In Clinical and Experimental Therapeutics

Current Issue
Previous Issues
Reprint Information
Back to The Journal of Applied Research
Click here for information on how to order reprints of this article.

Heat Tolerance in Patients with Type I and Type II Diabetes

Jerrold Scott Petrofsky, PH D*

Chris Besonis, BS, ATC

David Rivera, BS, ATC

Ernie Schwab, Ph D*

Scott Lee, MD*


*Department of Physical Therapy, Internal Medicine

Loma Linda University

Loma Linda, California

Department of Physical Therapy

Azusa Pacific University

Azusa, California


KEY WORDS: thermoregulation, diabetes, heat stress, sweat


Six control subjects and eight subjects with type 1 and type 2 diabetes were studied to understand the relationship between skin temperature, central body temperature, and sweat rate. The results of the experiments show that for all diabetic subjects (both type 1 and type 2) heat tolerance was poor. In fact, with a 30 minute exposure to an environmental temperature of 42˚C, even though subjects were at rest, central body temperature increased 1˚C more than that of controls. Further, skin temperature also increased. The reason for the increase in skin temperature and central body temperature appeared to be a failure of sweating. Sweating was lower at any skin temperature or at any skin location in diabetic compared with control subjects. Thus both Type 1 and Type 2 diabetic subjects were more susceptible to heat stress. This could have significant implications for heat disorders such as heat stroke for individuals with diabetes.


Diabetes is a major health care problem in the United States and around the world, affecting millions of people every year.1 Diabetes varies with race. For example, it has been estimated by the Centers for Disease Control2 that in the United States 7.8% of whites have diabetes whereas 10.8% people of African descent have diabetes. In the Mexican-American population, 10.8% of men and women were reported to have diabetes, whereas Pacific islanders have an incidence of over 15% for men and women. The risk for heart disease is twice as high in patients with diabetes. Diabetes is the leading cause of blindness, end-stage renal disease, and nontraumatic amputation.2 There are two basic types of diabetes: type 1 diabetes is an autoimmune disease in which the immune system attacks insulin-producing beta cells in the pancreas and destroys them. The most common type of diabetes, however, is type 2 diabetes, which is characterized by high insulin resistance.3

One of the causes for the increased risk of nontraumatic amputation seen in diabetic patients results from underlying damage to the microcirculation and sympathetic system found in the skin.4 Damage to the capillary bed in the skin can result in damage to the sympathetic cholinergic nerve fibers that innervate the sweat glands.5 This lack of circulation can result in severe damage to the sweat glands that innervate the skin.6 In some cases, however, damage to these sympathetic neurons can be reversed with exercise or hyperbaric oxygen.7 This type of sympathetic dysfunction has been described very early in the course of diabetes and has been seen in children with type 1 diabetes even before the onset of overt clinical symptoms.8

Sweat gland destruction has been reported to result in hypersweating in some areas of the body and no sweating in others.9 It can be anticipated then that damage to sweat glands would have a profound effect on the the thermoregulatory ability of patients with diabetes. Yet studies performed in heat chambers evaluating changes in skin have not been well controlled for important aspects such as central core temperature, nor have they been performed to quantify and isolate specific regional sweat rates in patients with diabetes. This was the purpose of the present investigation.



The subjects in this study were 4 patients with type 1 diabetes, 4 patients with type 2 diabetes, and 5 control subjects with no history of diabetes. The average age, heights, and weights are listed in Table 1. All procedures were approved by the committee on human experimentation, and all subjects signed a statement of informed consent. Three of the subjects with diabetes had impaired sensation in their feet on examination, and 5 had normal sensation. There was no report of any strength deficit in any muscle group in these subjects. Four of the 8 diabetic subjects were engaged in aerobic training such as walking or running on a daily basis. The other subjects were moderately active.


Sweat was measured by a sweat hygrometry system (Figure 1). The system is composed of a source of compressed dry air, a flow meter, sweat capsules, and a hygrometer. Compressed air at a pressure of approximately 1000 mm Hg was stored in a 30-L pressure reservoir. The output of the air was regulated through a regulator and flow meter such that an output of 50 mL/min flowed through individual sweat capsules applied to the skin (Figure 2). Each capsule was 2 cm2 in diameter and was round. The capsule had a cavity inside and was held to the skin by an elastic strap. The internal capsule had connections to two hoses. The air inlet provided a source of dry air and the air outlet was connected to a manifold and into a hydrometer system. The hydrometer consisted of a sensor using calcium chloride to measure the humidity in the air. The greater the humidity, the lower the electrical resistance of the calcium chloride. Therefore, an electrical output was provided proportional to relative humidity. By knowing the humidity, the gas temperature, and the flow rate of air across the sweat capsules and the cross-sectional area of the capsule, we could calculate the sweat rate per square centimeter of the skin. This technique has been described previously.10


The subjects were led into an environmental room in which the room temperature was initially at 22˚C. The subject was placed in the supine position on a tilt table. The subjects rested comfortably for a 10-minute rest period after which sweat rates, core temperature, and skin temperature were assessed. Sweat rate was assessed on the forehead, chest, forearm, and leg. Skin temperature was measured just distal to the knee. The table was the tilted at a rate of 200˚/min to an elevation of 45˚, and all measurement were repeated. After 3 minutes, the table was tilted back to horizontal, and the measurements were repeated again. The room temperature was then elevated to 42˚C. A light blanket was placed over the subject. Sweat rates, core temperature and skin temperature were assessed at 10, 20, and 30 minutes of exposure. At 30 minutes, all measurements before the tilt to vertical were repeated. The table was placed at 45˚ and the measurements were repeated again. One final set of measurements was taken with the table in the horizontal position (Figure 3).


The results of the determinations of sweat are shown in Figure 4, which illustrates the average sweat on the forehead (A), chest (B), calf (C), and forearm (D). Each graph shows the mean results of all subjects. In panel A, for example, each point illustrates the mean of all five control subjects. For the diabetic subjects, no statistical difference was seen in any sweat rates when comparing the rates for subjects with type 1 and type 2 diabetes. Therefore, for the purpose of presentation, all results from diabetic subjects have been averaged together and the mean response is shown on panels A, B, C, and D, respectively, for the forehead, chest, calf, and forearm. The forehead is a typical example of the response of the other areas of the body to sweating. For example, in panel A of the figure, the resting sweat rates in the cool environment were 0.38 ± 0.14 mg/cm2/min for the control subjects but only 0.2 ± 0.06 mg/cm2/min for the subjects with diabetes. The higher resting sweat rates of the control subjects was statistically significant (P < 0.01) compared with the patients with diabetes. There was a striking similarity in the sweat rates in all diabetic subjects.

The highest sweat rate recorded was in a patient with type 1 diabetes whose resting sweat rate was 0.24 mg/cm2/min. Most other subjects had sweat rates in the order of 0.14 to 0.19 mg/cm2/min. For the diabetic subjects, tilting the table from the horizontal to the vertical position had very little effect on sweat rates, whereas there was a slight diminution in sweat rates in the control subjects. The greatest increase in sweat rates was found during increasing the environmental temperature. In the warm room, sweat rates increased progressively. For example, for the subjects with diabetes, 10 minutes into heat exposure, sweat rate had increased to 0.32 ± 0.16 mg/cm2/min and at the 30-minute point resting sweat rate was 0.46 ± 0.16 mg/cm2/min. For the control subjects, sweat rates not only started higher but at the end of the 30-minute period, the resting sweat rate was 0.84 ± 0.48 mg/cm2/min. The difference in sweat rates between the controls and subjects with diabetes was statistically significant.

There was a tendency of both groups of subjects for sweat rates to drop when the table was tilted from the horizontal to the vertical posture. However, the difference in sweat rates were not statistically significant when changing body positions in either the cool or the warm room. The average sweat rate then for all areas in the cool environment for the control subjects was 0.36 mg/cm2/min compared with 0.21 mg/cm2/min for subjects with diabetes. With heating, sweat rates for the control subjects increased to an average of 0.81 mg/cm2/min. For the control subjects, the greatest sweat rate was 0.44 mg/cm2/min. In other words, control subjects were able to sweat at least at twice the rate of the subjects with diabetes.

Perhaps even more telling are the results shown in figure 5. This figure shows the average skin and core temperatures in the control and diabetic subjects at rest and throughout heat exposure. For the control subjects, average skin temperature was 32.95 ± 0.6˚C. During the 30 minute heat exposure, skin temperature increased to an average to 33.7 ± 1.05˚C. In contrast, the subjects with diabetes started with a resting skin temperature in the cool environment of 33.4 ± 1.0˚C. At the end of the 30-minute heat exposure, the average temperature was 34.5 ± 0.77˚C, an increase of 1˚C over the resting values. This increase was significantly higher than that of the control subjects (P < 0.05).

The same relationship was seen with core temperature. Resting core temperature of the control group was 36.7 ± 0.27˚C which increased at the end of the 30-minute heat exposure to 36.9 ± 0.26˚C. In contrast, for the subjects with diabetes, resting temperature started at 36.5 ± 0.18˚C and increased to 37.5 ± 1.1˚C. The increase in core temperature also averaged about 1˚C higher than in control subjects. It is not surprising then, as shown in Figure 6, that when the average sweat rate was plotted against the body temperature that the diabetic subject’s body temperature continued to increase with a small increase in sweat rate, but body temperature was maintained in the control because a very small increase in body temperature caused a very large increase in sweat rates.


The concept of damage to skin circulation and sweat glands in patients with diabetes is not new.11 However, previous studies analyzed sweating in patients with diabetes by painting the patient with corn starch and looking for either large amounts of sweat production or no sweat production in patchy areas.6 Previous researchers did not quantify the absolute amount of sweat or see how it compared with sweat in control subjects. In the present investigation, rather than simply warming the patient under an electric blanket, patients were exposed to a neutral and hot environment lying horizontal on a tilt table, and sweat rates were measured in relation to skin and central body temperature. Generally speaking, central body temperature is more accurate if tympanic or rectal temperature is used. However, oral temperature is acceptable for patients resting. With exercise, large amounts of ventilation can keep oral temperatures from producing the proper values when exercising in a cool environment. However, because subjects in this study were resting, their metabolic rate was low and oral temperatures were acceptable for measuring trends for central core temperature.

In our study, we were able to expand on these early experiments and isolate different areas of the body. We show dysfunction in sweat rates in all the areas of the body we assessed. On the calf during heating, for example (Figure 4C), although sweat rates were similar in control and subjects with diabetes, all eight subjects with diabetes had lower sweat rates. In other areas, such as the forehead, there was a greater difference in sweat rates than in the calf or forearm. Although earlier studies6 suggested that problems with sweat glands resulting in the lack of sweating are only found in only some areas of the body, our present data clearly show nonselective general damage to all areas of the body associated with diabetes. Furthermore, the damage that occurred did not appear to correlate with the duration of diabetes or type of diabetes. Thus it would appear that damage to the sweat glands with diabetes seems to be a great deal more widespread than originally thought and, similar to previous data, appears to occur early in the course of both type 1 and type 2 diabetes.

The overall consequence of this damage to the sweat glands resulted in significant thermoregulatory dysfunction. Earlier studies did not directly assess the effect that damage to the sweat glands can have on thermoregulation. However, our present investigation showed a clear correlation between abnormal noncompensatory rises in skin temperatures resulting in inappropriate raising of core temperature in subjects with diabetes. For example, in figure 6, as core temperature increased, control subject sweat rate increased proportionately to maintain core temperature constant. However for subjects with diabetes, sweat seemed to plateau irrespective of an alarming rise in core temperature. Thus, the diabetic subjects’ generalized inability to sweat across the body had a profound effect on core temperature. Extending this graph past 37.6˚C, it is not hard to visualize that with exercise in the heat sweat rates probably will not exceed 0.5 mg/cm2/min and body core temperature could rise to dangerous levels in diabetic subjects. It must be noted that in the present investigation, subjects rested passively in the heat. With any light exercise in the heat or with exposure to heat and high humidity, patients with diabetes could be extremely impaired. Further study is warranted to exactly quantify the relationship between sweat gland activity and thermoregulatory impairment with exercise in these subjects.


1. Ziegler D: Diabetic cardiovascular autonomic neuropathy: Prognosis, diagnosis and treatment. Diabetes Metab Rev 10:339–383, 1994.

2. CDC: Statistics on Incidence of Diabetes. Washington, DC: Centers for Disease Control and Prevention; 2002.

3. Manson JE, Rimm EB, Stampfer MJ, et al: 

4. Rendell M, Bamisedun O: Skin blood flow and current perception in pentoxifylline-treated diabetic neuropathy. Angiology 43:843–851, 1992.

5. Robertshaw D: Neuroendocrine control of sweat glands. J Invest Dermatol 69:121–129, 1977.

6. Fealey RD, Low PA, Thomas JE: Thermoregulatory sweating abnormalities in diabetes mellitus. Mayo Clin Proc 64: 617–28, 1989.

7. Tesfaye S, Harris ND, Wilson RM, Ward JD: Exercise-induced conduction velocity increment: A marker of impaired peripheral nerve blood flow in diabetic neuropathy. Diabetologia 35:155–159, 1992.

8. Meh D, Denislic M: Subclinical neuropathy in type I diabetic children. Electroencephalogr Clin Neurophysiol 109:274–280, 1998.

9. Gundersen J: Editorial: Blood pressure and blood flow in peripheral circulation Ugeskr Laeger 136:1861–1862, 1974.

10. Mitchell et al: 1976.

11. Accurso V, Shamsuzzaman AS, Somers VK: Rhythms, rhymes, and reasons: Spectral oscillations in neural cardiovascular control. Auton Neurosci 90:41–6, 2001.

Recommended Reading

    Bagg W, Ferri C, Desideri G, et al: The influences of obesity and glycemic control on endothelial activation in patients with type 2 diabetes. J Clin Endocrinol Metab 86:5491-5497, 2001.

    Hennekens CH, Speizer FE: Physical activity and incidence of non-insulin-dependent diabetes mellitus in women. Lancet 338:774-778, 1991.



Table 1. General Characteristics of Subjects


                       Age        Height     Weight         With

                        (y)          (cm)         (kg)         Diabetes


Type I               56        177.17      85.05            25

Type II              61        176.48      91.26           5.04

All Diabetes     58        176.82      88.15          15.02

Controls           37        176.59      80.38            n/a


Figure 1. This figure shows a schematic diagram of the experimental set up for measuring sweat rate. Pressurized air enters a capsule containing Dri-rite, a compound that removes all humidity from air. The dried air is then regulated through a precision flow meter and enters a small closed capsule on the surface of the skin. Air leaving the capsule, which gains humidity from the skin, then enters a humidity sensor. The resistive output humidity sensor is digitized through an A/D converter, and sweat rate is then calculated through a digital computer. Integrated with the humidity sensor is an air temperature sensor to calculate the temperature. This is used in humidity calculations to determine water content of the air.


Figure 2. This figure shows one of the sweat capsules used in these studies.


Figure 3. A subject in the experimental set up.


Figure 4a.


Figure 4b.


Figure 4c.


Figure 4d.


Figure 4. Sweat rates in the control and diabetic groups on the forehead (A), chest (B), calf (C), and forearm (D). The diabetic group averages all sweat rates determined in diabetic subjects with type 1 and type 2 diabetes as one large group. Each point represents the mean of eight subjects. The control group includes five control subjects, and each point represents the mean of the five control subjects. Sweat rates were measured with the patient in a cool room in the horizontal position (H), when changed from the horizontal to the vertical position (V), and back to the horizontal position (H). Sweat rates were also measured at 10 and 20 minutes into exposure to a 40˚C environment. After exposure, patients were again subjected to measurements of sweat in the horizontal (H), vertical (V), and back in the horizontal position (H).


Figure 5. The average skin and core temperature measurements in diabetic and control subjects measured in the cool room (1, 2, 3) during heating at 10, 20, and 30 minutes into heat exposure (4, 5, 6) and during transitions from the horizontal to vertical to horizontal position after exposure to the hot environment (6, 7, 8). Each point represents the mean of 8 subjects with type 1 and type 2 diabetes and five control subjects.


Figure 6. This figure relates the average whole body sweat rates for all control and diabetic subjects to the body temperature recorded during heat exposure and in the cool environment. Each point represents the mean of all subjects in the group.

©2000-2013. All Rights Reserved. Veterinary Solutions LLC is an authorized retailer for The Journal of Applied Research