Wunderlich’s measurements and the origin of “37°C”
In the 19th century Ferdinand von Wunderlich collected massive numbers of axillary temperature readings and reported an average of 37°C for healthy adults. He used glass mercury thermometers and measured under the arm, and his dataset became the implicit standard for generations. Those instruments, the measurement site, patient selection and clinical context all shaped the number in ways that matter for direct comparison with modern data.
How researchers discovered a long-term decline in average body temperature
Researchers re-analyzing historical records and modern clinical datasets found a steady decline in mean body temperature of roughly 0.03–0.06°C per decade over the last 150 years. Studies that control for age, sex, measurement method and time of day still show a downward trend, indicating the change is not just an artifact of different thermometers.
Three concrete drivers: infections, environment, and measurement
First, chronic and repeated infections were far more common in the 19th century; Persistent immune activation raises baseline metabolic rate and heat production. Second, improved housing, central heating and clothing changed thermal exposures so bodies expend less energy for temperature regulation. Third, earlier thermometers and axillary placement yield different absolute values than modern oral or digital devices. Each factor explains part of the decline; Together they produce the consistent signal seen across datasets.
Quantified examples and local case studies
A landmark analysis of US data spanning 1862–2017 estimated the mean adult temperature around 36.39°C in recent decades versus roughly 36.98°C in the 19th century—about a 0.59°C drop for men in that sample. A community-level example: the Tsimane population’s mean fell from ~37.0°C to 36.5°C across two decades of socioeconomic change, matching improvements in infection control, footwear, hygiene and vaccination.
Which temperatures should treat as red flags today?
Instead of treating a single number as the universal cut-off, should prioritize ranges with clear clinical implications. Values below 35°C generally indicate severe hypothermia or shock and require urgent evaluation. Values above 40°C suggest life-threatening hyperthermia, severe infection or dysregulated inflammation. Small deviations within the 36–37°C band usually warrant correlation with symptoms and the patient’s known baseline rather than automatic escalation.
Practical measurement guidance: make readings comparable and actionable
Follow these steps for reliable tracking:
Step 1: Choose a consistent method for longitudinal monitoring (same device type and same measurement site).
Step 2: Calibrate digital thermometers periodically and prefer validated oral or rectal devices when precise core estimates are needed.
Step 3: Record context (time of day, recent activity, antipyretic use, ambient temperature) to interpret single readings against the person’s baseline.
Why lifestyle and built environment matter
Widespread central heating and insulated housing reduce daily cold stress and lower metabolic heat production. Similarly, decreased outdoor exposure and changes in clothing norms narrow thermal variability. These behavioral and infrastructural shifts reduce baseline energy expenditure and therefore lower average measured body temperature at the population level.
Measurement bias: what to watch for when comparing studies
Axillary (underarm), oral, tympanic and rectal readings do not map 1:1. Axillary readings are typically lower than oral or rectal; older mercury thermometers may have different calibration curves than modern digital probes. When you compare studies, adjust for measurement site and device type or analyze only within-method subsets to avoid conflating method-driven differences with physiological change.
Research gaps that matter for future guidance
Design a prospective, multi-country cohort that uses standardized, calibrated devices across diverse climates and socioeconomic settings. Include repeated measures across seasons, record infectious disease indicators, inflammatory biomarkers, indoor environment data and clothing/behavior logs. Such studies can quantify how much of the temperature decline is due to lower infection burden versus altered thermal exposure or metabolic changes.
Actionable takeaways for labor and people tracking fever
Use a consistent thermometer and measurement site for follow-up. Treat extreme values (<35°C or >40°C) as clinical priorities. For small deviations, compare against the person’s baseline and the clinical context rather than an immutable “normal.” Understand that the historical 37°C reference may overstate contemporary population averages by several tenths of a degree.