Normal body temperature

The pandemic has made everyone aware of the need to monitor their body temperature to see if they might be carrying the virus. I wrote earlier about how my ‘normal’ body temperature when I have no symptoms of anything is much lower than the canonical value of 98.6F, which makes it hard to determine how to use my body temperature to determine if I am sick or if I have returned to normal after being sick. This article explains that there is a huge amount of variation because a multiplicity of factors affect the temperature.

Many trace this number back to an 1871 book on medical thermometry by a German physician named Carl Reinhold August Wunderlich who measured the underarm temperatures of thousands of people. Readers seem to have taken away the monolithic 98.6 number, but Wunderlich in fact wrote that based on his measurement “healthy persons under varying conditions” had temperatures that ranged from 97.16 F to 100.4 F.

Even accounting for that variation, a recent study suggested that, across the population, core temperatures have declined about 0.6 F (0.3 C) below the traditionally accepted levels and about 0.85 F (0.5 C) compared to the end of the 19th century and beginning of the 20th century.

The researchers acknowledge these observed reductions could be the result of changes in equipment and measurement technique. But they suggest the reduction could also reflect physiological shifts in metabolic rate and reductions in inflammation related to underlying medical conditions that people manage more effectively now than in the past.

Your body’s temperature is based on the heat from your energy production. Digesting food, physical activity and infection increase the metabolic rate and can all raise your temperature.

Babies may have slightly elevated normal body temperatures and caloric needs as a result of increased metabolic rate to support growth and development, while seniors may have slightly lower body temperatures due to lower rates of cellular growth and repair.

In addition to changes that occur as the result of age, everyone has variation throughout the day based on their circadian rhythms. Both the 1871 book by Wunderlich and more recent research have demonstrated that people typically experience their lowest body temperatures overnight. Similarly, body temperature will fluctuate through the varying phases of females’ menstrual cycle.

So basically, it seems like each of us has our own ‘normal’ temperature that we can use as a baseline, that is obtained by measuring at the same part of the body using the same device at the same time of day.


  1. blf says

    I’m another individual who is in or perhaps a tad below the “low” end of the “normal” temperature range. It’s never (knowingly) caused me any problems, but I do start to wonder what’s going if it climbs into the mid- or upper-“normal” part of “the” range… mostly, it seems, just being hot from exercise or ambient temperature.

  2. DrVanNostrand says

    I recently had a situation where I had my temperature taken about 2x a week for about 2 months. I also typically run a little low, 97.5 -- 98 F. I wonder if that’s why I hate hot weather…

  3. Pierce R. Butler says

    In my experience, most US-pharmacy-available digital body thermometers -- the only kind available since mercury thermometers went off the market -- are cheaply-made (if not -sold) crap that stop functioning after a few months.

    Even if I could afford to replace them continually, I wouldn’t trust their accuracy, or the consistency of a record based on a series of mediocre instruments.

  4. says

    It’s amazing how the paranoid act about thermometer “guns” that take your temperature without touching. I’ve heard ridiculous stuff (“kills brain cells!”), just as ridiculous as anti-vaxxer nonsense (and probably the same people).

    These “guns” are better, measuring from the forehead without touching skin, and not the ear or other places. My school uses them which allows us to process dozens of kids in a couple of minutes.

  5. Bruce says

    In almost every country in the world, normal body temperature is 37 C.
    That is about 97 to 100 F.

    If a mathematically illiterate person thinks that 37 is the same as 37.0, then they will think that “normal” body temperature is 98.6 F.

    The fact that a math average exists does not imply that it has 0.1% error bars.
    If a non-digital thermometer is used, nobody expects there to be much meaning in differences between numbers that all round off to 37C.
    The fact that we now have digital thermometers does not mean that human variation is magically reduced suddenly.

    If a pizza is cut into six slices, and a calculator exists, that does not mean that each slice is 16.6666667% of the pizza, with 9 digit precision.
    Part of understanding how to use numbers is understanding the limits of the application of those numbers to objects in the real world.

    I don’t think anyone on this blog is confused about any of this, but I think most times we hear others speak of 98.6 F, this is affecting their thinking.

  6. Dunc says

    Also, you need to spend a reasonable chunk of change to get a thermometer that’s actually accurate to better than +/- 1°C. Sure, you can pick up a cheap digital that will read to 0.1°C, and it will read repeatably, but it’s unlikely to be accurately calibrated except by luck. And then there’s the question of keeping it calibrated… If it doesn’t have a current calibration certificate, it’s not that accurate.

  7. anat says

    Bruce, back in the day of mercury thermometers my mother taught me to treat 36.6 C as ‘healthy’ and 37.0 C as ‘not healthy enough to be up and about’. Of course we never took temps unless we suspected illness or were on the recovery from illness, so I don’t know how much measurable variation there really was.

  8. says

    Back when I was a child I was taught that 36.6 C is normal. Anything above 37.0 C meant that you were sick. My normal body temperature is about 37.1 C. Back when I was a teen I worried about whether this could mean that I had some chronic illness.

  9. Sam N says

    I’m skeptical of many of the claims regarding thermometer reliability here. I have a cheap $20 model, and it reliably measures me between 37.1 and 37.5 C. I have no idea of it’s accuracy, but unless I have wild fluctuations in temperature, it appears to be precise.