Total Body Water Calculator

1. How to calculate total body water
2. Formulae for body water estimation
3. Limitations of total body water estimates
4. Applicability and limitations of the Watson formula
5. Limitations of the Hume-Weyers formula
6. TBW in children, including patients on peritoneal dialysis
7. Normal body water percentage

    How to calculate total body water

Total body water is the entire volume of water in a human body, including that found in the blood, bones, organs, and soft tissues. It is often used in a clinical setting when treating kidney diseases. For example, the TBW estimate is critical for patients on peritoneal dialysis (PD) when determining the delivered dialysis dose.

To calculate an estimate of the body water of a person our tool uses simple body measurements and anthropological data. The required input consists of the person's age, sex, height, and weight. The TBW calculator then uses predictive equations developed through scientific studies to produce an estimate of total body water volume (in US fluid ounces or litres), weight (lb, kg), and as percentage of the body weight (body water percentage).

The accuracy and validity of the body water estimates is discussed below in general and for each formula separately. For children the calculator employs equations developed specifically for children of one month of age to the age of 17, and for young adults. In all of these equations age is not used as input so when using the calculator for infants between the age of one month old and one year old users should enter "1" as the age.

    Formulae for body water estimation

Our calculator uses several formulas, all of which are regression equations based on studies on samples of the relevant populations. For adults, the two equations are the Watson formula and the Hume-Weyers formula presented below.

    Watson body water formula

Watson (1980) ^[1] studied a total of 714 people to derive two equations for total body water, one for men:

TBW = 2.447 - 0.09516 · Age + 0.1074 · Height + 0.3362 · Weight

and for women:

TBW = -2.097 + 0.1069 · Height + 0.2466 * · Weight

The input for both is in years, centimeters, and kilograms, and the output is in litres. Conversion to lb or kg is trivial given the density of water, and calculating the body water percentage is a matter of simple percentage calculations.

Accuracy of the Watson formula

The standard deviation of the resulting estimate for males is 3.76 litres, whereas for females the standard deviation is 3.60 litres ^[1]. This means 68% of estimated body water volumes will fall within ±3.76 / 3.60 of the true TBW, while ~95% of estimated values will fall within ±7.52 / ±7.20 of it.

The squared correlation coefficient r² for men is 70.4% and it is 73.6% for women, meaning that the equations account for 70.4% and 73.6% of the variability in observed body water content, respectively. Despite the initial equation being derived for a mostly European-descent US population, the Watson equation was found to offer a good match for South-Koreans by Seoung (2001) ^[6].

    The Hume-Weyers formula

The equations derived in Hume & Weyers (1971) ^[2] are based on a study of total body water of 30 male and 30 female volunteers with no clinical evidence of fluid retention or malnutrition which might bias the outcomes of the study.

TBW = 0.194786 * · Height + 0.296785 · Weight - 14.012934

TBW = 0.34454 · Height + 0.183809 · Weight - 35.270121

The input for both is in centimeters and kilograms, while the output is in litres.

The Mellits-Cheek equations

The work of Mellits & Cheek (1970) ^[3] had the goal of deriving "equations which accurately estimate total body water from anthropometric measures in normal individuals from infancy to adulthood". Using a sample of 251 observations (168 males, 83 females) they derived four regression equations. Two for men:

(1) TW = -6.773 + 0.170 · Height, when Height ≤ 132.7 cm

(2) TW = -66.751 + 0.623 · Height, when Height > 132.7 cm

and to for women:

(3) TW = 3.806 + 0.127 · Height, when Height ≤ 110.8 cm

(4) TW = -27.846 + 0.344 · Height, when Height > 110.8 cm

The observed standard deviations are as follows: 0.80 liters for equation (1), 3.60 l in eq. (2), 0.65 l in (3), and 2.17 l in (4). Our calculator shows their body water estimates if the entered age is below 31.

The Morgenstern formula (PPDSC)

While all the aforementioned formulae were developed mostly based on and for healthy individuals, the study by Morgenstern et al. (2001) ^[4] focuses exclusively on children on peritoneal dialysis (PD). It was conducted by the Pediatric Peritoneal Dialysis Study Consortium (PPDSC) and included taking deuterium oxide (²H₂O also written as H₂¹⁸O) measures of total body water in a group of 24 children with ESRD. Their ages were from 4 months to 19 years old and all were receiving chronic PD.

The study found Mellits-Cheek gives larger errors compared to a new set of equations Morgenstern et al. developed based on their sample of children on dialysis. The equations are:

TBW = 0.074 · (Height · Weight)^0.66 for boys

TBW = 0.117 · (Height · Weight)^0.59 for girls

    Limitations of total body water estimates

Each estimation method has its limitations, appropriate and inappropriate applications, and other limitations. General limitations shared by all total body estimation equations have been found by the comprehensive study of their performance by Tzamaloukas et al. ^[7]. The study advises that such estimates be used with care, especially when applied to individuals belonging to certain subgroups.

While the study finds the formulas generally fit-for-purpose: "Estimates of the average body composition provided by the anthropometric formulas agreed with estimates provided by the standard methods." it is quick to add the caveat that "these formulas have the potential of introducing large errors when estimating body composition in individuals differing from the average subject". Specifically, their evaluation finds that "all formulas systematically overestimate body water in subjects who are obese or experiencing volume excess."

The general limitation the paper concludes with is that "Anthropometric formulas [...] can provide only approximations of body composition and may be the sources of large errors in evaluating body composition in peritoneal dialysis patients. [...] These errors may alter the interpretation of urea kinetic studies in certain categories of peritoneal dialysis patients (e.g., obese subjects).".

    Applicability and limitations of the Watson formula

The major limitations of the Watson formula concern its application to children and adolescents as well as to the obese and very obese.

Only for adults aged 17 and above

This table is an extract from Table 1 found in Watson, Watson & Bett (1980) ^[1] and shows the sample used to derive their body water equation.

Sample used in deriving the Watson formula

Characteristic	Men	Women
Number of subjects	458	256
Age range (years)	17 - 86	17 - 84
Height range (cm)	132 - 201	124 - 181
Weight range (kg)	36.4 - 148.3	31.4 - 186.4
TBW as % of BW (range)	38.5% - 73.5%	27.4% - 70.9%
TBW as % of BW (mean)	58.3% ±6.7%	48.5% ±8.6%

A quick look at the sample used to derive the Watson equation shows it shouldn't be applied to people below the age of 17, meaning to young children and adolescents, since that population was not represented in their sample. The authors state as much in their work: "These values were used to derive total body water prediction equations for adults of any age." (emphasis ours). This is the reason this body water calculator does not show results from the Watson formula for anyone below the age of 17.

Overestimation in very obese individuals

Extreme variation of the percentage of body fat is the main factor affecting the accuracy of TBW prediction from equations based on bodily measures and anthropological information. According to Watson ^[1] "The effect is much more pronounced in the grossly obese than in the very lean." This has mostly to do with the relatively low water content in adipose (fat) tissue (10 to 30% water) which, according to the authors leads to "TBW prediction formulae derived from subjects with a wide range of body fatness will tend to overestimate TBW in the grossly obese". Gross obesity was defined by the authors as a body fat percentage above 40% for males and above 53% for females.

Since there were only a handful of people in the study with such a body fat percentage, the equation produced cannot accurately account for them. This was independently verified by Tzamaloukas et al. ^[7] which states that "formulas systematically overestimate body water in subjects who are obese or experiencing volume excess.". You can get a quick estimate of your body fat percentage using our body fat percentage calculator.

    Limitations of the Hume-Weyers formula

The age range of the participants' in the Hume-Weyers study was 35 to 71 years old for males and 33 to 84 years old for females, which means younger adults, children and adolescents were not represented at all in their sample. The relatively small sample size makes the generalizability of this formula potentially worse than that of Watson's and it is likely that it suffers from issues similar to limitations similar to those characteristic of Watson's.

Despite the above, the Hume-Weyers equation produces body water estimates close to those of Mellits-Cheek across the entire age range from age 1 to age 30, and it gives results close to Watson's formula for ages 17 and above. This is why our total body water calculator also includes it in its results.

    TBW in children, including patients on peritoneal dialysis

Unlike the other two formulae, the Mellits-Cheek equations were derived specifically for children as their sample included children as young as 1 months old. Mellits and Cheek consider their equations applicable from the age of 1 month old to 34 years of age for males, and from 1 month to 31 years of age for females ^[3] which is supported by their sample population. Their equations seem to be well-suited for estimation in healthy children, adolescents, and young adults.

For children on peritoneal dialysis however one should prefer the equations developed by Morgenstern and his team. The superiority of the PPDSC equations for use in children on PD has been confirmed in a comprehensive comparison by Mendley, Majkowski & Shoeller (2005) ^[5] seeking to validate estimates of total body water in pediatric dialysis patients by deuterium dilution.

From the paper: "Current K-DOQI recommendations include the use of the Mellits and Cheek formula for estimation of TBW, but our data demonstrate systematic errors in the estimate of TBW using this recommended model. The formula proposed by the Pediatric Peritoneal Dialysis Study Consortium provided the best estimate of TBW and, although it must be recognized as still relatively imprecise, we recommend its adoption in this setting. The 95% confidence interval for this estimate (-.86 to.84 L) makes it useful for all but the smallest of children receiving PD."

    Normal body water percentage

From the table showing the sample used in the Watson study ^[1] the average body water percentages differ significantly between men and women, so it is best to consider the two genders separately. The average for men is 58.3% with a standard deviation of 6.7% meaning 95% of the men from the sample had a body water percentage between 45.2% and 71.5% (normal quantile calculation). The average body water percentage for women is 48.5% with a standard deviation of 8.6% meaning 95% of the women had a TBW percentage between 31.6% and 65.4% (normal quantile calculation). These should be representative of the U.S. adult population as of 1980, but may be on the higher end as of 2026 given the well-established increase in the prevalence of obesity found in nationwide studies reported by the CDC.

The large difference of roughly 10 percentage points between men and women is likely mostly due to the propensity for the female body to have a higher body fat percentage. This is also reflected in the army body fat standards which allow a much higher body fat percentage for candidates and enlisted women (6 p.p. and 10 p.p. higher, respectively). As discussed above, adipose tissue contains a lower proportion of water compared to the average of the rest of the body.

Breakdown by age group

Given that the body water percentage varies somewhat with age, it can be useful to examine the averages across major age groups. The table presents average BW percentage by age and gender from Table 2 of the Watson 1980 paper ^[1].

Body water percentage by age group and gender

Age group (years)	Men	Women	Difference
20 - 29	60.5%	49.6%	10.9 p.p.
30 - 39	57.8%	48.8%	10.3 p.p.
40 - 49	57.0%	47.5%	9.5 p.p.
50 - 59	56.4%	47.1%	9.3 p.p.
60 - 69	56.5%	44.5%	12 p.p.
70 - 79	54.0%	47.0%	7 p.p.
80 - 89	56.3%	52.0%	4.3 p.p.

The age distribution shows that in both sexes the water percentage generally declines with age. For women it reverses the trend and in fact reaches higher values in the 80-89 age group compared to the 20-29 age group. The tendency in older age is similar in men, but not as pronounced. The difference between the genders remains relatively consistent except for the 70 - 79 and 80-89 age groups where it shrinks significantly to 7 p.p. and 4 percentage points respectively.