 30%. I'll correct the original post. Thank you for making me re-evaluate! Please let me know if this sounds better.

Glad it spurred something useful to you. The ranges seemed a bit constricted to me, too. Seventy-six animals spread across eleven species is not a large sample size, though. I would go with a hundred per species, if available, for improved statistical power. For instance, some sulcatas are relatively high-domed and many are less pronounced. If I had a heterogeneous mixture of these in a small sample size, my data quality could be severely compromised. Especially so if favoring one extreme by happenstance. It is part of why I enroll thousands of patients in my trials instead of a handful. One or a few freaks can send the whole exercise to Hades.

Still, regardless of qualifications, reference, or animals examined, this does not determine composition, which is the essence of my most central point. Measuring a different dimension a greater number of times does not change that.

I consider the effort a positive step as long as it is not over-interpreted beyond what it is capable of determining. I also am not a fan of taking a one-dimensional measurement and just popping it in as a three-dimensional value (regardless of correction factor).
 

Redfoot NERD

Well-Known Member
10 Year Member!
Joined
Dec 5, 2007
Messages
3,666
Location (City and/or State)
Tennessee
Zamric said:
Hmmmm according to this formula, WalkingRock is 0.786 ... oh my .... I think he's dead... nope, I see movement! If this is accurate then WalkingRock must be in very serious trouble! He shows no signs of distress or I'm not as good at this as I thought! I thought 75 lbs was a good size but according to this chart, he should be over 95 lbs.

36_12_6.gif
... I luv it "Z" .. how do you make tortoise keeping a science??? You might be able to chart activity on a regular basis.. but everytime you get something that they do the same 2-3 times during a short time frame they go and change on you and it throws it all out of sequence! How is it possible to establish a given when you can't totally control ALL of the parameters?

Like a fellow redfoot breeder friend of mine said - "Ya don't need to go to a library to raise a redfoot tortoise".
 

Zamric

Well-Known Member
10 Year Member!
Joined
Oct 29, 2011
Messages
3,301
Location (City and/or State)
The Crystal Unicorn
I was just seein' if the science fit into my reality... It missed by a long shot! WR is healthy as a ....well a Tortoise!
 

Madkins007

Well-Known Member
Moderator
10 Year Member!
Joined
Feb 15, 2008
Messages
5,393
Location (City and/or State)
Nebraska
Zamric said:
I was just seein' if the science fit into my reality... It missed by a long shot! WR is healthy as a ....well a Tortoise!

At .78, he is one standard deviation from the norm. Based on the recalculated figures, that means there is about a 15% chance it is a normal body weight for him.

Just out of curiosity, how do you know he is not underweight?

Redfoot NERD said:
] ... I luv it "Z" .. how do you make tortoise keeping a science??? You might be able to chart activity on a regular basis.. but everytime you get something that they do the same 2-3 times during a short time frame they go and change on you and it throws it all out of sequence! How is it possible to establish a given when you can't totally control ALL of the parameters?

Like a fellow redfoot breeder friend of mine said - "Ya don't need to go to a library to raise a redfoot tortoise".

Making anything a science is quite easy- you do something, observe and react to the results, and repeat. Making it measurable and repeatable is nice, and makes thing neater and easier to pass on, but the science part is in the observations and reactions.

You are also being hypocritical about this... again. You use science all the time. Your preferred incubation temps, your understanding of the average clutch sizes and hatching times, your understanding of growth rates, your use of your diet plan... that is all science.

Going a different route- why should raising tortoises be any different than raising horses, hamsters, or human babies? Almost all animals and people we raise are measured and analysed so we know the growth curves, diet plans, etc. Science.

That does not mean we have always locked it down 100% or that we cannot change our minds. One of the great things about science is that it is can change when new data or results appear. Only idiots do not change their minds or methods in the face of good data.

As for your friend, about the library? That sounds like something an idiot would say. A more rational person would take the time to do some basic research about any animal they wanted to raise and were unfamiliar with. You know- like we tell people to do all the time here or on any other hobby forum.

The library may not be the BEST source for info on a given specialty, but it is often a great start, and one must presume that the 'friend' intended the anti-science comment to include web sites, etc. After all- isn't YOUR SITE basically a library?????

Baoh said:
Glad it spurred something useful to you. The ranges seemed a bit constricted to me, too. Seventy-six animals spread across eleven species is not a large sample size, though. I would go with a hundred per species, if available, for improved statistical power. For instance, some sulcatas are relatively high-domed and many are less pronounced. If I had a heterogeneous mixture of these in a small sample size, my data quality could be severely compromised. Especially so if favoring one extreme by happenstance. It is part of why I enroll thousands of patients in my trials instead of a handful. One or a few freaks can send the whole exercise to Hades.

Still, regardless of qualifications, reference, or animals examined, this does not determine composition, which is the essence of my most central point. Measuring a different dimension a greater number of times does not change that.

I consider the effort a positive step as long as it is not over-interpreted beyond what it is capable of determining. I also am not a fan of taking a one-dimensional measurement and just popping it in as a three-dimensional value (regardless of correction factor).

Like I said, I am really not qualified to debate Dr. Donoghue's study, assumptions, etc.

That said- what do you mean by composition? I assume it is something like the ratios of various tissues of differing densities- bone vs. muscle vs. voids, etc. but wouldn't the basic ratios be about the same in healthy tortoises? I would also have to assume that this very topic is why she made such a big deal about the standard deviations. In one example, she mentions a tortoise that is 2 standard deviations from the norm and that there is only a 1 in 40 chance that it is the normal weight for that individual. I tried to better reflect that idea in the re-writing of the OP.

And, if I may say, all in good humor, your comment about "I also am not a fan of taking a one-dimensional measurement and just popping it in as a three-dimensional value (regardless of correction factor)." is pretty much exactly how I feel about 'average weight/height charts' as an overweight sorta guy! :)
 

Zamric

Well-Known Member
10 Year Member!
Joined
Oct 29, 2011
Messages
3,301
Location (City and/or State)
The Crystal Unicorn
I dont know that's he's not underweight. I know he eats his fill of fresh grass and fresh veggies and more Mazuri than than you might think, I know he walks his butt off!

...but he could be under weight... he does show signs of dehydration SOMETIMES (foamy eyes) but that always dissapers after a good soak. After all, he does fall within the parameters even if it is the low end.
 

Baoh

Well-Known Member
10 Year Member!
Joined
Nov 18, 2007
Messages
1,827
Location (City and/or State)
Florida
Yes, differing tissues with their differing densities. I would not peg the tissue proportions to be the same in a healthy animal or across healthy because health is a matter of significant range. A short sedentary person at 25% bodyfat with decent blood values, a higher weight class shotputter, and a low-weight marathon runner may all be considered healthy humans and all have very different tissue distributions. I hovered around 119-125 when I was a kickboxer. I sometimes force myself up into the 250s and semi-lean as a weight lifting enthusiast. I typically sit at 190-210 and am leaner still by playing with my diet for the fun of it. All three occurred at the same height and all three had/have associated values saying I had/have excellent health and that is in just one individual case.

All good times. While there may be quite a number of ways to do things incorrectly, I think there are also multiple approaches to doing some things correctly.
 

Madkins007

Well-Known Member
Moderator
10 Year Member!
Joined
Feb 15, 2008
Messages
5,393
Location (City and/or State)
Nebraska
Which sounds like a perfect description of the purpose of 'standard deviations'. While a trained athlete will have a very different BMI than a more typical person, 85% (or whatever the ratio would be) of 'healthy' people of a given height will be a certain weight. If your BMI is two standard deviations from the norm, there is still a chance (15% if I recall correctly) that you are healthy, but there is a statistically significant chance that there is something wrong instead.

By the way- as a gravitationally enhanced person, the idea of 'playing with a diet for the fun of it' marks you as a very sick person. :)
 

SulcataSquirt

Member
5 Year Member
Joined
Jun 13, 2011
Messages
878
Location (City and/or State)
Missouri
I just tried out the formula and it was only off by two grams! interested to do it again in a few months.
 

Baoh

Well-Known Member
10 Year Member!
Joined
Nov 18, 2007
Messages
1,827
Location (City and/or State)
Florida
Madkins007 said:
Which sounds like a perfect description of the purpose of 'standard deviations'. While a trained athlete will have a very different BMI than a more typical person, 85% (or whatever the ratio would be) of 'healthy' people of a given height will be a certain weight. If your BMI is two standard deviations from the norm, there is still a chance (15% if I recall correctly) that you are healthy, but there is a statistically significant chance that there is something wrong instead.

By the way- as a gravitationally enhanced person, the idea of 'playing with a diet for the fun of it' marks you as a very sick person. :)

Correct, however, the sample sizes given do not lend enough statistical power to establish readily defensible norms. Plus, when applied to humans, the morphometrics should be far less variable compared to between several species. That could easily compound the errors inherent.

That is speaking of statistical significance, which can be far removed from clinical significance. A five percent change in mass may be statistically significant, but if its accumulation is spread across the body (typical) and not held in one concentrated area (atypical outside of depot sites), it will be of no clinical significance unless you were already at the end of a tail on the Gaussian distribution. For another easy/simple example of statistical versus clinical metrics, if a 100-pound adult human who eats a decent amount of fibrous vegetable matter would properly use magnesium citrate to basically clear out the entire GI tract, the human would lose a statistically significant amount of weight, yet would look pretty much no different and be in no different a state of health. Glycogen depletion with its associated water shedding is another example if you want to talk about true body mass changes instead of GI tract contents.

BMI was created to be easy metric more so than it was to be a high quality metric.

Perhaps true, but I consider my body a biological tool for my own purposes. If it does not fit one purpose, I modify it to some degree such that it approximates the ideal for that purpose as much as my genetic potentials will allow. There are always limits, of course. Some of the things I have done have helped me understand administration and the clinician-patient interface better, too, which is a side benefit of my "hobbies" that have helped in my career.
 

Madkins007

Well-Known Member
Moderator
10 Year Member!
Joined
Feb 15, 2008
Messages
5,393
Location (City and/or State)
Nebraska
OK, so enough of this banter. Bottom line- is the idea of a tool like this beneficial to some degree, when properly explained, and is it explained properly? What can we do to make something useful from all this?
 

Zamric

Well-Known Member
10 Year Member!
Joined
Oct 29, 2011
Messages
3,301
Location (City and/or State)
The Crystal Unicorn
....well, if we can get it to new members who are adopting or rescuing older torts, it will give them a general idea of what they may be up against as far as the health of their tortoise. Tho I do belive it could use some tweeking to give a better idea of true wieght-vs- size. extraplating a weight from just the shell lenth seams kinda dumb! Wouldn't we need to know the width of the shell as well or all tortoise proportions identicle? Would shell dome height also be concidered? I think that by adding the width and the height, you can eliminate the tiered results and be able to say LxWxH= approx. healthy weight... or am I trying to make it simpler than it really is? (I do that sometimes)
 

Neltharion

Member
10 Year Member!
Joined
Jun 10, 2011
Messages
253
Location (City and/or State)
Northern California
Madkins007 said:
OK, so enough of this banter. Bottom line- is the idea of a tool like this beneficial to some degree, when properly explained, and is it explained properly? What can we do to make something useful from all this?

A local tortoise club member (also a Vet) uses the same formula as a gauge to help determine if a tort is safe for brumation (notice, 'help' determine as this is not the only determining factor). At one of the club meetings every Fall, the members are free to bring their tortoises, and get them weighed. If the torts fall very low on the spectrum or very high on the spectrum, he recommends not allowing them to brumate. He also caveats that the measurement is not really applicable to species that tend to have shorter but wider shells.

This is very similar to the general height and weight formula used in humans. I would say that its generally a fair indicator, but there are exceptions.

Like in humans, a bodybuilder carrying a lot of muscle mass with a low body fat percentage would be considered overweight based on strictly height and weight measurements. There are going to be exceptions to the norm. Thank you for sharing this.
 
Status
Not open for further replies.
Top