History of Medicine

OUR “NORMAL” TEMPERATURE 

     No one today thinks twice about taking the temperature of someone presenting with illness, but that wasn’t always so. Physicians in the U.S. rarely recorded temperatures before, and even during, the Civil War. World-wide the practice was much the same. What made the change?

     Human temperature seems to have been first measured by the remarkable Sanctorio Sanctorius, professor at the University of

Sanctorio Sanctorius (from Wikipedia)

Padua in the early 1600s. In addition to weighing intake and output to measure insensitive losses, he measured temperatures using a graduated thermoscope, an open tube sensitive to atmospheric pressure as well as temperature. Closed tubes, to eliminate atmospheric pressure effects, were soon invented. Then in 1714 Daniel Gabriel Fahrenheit, a skilled instrument maker working in Holland, fashioned a mercury-filled closed tube thermometer with a new scale that proved quite accurate. He

Daniel Fahrenheit (from Wikipedia)

picked zero as the temperature of a mixture of ice, water, and sea-salt, 32 of water and ice alone, and boiling water measured 212. The astronomer Anders Celsius, in 1742, brought back the centigrade scale first promulgated by Christiaan Huygens in the previous century. Both scales continued in use.

       Thermometry still had no sex appeal, however. Of a few prominent medical men who adopted it in their practice one stands out: Anton de Haen, a former pupil of Hermann Boerhaave. As professor of medicine at the University of Vienna from 1754-76, de Haen recorded the levels and diurnal variations in temperatures of normal and diseased subjects, correlated high temperatures with high pulse rates, and noted the association of rising temperatures

Anton de Haen (Wellcome Library)

 with chills. But most clinicians saw no practical value in measuring temperatures. Concepts of the mechanisms of heat production and regulation, both in physics and biology, were not developed until the early 19^th century, and fever was still a disease, not a symptom.

     The person most influential in bringing the medical world around to using the thermometer was Carl Wunderlich. Wunderlich studied medicine at the University of Tübingen, graduating in 1837. Grumbling about the backward status of German medicine at the time he teamed up with two of his schoolmates, Wilhelm Griesinger and Wilhelm Roser, to literally usher in a new era. The three founded a new journal, the “Archiv für Physiologische Heilkunde” (Archives of Physiologic Medicine, or Medical Science) to emphasize the importance of physiologic investigation in addition to clinical observation. It was the beginning of a golden age in German medicine, assisted by government subsidies and attracting students from around the world.

     Wunderlich ended up as professor of medicine at the Univ. of Leipzig, where he conducted his temperature studies (and where clinical physiology achieved world renown). With Germanic

Carl Wunderlich (National Library of Medicine)

thoroughness he recorded around-the-clock temperatures of some 25,000 patients, collecting over a million measurements. Both normals and the sick of all ages were included. The work culminated in a book, Das Verhalten der Eigenwärme in Krankheiten (On the temperature in Diseases), that exerted wide influence.

    Wunderlich used a mercury bulb thermometer and preferred temperatures taken in “a well-closed axilla” over oral or rectal temperatures. His work established the now familiar “normal” average temperature of 37C (98.6 F) degrees, with oscillations seldom exceeding 0.5^o C each way. He demonstrated rather typical fever patterns for various diseases, such as typhoid, typhus, relapsing fever, smallpox, measles, etc., thus using the thermometer as a diagnostic aid. Fever patterns, especially high temperatures, could also help with prognosis.

     In spite of the influence of the work, there were problems. Statistical methods were not well developed at the time and Wunderlich’s measurements are not tabulated, but rather summarized. More important, how accurate were the

English version of Wunderlich's book
second edition (Hathi Trust)

measurements? Wunderlich himself states, “Errors that do not exceed half a degree Centigrade are hardly worth mention”, suggesting that high precision was not a priority.

     And another question: is 37^oC really the “normal” average temperature? In 1992 a group of 148 normal subjects aged 18 to 40 were studied over three days with modern Diatek electric oral thermometers (700 readings). Their mean (and median) temperature turned out to be 36.8^oC +/-0.4 degrees (98.2^oF +/- 0.7 F).

     Why the difference? Wunderlich preferred axillary temperatures, known to be lower than oral ones. He used glass thermometers with a mercury column, most about 12” long, that could take 15 to 20 minutes to equilibrate – frustrating to a busy nurse. Another possible source of error is calibration of the thermometers. The first internationally accepted temperature scale was not established until 1897, ten years after Wunderlich’s death. We cannot test Wunderlich’s own instrument but a thermometer belonging to one of his students is available (now in the Mütter Museum in Philadelphia). It is 22.5 cm long (about 9”), and in a water bath it gave readings 1.6^oC to 1.8^oC higher than modern electronic thermometers, partially offsetting the lower readings formerly obtained in the axilla. Physical changes from long storage, such as change in bulb size, were thought to play a minor role in such a difference. Finally, as mentioned, Wunderlich did not worry about less than a half degree variance in measurement.

     So we are “cooler” than we thought, though not by much. But Wunderlich’s work, in spite of some inaccuracies and fuzzy statistical methods, established thermometry as a permanent clinical practice.

SOURCES:

Wunderlich, W A. On the Temperature in Diseases: A Manual of

      Medical Thermometry. Eng trans by W B Woodman, London,

      1871.

Mackowiak, P and Worden, G. “Carl Reinhold August

       Wunderlich and the Evolution of Clinical Thermometry”.

      1994. Clin Infect Dis v18(3) pp 458-67.

Dominguez, A, et al. “Adoption of Thermometry into Clinical

       Practice in the United States”. 1987. Rev Infect Dis v9: 1193-

      1201.

Mackowiak, P A, et al. “A Critical Appraisal of 98.6^oF, the Upper

       Limit of the Normal Body Temperature, and Other Legacies

       of Carl August Wunderlich”. 1992. JAMA v268: 1578-80.

Gershon-Cohen, J. “A Short History of Medical Thermometry”.

       1964. Ann N Y Acad Sci. v121:4-11.