Lower diagnostic threshold for fever in older people may help detect COVID-19 cases earlier

By Dr. Liji Thomas, MDFeb 2 2021

In the current coronavirus disease 2019 (COVID-19) pandemic, fever has been noted to be among the typical symptoms reported by individuals with symptomatic infection. The normal body temperature ranges throughout the day, but is on average about 37°C (98.6°F), varying up to 0.25 to 0.5°C from day to day.

Temperature regulation in humans is the function of multiple body systems, including the heart and lungs, the respiratory system, and the muscles. The role played by individual biological factors, such as genetics, is not well known.

A new preprint on the medRxiv* server discusses heritable factors, age and body mass on temperature variation between humans. One notable finding is that the current threshold for detecting fever in older adults may be too high, precluding the ability to detect COVID-19 early on.

Study: Individual factors underlie temperature variation in sickness and in health: influence of age, BMI and genetic factors in a multi-cohort study. Image Credit: Yuganov Konstantin / Shutterstock

This news article was a review of a preliminary scientific report that had not undergone peer-review at the time of publication. Since its initial publication, the scientific report has now been peer reviewed and accepted for publication in a Scientific Journal. Links to the preliminary and peer-reviewed reports are available in the Sources section at the bottom of this article. View Sources

Defining fever

Fever is a response to infection, but has been defined in varying ways, including a single temperature reading of >37.8°C (>100°F) in the mouth, repeated readings >37.2°C (>99°F) in the mouth, or rectal temperatures >37.5°C (>99.5°F); or simply, having a rise in temperature of >1.1°C (>2°F) above baseline.

Age is known to be associated with a lower normal body temperature, and older people are less able to cope with the extremes of temperature. The observed reduction in basal body temperature with age may be due to the waning function of the systems involved in thermoregulation, as well as inadequate homeostasis in response to changes in the temperature of the surrounding environment.

Fever thresholds may miss older people

These reference standards are too high for older people, say many researchers. In fact, if a reading of >37.2°C is substituted in the first definition, for instance, more older people with fever would have been correctly identified, since an earlier study shows that in people aged 60 years or more, only 20-30% show temperatures of 37.8°C or above in the presence of infection.

The reasons for this difference in presentation could be immunosenescence, or the waning of immune function with age. Both innate and adaptive immunity are weakened, which may explain why older adults are more often and more severely affected by infections.

Both T cell and neutrophil function are affected by aging. The poor immune function in turn affects pyrogen production and hence the occurrence of fever.

In COVID-19 in the elderly, atypical presentations are recorded, such as weakness and headache, or delirium, without fever, or with hypothermia (temperature <36.0°C). The current study aims to understand how age is related to temperature in a healthy group and a group with viral infection, so as to define a more sensitive fever threshold that is capable of identifying fever in older adults with infections, including COVID-19.

Study details

In the current study, the researchers examined the links between age, sex, body mass index (BMI) and basal temperature, as well as genetic factors influencing body temperature, in a group of healthy twin-pair volunteers from the TwinsUK registry. This was followed by comparing these associations with those present in community-based and hospitalized groups of COVID-19 patients with acute infection.

The researchers found comparable associations between age and temperature in healthy volunteers, and in both types of COVID-19 patient groups. As reported in multiple studies, older and frailer adults may not respond to infection with a significant rise in temperature, both because they have a lower basal temperature and because their immune system responds with a lower pyrogenic response to infection.

This could be mediated or contributed to by age-associated reductions in pro-inflammatory mediators following infection, such as interleukin-1 and tumor necrosis factor. As a result, other signs of infection should be given due weight, even in the absence of fever.

Sources of variability

An earlier twin study was carried out in 53 twin-pairs, all adult females. Using continuous wrist temperature-sensing monitors, the researchers showed that the circadian system-dependent variation in temperature, as well as many other physiological parameters, was more closely correlated within pairs of monozygotic twins than for dizygotic twins. Genetic factors contributed 46-70% of the variance.

In the current study, the researchers observed that a higher BMI is associated with higher baseline and fever temperatures. This seems to support the link between abnormally high body mass and inflammation.

The association between BMI and temperature was independent of age in both healthy volunteers and COVID-19 cases in the community. This is supported by the CoLaus study in men and postmenopausal women, where obesity and abnormal insulin metabolism were correlated with higher temperatures.

The current study also shows, from its analysis of correlations within monozygotic twin pairs, that the baseline temperature of an individual is partly determined by the genetic makeup. These make up 47% of variance after adjusting for age and sex. If BMI is also added in, heritability contributes 44% of variance.

The individual’s genetic constitution, in terms of heritable physiological factors such as vasomotor sweating, skeletal muscle responses and the perception of temperature with the resulting physical behavior, also affects temperature regulation. Thus, this also contributes to variations in temperature between different age groups and people with different BMI.

However, fever in infectious cases, as defined by the first definition (a temperature of 37.8°C or above), had no significant genetic component, indicating the primary role of environmental factors like infection in deciding the incidence of fever in a population.

Lower fever threshold

The study adds weight to previous observations and clinical recommendations that fever thresholds as currently defined are not able to pick up fever in older adults. If a temperature of 37.4°C is taken as the discriminatory threshold for fever in this group, of adults at or over 65 years of age, it will perform as sensitively as the usual threshold of 37.8°C in younger adults. In either case, the sensitivity is about 33%, and the specificity near 100%, for distinguishing the group of healthy volunteers from hospitalized COVID-19 patients.

What are the implications?

Baseline and infection-related temperatures increase with the BMI, but decrease with increasing age.

The associations between baseline and fever temperature and individual determinants, especially the BMI, imply that a lower temperature threshold should be used to define fever in older and thinner adults. This will help to raise timely clinical suspicion of infections, including COVID-19, leading to earlier testing, and isolation, of possible COVID-19 cases in this group in hospitals.

This is especially important for residential care facilities since most residents are older people. However, the observed link with community COVID-19 infections could be useful in helping to detect community cases and contain the spread of such infections.

This news article was a review of a preliminary scientific report that had not undergone peer-review at the time of publication. Since its initial publication, the scientific report has now been peer reviewed and accepted for publication in a Scientific Journal. Links to the preliminary and peer-reviewed reports are available in the Sources section at the bottom of this article. View Sources