Materials and Apparatus by mikeholy


									               Bioengineering 210

                  Laboratory II

                 Final Project:

Analysis of Ionic Concentration of Sweat Using
  the Atomic Absorption Spectrophotometer

                                    Professor Dr. Mitchell Litt

                                              SPRING 1998
                                                  Group W6
                                                    Paul Kim
                                           Catherine LaRocco
                                                    Dale Yoo
                                                     Hua Zhu

        In this experiment, three sweat collecting methods and devices were designed and
tested. Among the three methods and devices, a simple sweat-catching method using test
tubes and a slightly more complicated method using a flexible pouch constructed with
Opsite-type dressing and parafilm were tested and determined to be inaccurate or
inefficient because of various factors. The third device was constructed with a 100 mL
square petri dish and syringe. It was tested under experimental conditions and proved to
be the most efficient for easy sweat collection. It was used to collect all of the sweat
samples in this experiment because of its ability to prevent contamination of the sweat
sample, thus, ensuring the accuracy of the measurements and data.
        In the second part of the experiment, sweat samples were collected in dry and
steam sauna conditions with mean temperatures of 185 F and 156 F, respectively. The
relative humidity was negligible in the dry sauna and 100 % in the steam sauna. The ion
concentration of Na+, K+ and Ca2+ in the sweat samples were determined using an atomic
absorption spectrophotometer. According to previous research conducted regarding
sweat content, the first data points of the potassium and calcium concentration do not
truly reflect the concentration of ions in the sweat. This was supported by our data.
Additionally, we found no obvious effect of relative humidity on ion concentration in
sweat except for Ca2+ ion concentration, where the dry trials yielded much higher calcium
concentrations than in the steam sauna. Since there was no distinction between the
physiological conditions overall, we calculated the mean ion concentrations of all trials.
The mean concentration of Na+, K+ and Ca2+ in sweat was found to be 1884 ppm, 236.0
ppm, and 8.441 ppm, respectively. The experimental sodium ion concentration was
lower than the normal blood ion concentration as well as literature values for drug-
induced sweat production but was higher than the literature value for thermally induced
sweat. On the other hand, the experimental potassium ion concentration was higher than
the normal blood ion concentration and the drug-induced sweat but lower than the
thermally induced sweat. Calcium ion could not be compared to literature values because
of the lack of data for this ion.


        There are many forms of sweating that occur in the human body: eccrine,
emotional, gustatory, and apocrine, although eccrine is the only form to secrete a
significant amount of fluid and electrolytes. The other forms occur sporadically and
often have inconsequential fluid loss. Under stress, a man is capable of sweating at rates
as high as 2 liters/hr and total volumes in excess of one quarter their total body fluid in a
single day.[1]
        The secretion of sweat glands, a type of exocrine gland, cools the body by
evaporation and, at least in lower life forms, produces a sexual attractant odor.[2] The
millions of eccrine glands produce an odorless, light and watery[3] diluted salt solution for
thermoregulation,[4] while the bacteria around the apocrine glands produce odor-causing
byproducts as waste.[3] These bacteria are nourished from the glands’ protein-rich
secretion.[4] Deodorants are antimicrobial agents that minimize the number of bacteria
and fragrances to mask any remaining undesirable smells. Antiperspirants are aluminum
salts such as the very potent and irritating aluminum chloride or the least powerful and
least irritating aluminum chlorohydrate, which plug up sweat gland ducts to stop odor and
decrease the amount of wetness.[3]
        The literature shows that pharmacologically induced sweat contains 3381 ± 253
ppm of sodium and 195.5 ± 39.1 ppm of potassium.[5] (There are no values for the
concentration of calcium in sweat because it is found in trace amounts.) Various
researchers agree that the range of ion content of sweat falls within or close to the range
of ion content of blood (Table 1).[5,6]

          Sodium Reference Range                              3128 ppm – 3358 ppm
         Calcium Reference Range                             48.50 ppm – 52.91 ppm
        Potassium Reference Range                            136.85 ppm – 195.5 ppm
                                 Table 1: Ion Content of Blood[6]

                                        It has been found, though, that the human sweat
                                duct utilizes an ATP-dependent pump (Figure 1). The
                                research of Sato, Fiebleman, and Dobson indicates that
                                this pump can cause lower sodium and higher potassium
                                concentrations than expected from pharmacological
                                data.[7] This type of pump, or primary active transport
   Figure 1: Classic Na+/K+     carrier,[8] uses the energy derived from ATP to drive the
   Electrogenic Pump[9]
                                simultaneous active transport of Na+ ions out of and K+
ions into the cell,[9] both against their respective concentration gradients. One cycle of
the constantly active pump [8] moves three Na+ ions and two K+ ions. This pump exists
in animal cells and uses a large percentage of the metabolic energy of the cells.[9] It is
what keeps the constant[10] low [Na+] and high [K+] in the cytoplasm.[9]
       In this study of sweat, it was determined that an
analysis of a sports drink would help one to understand
thermally induced sweat. As a case study, Gatorade
Thirst Quencher®, a registered trademark of Stokely-Van
Camp, Inc., was investigated. According to this
company, a sports drink should do all of the following:
stimulate rapid fluid absorption, assure rapid
rehydration, provide carbohydrate energy to working
muscle, and encourage more drinking of fluids.[11]
Gatorade Thirst Quencher® contains 0.4583 g/L (458.3
ppm) of sodium and 0.1250 g/L (125.0 ppm) of
                                                              Figure 2: Gatorade Label [12]
       Stokely-Van Camp, Inc.’s research has shown
that beverages containing about a 6 % carbohydrate level and a small amount of sodium
are ideal for rapid fluid absorption. But as the carbohydrate percentage of a drink is
increased, absorption is slowed. For this reason, drinks that contain carbohydrate levels
that are greater than 7 % are not recommended. Complete restoration of body fluids is
best when the sodium that was lost in sweat is replaced along with fluids. In addition, the

sodium and glucose levels in Gatorade stimulates people to drink more fluid voluntarily
until the body is rehydrated.[13]
        Sweat is a biological fluid that is often
overlooked, so there still remains many technical
problems such as dilution, condensation,
contamination, and evaporation, that must be
solved when developing a sweat collection device.
Brisson et al. have developed a technique using an
OpSite-like medical dressing (Figure 3). Before
application on a cleaned and dried surface of the
mid-lumbar region of the subject’s skin, the
                                                         Figure 3: Design 2 Schematic of Sweat Collector
adhesive side of the dressing is exposed and a
                                                         using Smith & Nephew's OpSite Wound
piece of permeable laboratory film paper, such as        Dressing[14]

parafilm, is attached. The dressing is then transferred to the skin forming a “pocket”. A
small opening is created in the upper part, but it is kept closed when the sweat is not
being extracted.[14] The sweat can be suctioned from the dressing by using a Vacutainer®
tube inserted in a tube holder and fitted with a long dull needle.[15]
        Boisvert et al. modified the previously mentioned device so that it would allow
collection in a warm (30 C) and humid (relative humidity 80 %) environment. This new
technique would reduce leakage and hidromeiosis during intense or prolonged exercise.
The first modification was the application of hypo-allergic glue, which is made of
benzion simple tincture and binds the OpSite-like membrane to the surface of the skin.
This has been shown to prevent leaks when collection extends over one hour or is
performed in humid environments. Secondly, the excreted sweat was diverted to a pouch
chamber attached to the lower end of the device. This permitted the sweat to accumulate
away from the skin to minimize potential hidromeiosis and changes in skin
        Another method, by Taylor, Polliack, and Bader, utilized a 4 cm  4 cm square of
Whatman cellulose chromatography paper for sweat collection. This paper is covered
with a water-impermeable sheet of 50 m thick polypropylene and sealed around the
edges with surgical tape. After collection, the cellulose pad was placed in a tube and

deionized water was added. The tubes were centrifuged at 3000 G for 10 minutes in
order for the sweat to accumulate.[16]


Preparation of Calibration Solutions
 Solutions of 1000 ppm K+, Ca+2, and Na+
 Fisherbrand® plastic 15-mL and 50-mL disposable test tubes and Nalgene® test tube
 4 plastic 100-mL disposable beakers
 2 plastic 1000-mL disposable beakers
 Denville XL 3000 20-200 L and 100-1000 L pipettes
 Fisherbrand® Redi-Tip 200-L and 1000-L pipette tips
 Drummond Scientific Co. pipet-aid®
 Fisherbrand® disposable 10-mL pipettes
 Scoopula
 Deionized water

Sweat Collection
 Becton Dickinson Vacutainer® Blood Collection Set
 Puncture needle
 Long dull needle
 Johnson & Johnson Medical Inc. Bioclusive* transparent dressing
 Paraflim
 Fisherbrand square petri dish with grid (13-mm  33-mm  33-mm)
 Tycon tubing (1/32 inch od - 3/32 inch od, 1/16 inch id - 1/8 inch od, 1/8 inch id - 1/4
   inch od)
 Super glue
 Cyclohexanone
 20-mL Becton Dickinson syringe
 3-foot  10-inch (approximate) latex Thera-Band® strip
 10-mL test tubes
 Kendall Webcol® Alcohol Prep, Sterile, Saturated with 70 % Isopropyl Alcohol
 Paper towel
 Deionized water

Determination of Ion Content
 Perkin-Elmer Atomic Absorption (AA) Spectrophotometer 4000
 Microsoft Excel®
 Deionized water


       There were two parts in this experiment: 1) to design an efficient, reliable and
reusable sweat collecting device and 2) to determine the concentration of sodium,
potassium, and calcium ions in sweat.

Design 1
       Our first method, which was the most primitive one, was to swab a 10-mL test
tube against the subject’s skin to collect the sweat.

Design 2
       The second method was a simple disposable device designed after studying
several scientifically related research papers and consulting Dr. Henry R. Drott, an expert
in the field of cystic fibrosis and sweat collection at the Children’s Hospital of
Philadelphia (see Acknowledgements). This device was designed to apply the special
properties of OpSite-type medical dressing. It was constructed with a piece of parafilm
cut into the shape that would collect sweat at the bottom of the device when applied
vertically on the lumbar region of the subject’s back (Figure 3). It was fixed by applying
a piece of OpSite-type medical dressing covering the entire device, thus, creating a small
airtight pouch under the parafilm for collecting sweat. Then, the sweat collected at the
bottom of the pouch would be extracted with a needle connected to a Vacutainer®, which
sucks the sweat from the pouch into the tube. Because of the special properties of the
dressing, it would adhere to the skin firmly even under humid conditions, such as those
created by sweating.

Design 3
       After receiving advice from Gia Hchevia, a biotechnologist and researcher at
Hospital of the University of Pennsylvania (see Acknowledgements), a third device was
designed to solve the problems encountered in the first two. This device was made of a
square petri dish of 13-mm  33-mm  33-mm. A size 1/32 inch inner diameter - 3/32
inch outer diameter tube was glued to a hole (diameter 2/32 inch) drilled at one of the

corners of the petri dish. At the other end of the tube, a 20-mL syringe was attached for
collection (Figure 4). The device was applied to the lumbar region of the subject’s back
and fixed with a wide latex strip (Figure 5). An
airtight space was created between the petri dish
and the subject’s skin, and the sweat was
extracted from the plastic tube by a syringe.

Sweat Collection
       Using the third design, which was
determined to be the most efficient and accurate,
sweat samples of the same subject from four
trials under two distinct conditions were collected Figure 4: Design 3 Sweat Collection Apparatus
and analyzed. Two trials were carried out for
each of the two environmental conditions, which were created by a steam sauna and a dry
sauna. In the dry sauna, the temperature was approximately 185 F with negligible
humidity. In contrast, the temperature in the steam sauna was 156 F with relative
humidity of 100 % since the air was saturated with water vapor. A sweat sample was
collected every 15 minutes in each trial and the
subject was allowed to rest for 5 minutes during
extraction of the sample to prevent severe
dehydration. Between each trial, the subject rested
1 hour and drank 1 liter of deionized water to
recover to his normal state.

Preparation of Calibration Curves
       To analyze the ion content of sweat using
an atomic absorption spectrophotometer,
calibration curves must first be constructed for
                                                           Figure 5: Application of Design 3 onto
each ion that will be tested before each run with          Lumbar Region of Subject

the machine. The 1000 ppm stock solutions of potassium ions (K+), calcium ions (Ca+2),
and sodium ions (Na+) must be diluted in order to find the linear region of the absorbance

versus concentration curve. Using the method of parallel dilutions, these solutions were
diluted to 100 ppm, 50 ppm, 10 ppm, 8 ppm, 6 ppm, 4 ppm, and 2 ppm, making 50 mL of
each. The absorbance of each of these solutions was determined three different times and
then the averages of these three determinations were plotted as a function of their
respective concentrations. A best-fit line was formed by linear regression and its
equation was calculated using the Microsoft Excel® computer program. By plotting such
low concentrations, we would obtain an accurate linear region for the calibration curves,
which typically look like Figure 6.

Determination of Ion Content
       Before actually measuring any
values from the sweat, we diluted the
sweat samples for two reasons. First,
we only had approximately 1 mL of           Absorbance

the each sweat sample so if we ran the
sample through the AA
spectrophotometer without diluting it,
we would have lost all of our sample
in that one measurement. Secondly, the sweat        Figure 6: Typical Absorbance        versus   Ionic
                                                    Concentration Calibration Curve
sample was too concentrated to fall into the
linear range of our calibration curves.
       At the start of our experimentation, we had literature values from Bijman[5] to
predict what the concentrations of the sweat samples might be. So we decided to dilute
the sample by different dilution factors, which were calculated specifically for each ion.
We initially decided to dilute all the samples from both of the dry sauna trials by a factor
of 50. We found that at this dilution factor, the Na+ concentrations in the samples were
still above the linear region of the calibration curve ([Na+] >10 ppm). The K+
concentrations fell consistently within the linear range (2 ppm < [K+] < 10 ppm) while
the Ca+2 concentrations were below the linear range ([Ca+2] < 2 ppm). Therefore, we
decided to further dilute the samples to a dilution factor of 1000, and we observed that
the [Na+] readings finally fell within the linear region of the calibration curve for Na+.

We determined that it was acceptable for the calcium ion to have concentrations below 2
ppm because we assumed that the amount of absorption was zero at zero concentration.
Hence, we could extrapolate the best-fit line by adding the point (0,0) to our regression
line. [Ca+2] of the sweat samples would now fall into the linear region of the calibration
curves. So for the wet sauna samples collected the week after, we diluted each sample by
a factor of 10, 50, and 1000 to test for the [Ca+2], [K+], and [Na+], respectively. Note that
we changed the dilution factor of the calcium ion from 50 to 10 so that the data point
would be closer to the range that we actually investigated (2 ppm < X < 10 ppm). The
[Ca+2] values were still lower than 2 ppm at a dilution factor of 10; however, we could
not use a lower dilution factor because we would not have enough of the sweat sample to
obtain all the appropriate absorption readings from the AA spectrophotometer. For each
sweat sample, we took three readings and averaged them. We used these averages to
determine their respective concentrations by using the calibration curves.


       The following graphs represent all of the collected data for each of the three ions
after using the appropriate calculations:

                                                               Concentration versus Time for Na+

                      Concentration of Na + (ppm)


                                                    2000                                     Dry Sauna Trial 1
                                                                                             Dry Sauna Trial 2
                                                                                             Steam Sauna Trial 1
                                                    1000                                     Steam Sauna Trial 2


                                                           0     20      40        60   80
                                                                      Time (min)

                     Figure 7: Calculated Concentration of Sodium Ions versus
                     Duration of Experiment - The four trials are represented by
                     the different data point patterns, as indicated in the legend.

                               Figure 8: Calculated Concentration of Calcium Ions
                               versus Duration of Experiment - The four trials are
                               represented by the different data point patterns, as indicated
                               in the legend.

                         Figure 9: Calculated Concentration of Potassium Ions versus
                         Duration of Experiment - The four trials are represented by the
                         different data point patterns, as indicated in the legend.

        The following table represents the results of t-tests that were run to compare the
two separate trials within the same condition for each ion.

      Type of Trial Comparison Mean - Trial 1 Mean - Trial 2              t Stat    t Critical two-tail
            [Na ] in Dry         1596.89        2249.06                  -1.997            2.447
            [Na ] in Steam               1735.83           1952.81       -2.195            2.447
              [K ] in Dry                 155.35           245.03        -2.229            2.447
             [K ] in Steam                178.81           208.69        -1.118            2.447
             [Ca ] in Dry                 35.07             19.68        4.258             2.447
           [Ca ] in Steam                  5.38             6.81         -1.123            2.447

Table 2: T-Test Results - This table shows the comparison of means between the two trials in each of the
two conditions (dry sauna and steam sauna) for each of the three ion concentrations ([Na +], [K+], and
[Ca+2]). Also, their respective t Stat and t Critical two-tail values are shown.

                                                                              Consistency of Trial 2 and Trial 1

                                   Trial 2 Concentration (ppm)
                                                                 200                                                 y=x
                                                                 150                                                 R2 = 1

                                                                          0   50       100             150    200       250        300
                                                                                       Trial 1 Concentration (ppm)

              Figure 10: Ideal Graph of Trial 1 Concentration Data versus Trial 2 Data
              – This figure shows the ideal consistency between two trials. Since both
              trials are performed under the same condition (i.e. steam sauna or dry
              sauna), it is assumed that the ion concentration as a function of time is the
              same for both trials. Hence, in an ideal situation, the ion concentration at a
              particular time is the same for both trials since the external conditions have
              the same effect on both trials. It is assumed that no other variables exist that
              affect ion concentration between trials

                                                                 Steam Sauna Trend Comparison for [Ca+2]

    Trial 2 Concentration (ppm)

                                                                                       R = 0.961




                                                                 0            5                   10          15              20
                                                                               Trial 1 Concentration (ppm)

Figure 11: Graph of Trial 2 [Ca+2] versus Trial 1 [Ca+2] – This figure shows
that there is consistency between the Trial 1 and Trial 2 data for calcium ion
concentration in the sweat samples that were collected in the steam sauna. The
R2 value of 0.961 shows that there is a strong correlation between the two sets
of data.

            Ion         Mean Sweat         Mean Blood Ion                 Percent
                     Concentration (ppm) Concentration (ppm)            Deviation (%)
        Sodium               1884                     3243                   -41.9
        Potassium            236.0                    166.2                   42.0
        Calcium              8.441                    50.71                  -83.4

Table 3: Mean Sweat Concentrations Compared to Blood and Reference Sweat Concentrations of Ions –
Sodium, potassium, and calcium mean ion concentrations in our experiment in comparison to blood ion

             Ion              Experimental             Drug Induced     Thermal Induced
                            Sweat Conc. (ppm)        Sweat Conc. (ppm) Sweat Conc. (ppm)
   Sodium                            1884                  3381                   648.0
   Potassium                         236.0                 195.5                  246.0
   Calcium                           8.441                  ---                    ---

                     Table 4: Pharmacological versus Thermally Induced Sweat


       Our experiment investigated the ion content of sodium, potassium, and calcium in
sweat. Since the analysis of sweat entailed the actual sampling of sweat, the first aspect
of our experiment was to design a device for collection. We tested three devices as
shown in the Methods section. Device 1 simply involved the catching of sweat droplets
as they fell from the subject’s forehead. Very quickly, this method was eliminated for
several reasons. First, the catching of sweat into a test tube proved to be too tedious and
complicated. We thought that we could just collect the sweat into a larger tube or funnel
but that proved to be impossible because the sweat droplets adhered to the sides of the
container and the funnel when they were transferred into smaller tubes. Additionally, this
method would not yield accurate data because the water in the sweat had time to
evaporate as the drops were collected. In addition, in trying to swab the sweat into the
tubes, some epithelial cells were inevitably collected along with the sweat sample. These
cells would need to be centrifuged out of the sweat. Otherwise, these cells would
ultimately clog the aspirator of the AA spectrophotometer. Furthermore, this method was
an “open” collecting design, which means that the collected sweat was in direct contact
with the air and may have become contaminated. Boisvert[15] et al. previously reported
that there was a great discrepancy of concentration data between sweat collected from an
open system and a closed system. Therefore, some type of closed collecting system
needed to be devised.
       The second device that was created was an imitation of a sweat collecting device
described in an article by Brisson et al.[14] It consisted of a flexible “pouch” attached to
the body that would be essentially airtight and prevent evaporation. Since evaporation
would increase the measured concentrations of ions, measurements of samples from an
open system would not be purely representative of sweat secreted by the body. Hence,
this device should yield better results than the open system. However, this device had
several problems. The OpSite-type dressing used to fix the device onto the subject’s
body lost its adhesiveness when the skin under the dressing began to excrete sweat
causing leakage from the pouch and making the collection of a sufficient amount of sweat
within the designated time impossible. Furthermore, the sweat that was collected was too

concentrated since much of the water had evaporated from the sample from the leak in
the pouch.
       The third device that was used was determined to be the most efficient and
accurate device. It eliminated possible evaporation, minimized leakage, and ensured that
all the sweat collected was from the designated region. Evaporation was obviously
minimized because of the rigidity of the container as well as the airtight withdrawal
system, which involved the use of a syringe. As the plunger was pulled, negative
pressure was created within the petri dish, hence, maximizing the collection of sweat at
the hole. Additionally, the specific location of the petri dish in the middle lumbar region
was helpful because the skin surrounding the dish was flexible; this region was
recommended by Brisson et al.[14] and Dr. Drott (see Acknowledgements).
       The second half of our experiment dealt with the actual measurement of sodium,
calcium, and potassium ions in sweat during two different physiological conditions and
comparisons of the results. The first physiological condition was a dry sauna with an
average temperature of 185 F and negligible relative humidity. The second condition
was a steam sauna with an average temperature of 156 F and relative humidity of 100 %
since the air was fully saturated with water vapor. When analyzing the sweat samples
that were collected under these two conditions, we tried to determine if there was any
correlation between the sweat ion concentration and the humidity of the environment, in
addition to the relationship between concentration and time. We predicted that the ion
concentration of the sweat would be higher in the steam sauna than in the dry sauna. Our
predictions were based on the fact that the human body tends to excrete higher
concentrations of ions with increases in the sweat rate as previously documented by
Boisvert[15] et al. in 1993 and Gibson and di Saint’Agnese[17] in 1963. This would occur
because of the higher rate of sweat evaporation in the dry sauna than in the steam sauna
where the relative humidity was approximately 100 %.[15] This means that the subject, in
the dry sauna, would not need to sweat as much in order to cool down the body since the
sweat evaporates more quickly than the sweat in the steam sauna. On the other hand, in
the steam sauna, where relative humidity is high, the body could not cool down as
quickly since the sweat on the skin could not evaporate into the saturated air. From the
data that we obtained, we found that there was no significant difference between the

results from the dry sauna and the steam sauna trials for sodium and potassium ions.
There was no observable trend between the two external conditions. The data sets from
the steam sauna and the dry sauna for the potassium and sodium ions overlapped when
concentration was graphed versus time (Figures 7 and Figure 9). A statistical analysis
comparing the two physiological conditions was impossible because the number of trials
was too small. Nonetheless, observation of these graphs will show that there is no
distinction between the trials performed in the dry sauna and the trials performed in the
steam sauna since there is so much overlapping between the plots. However, for the
calcium ion, both trials in the dry sauna yielded higher concentrations than the trials in
the steam sauna (Figure 8). This result contradicts our prediction that the ion
concentration in the dry sauna should be less than that of the steam sauna. From Figure
8, we may believe that the difference in relative humidity affects the concentration of
calcium in sweat, but further experimentation must be done to prove this correlation.
Additionally, no previous literature indicates any significant concentrations of calcium in
normal sweat. This prevents us from performing a historical comparison in order to
verify the validity of our data.
        One important observation to be made from the potassium graph (Figure 9) is the
fact that the first concentration point for potassium in each trial tends to be much higher
than the rest. This type of characteristic was also noted by Sato and Sato in 1990.[18] The
first data point represents the measurement of sweat during the first 15 minutes. During
this time, the body is attempting to restore ion equilibrium and adjust to the
environmental conditions. At temperatures above 37 C, the body needs to sweat in
order to maintain its homeostatic internal conditions. Furthermore, the first sweat
produced by the body contains many ions and impurities that tend to inflate the
concentrations of the first set of data. This anomaly can also be explained
physiologically. The ion pumps located in the basal and ductal sweat cells may not be
completely active at the start of sweat excretion. The sodium-potassium pumps may not
become active until the body begins to regulate its temperature. At this point, the ductal
cells actively pump potassium from the extracellular fluid into the cell. This pump
decreases the potassium concentration in the sweat for some time as correlated to the
rapid decrease in potassium ion concentration between the 15 minute measurement and

the 30 minute measurement as shown in Figure 9. Concurrently, the sodium
concentrations remain fairly constant in the sweat. Without any active pumping, the
sodium concentrations should decrease over time because the body needs to retain
sodium ions. The sodium ion concentrations remain fairly constant for two reasons.
First, the sodium pumped into the sweat in exchange for potassium ions. Secondly, as
sweat rate increases with time as shown by Boisvert,[15] the sodium ion concentration
increases slightly.[15] It can not be firmly concluded that our sodium concentration
increases over time, but it is evident that the sodium concentrations are constant over
time whereas the other two ions show a drastic decrease in ion concentration between the
15 and 30 minute measurements. Regardless, the consistency of the data after the first 30
minutes for all ions justifies the conclusion that the first set of data is not completely
representative of the data throughout the experiment.
       A major limitation of our experiment was the fact that we were only able to
perform two trials for each physiological condition (dry and steam sauna) because of time
constraints. Therefore, statistical analysis was very difficult and observing trends within
trials are virtually impossible. Two different approaches were taken to compare the
trials. First, we observed that the data points at 15 minutes were much higher than the
other four points in the trial for potassium and calcium ions. For all three ions, there is a
region on the graphs (Figures 7-9) where the concentrations are fairly constant. In order
to treat the data as though the concentrations were not dependent on time (since the
concentrations remained fairly consistent after 15 minutes), we excluded the data points
at the first 15 minutes and used the other four points to find the means and perform a t-
test analysis. This was also justified by the fact that sweat rate tends to become more
constant after the body has had time to adjust to the external stimuli. Hence, we
performed t-tests comparing the mean of one trial to the mean of the other trial within the
same physiological condition. This data is listed in Table 2. From the Table, one can see
that the data for the sodium ion and the potassium ion are statistically indifferent for both
external conditions (dry sauna and steam sauna). Additionally, the data for concentration
of calcium ion in the steam sauna seems to be statistically indifferent. However, for the
calcium ion in the dry sauna, there is a statistical difference between trials. The t-stat
value for these trials is 4.258 which is greater than the t-critical value of 2.447 for 95 %

confidence and n = 4. Secondly, we performed a unique statistical analysis comparing
the two trials of data within each environmental condition. A method of comparing two
trials regardless of their dependency on time was defined to analyze any consistency in
the data. This method differed with the t-test method because it assumes that there was
no time dependency on concentration and that the concentrations remained fairly
constant. This method assumes that there may exist a strong relationship between time
and concentration and tests this consistency between the trials. Specifically, the data for
the concentration of a particular ion in trial one was plotted against the concentration of a
particular ion in trial two. Ideally, if the two trials were identical to each other, it would
be logical to conclude that these two trials were not statistically different. Therefore,
time was ignored completely since, in an ideal situation, the time dependency for ion
concentration would be the same for the two trials when run at identical conditions. If
two sets of data within the same external condition (i.e. the steam sauna) yielded identical
ion concentrations at every timed measurement, the graph of the ion concentration in trial
two versus the concentration of the same ion in trial one would be perfectly linear with a
R2 value of one. Once the graph of ion concentration for trial one versus trial two was
made, a linear regression was performed, setting the y-intercept equal to zero. Figure 10
shows what an ideal graph of two trials of data would look like. We performed this
statistical analysis, comparing trial one and trial two for each of the ions in each
environmental condition. We evaluated the precision by observing the R2 value of the
best-fit lines. Only the two trials for the calcium ion in the steam sauna showed
consistency. Figure 11 shows that the R2 value of the linear regression between the trials’
data is 0.961. The low R2 values of the linear regressions for the rest of the data indicate
that there were many other variables that caused variation between the two trials. Food
and drink consumption in between the trials in addition to different pre-sauna dietary
habits may have affected the consistency between trials. In order to improve the
consistency in the future, it is necessary to perform sufficient trials under the same
conditions in order to define a relationship between ion concentration and time.
        Another important aspect of this experiment was comparing the ion
concentrations found in the sweat samples with those found in normal blood. As shown
in Table 3, the normal ion concentration in blood of sodium, potassium, and calcium are

3243 ppm, 166.2 ppm, and 50.71 ppm, respectively.[6] Our experiment yielded mean
concentration values for sodium and calcium for both conditions to be 1884 ppm and
8.441 ppm, respectively, which are well below the blood ion concentration levels
mentioned above. This would be expected since the sweat cells would attempt to
maintain ion equilibrium within the cells while excreting more dilute solutions of sweat
to cool down the body. However, we found the ion concentration for potassium to be
236.0 ppm. This value was above the normal blood concentration for potassium. When
one eliminates the first potassium data point, which was discussed to be an anomaly
within the data, the mean potassium concentrations in both steam and dry sauna trials
become comparable to that of blood concentration for potassium. However, the
potassium ion concentration is still a bit high since it would normally be expected to have
all ion concentrations less than that of blood. An explanation of this was briefly
mentioned in the Background section. Since each ductal sweat cell has a sodium-
potassium pump to regulate ion concentration in the body, an increase in the potassium
ion will correlate with a decrease in the sodium ion concentration.[7] This would explain
that sodium ion concentration was 44 % less than that of blood. As shown in Table 3, our
potassium concentration was almost equal to that of blood while the sodium
concentrations were well below the level in normal blood. Sato, Feibleman, and Dobson
also previously observed this trend in 1970.[7]
       When one compares the ionic concentration of the sports drink, Gatorade Thirst
Quencher, with the experimental values in this experiment, one finds that the sodium
ion concentration of the drink is approximately a quarter of that found in the sweat while
the potassium ion concentration of the drink is approximately half of that found in the
sweat. This is an interesting observation because this sports drink’s purpose is to
replenish any losses in fluids and ions. The reason why the sodium ion concentration may
be so low is because smaller amounts of sodium are ideal for rapid fluid absorption. For
the ionic concentration of potassium, there is not a clear explanation why it is so low in
the drink. Perhaps there are industrial or economic reasons why Stokely-Van Camp, Inc.
cannot increase the [K+] concentration or maybe they ran sweat studies solely on athletes,
which may have sweat that is more diluted in ionic concentration than the general

       The concentrations of the ions investigated in our experiment were also compared
to those of literature values for sweat under other conditions. Table 4 shows the ion
concentrations of sodium and potassium in sweat when isolated both pharmacologically
and thermally. Drug induced sweat experiments are very common because of their
efficiency and ability to control the degree of sweating by varying the concentrations of
the drug. The most common drug used was pilocarpine.[7] Although pharmacologically
induced sweat may have been easier to manipulate, the actual mechanism of sweat
production tends to skew data. Since the drugs pacify the sodium-potassium pumps in
the sweat cells, there was a higher overall sodium ion concentration in the sweat that was
produced pharmacologically than that in sweat that was produced thermally. For our
experiment, we only investigated the thermal effects of sweat production. The majority
of research done on sweat ion concentration in the past has used the pharmacological
approach in the collection of samples. Additionally, the drug-induced samples were
collected from the forearm, which may have caused some deviation from their values and
the ones that were collected in this experiment, but the main difference occurs because
pharmacological effects tend to increase ion concentrations in sweat. However, one
particular experiment performed by Taylor et al. found the median [K+] in sweat to be
648 ppm and the [Na+] to be 246 ppm when thermal effects were the only factors
initiating sweat production.[16] These values are significantly lower than our
experimental mean values for these ions. There are a few reasons for this. First, Taylor’s
experiment entailed collection of sweat over a nine-hour period. Ion concentration by the
end of the experiment would have been very low. Second, the ambient temperature was
only slightly greater than body temperature at 32 C, which meant that the sweat rate was
slow. Since the sweat rate was considerably slower than that of this experiment, it can be
concluded that the difference in sweat rate was the main cause for the concentration
discrepancy. From these varying values of concentration from many different sources, it
can be concluded that ion concentrations in sweat changes considerably from one
experiment to another. As a result, our data as well as data in the literature are not
completely comparable. In order to improve sweat collection analysis for the future,
many more trials must be performed under the same conditions, and all data from various
experiments need to be pooled and directly compared.


Dr. Henry R. Drott, Chief Investigator for Cystic Fibrosis Sweat Collection at the
       Children’s Hospital of Philadelphia

Gia Hchevia, Biotechnologist and Researcher at Hospital of the University of


[1] Quinton, Paul M., “Physiology of sweat secretion”, Kidney International, Vol. 32,
      Suppl. 21 (1987), p. S-102 – S-108.

[2] Fox, Stuart Ira, Human Physiology, 5th ed., Wm. C. Brown Publishers: Boston. 1996.
       p. 16.


[4] Fox, Stuart Ira, Human Physiology, 5th ed., Wm. C. Brown Publishers: Boston. 1996.
       p. 12.

[5] Bijman, Jan, “Transport process in the eccrine sweat gland”, Kidney International,
       Vol. 32, Suppl. 21 (1987), p. S-109 – S-112.


[7] Sato, Kenzo, Cary Feibleman, and Richard L. Dobson, “The Electrolyte Composition
       of Pharmacologically and Thermally Stimulated Sweat: A comparative study”,
       The Journal of Investigative Deratology, Vol. 55, No. 6 (1970), p. 433-438.

[8] Fox, Stuart Ira, Human Physiology, 5th ed., Wm. C. Brown Publishers: Boston, 1996.
       p. 132.


[10] Fox, Stuart Ira, Human Physiology, 5th ed., Wm. C. Brown Publishers: Boston,
       1996. p, 136.




[14] Brisson, G. R., P. Boisvert, F. Péronnet, H. Perrault, D. Boisvert, and J. S. Lafond,
      “A simple and disposable sweat collector”, European Journal of Applied
      Physiology and Occupational Physiology, Vol. 63 (1991), p. 269-272.

[15] Boisvert, P., K. Nakamura, S. Shimai, G. R. Brisson, and M. Tanaka, “A modified,
      local sweat collector for warm and humid conditions”, European Journal of
      Applied Physiology and Occupational Physiology, Vol. 66 (1993), p. 547-551.

[16] Taylor, Richard P., Adrian A. Polliack, and Dan L. Bader, “The analysis of

       metabolites in human sweat: analytical methods and potential application to
       investigation of pressure ischaemia of soft tissues”, Annals of Clinical
       Biochemistry, Vol. 31 (1994), p. 18-24.

[17] Gibson, Lewis E. and Paul A. di Saint’Agnese, “Studies of salt excretion in sweat”,
      The Journal of Pediatrics, Vol. 62, Number 6 (1963), p. 855-867.

[18] Sato, Kenzo and Fusako Sato, “Na+, K+, Cl-, and Ca2+ concentrations in cystic
       fibrosis eccrine sweat in vivo and in vitro”, Journal of Laboratory Clinical
       Medicine, Vol. 115, Number 4 (1990), p. (504-511).


To top