Miscarriages and maternal temperature
In other sections, observations from a variety of sources suggest that the 7cv constraint is connected to the requirement in mammals at a particular period in development to occur in a narrow temperature range. This chapter will examine the range of maternal temperatures observed in humans and incidence of miscarriages.
The developing embryo is exposed to the core temperature of the mother. Only recently have techniques been developed to make such measurements non-invasively ( ). Published temperatures that rely on surface or peripheral measurements likely underestimate core temperatures and overestimate core temperature variability.
Average body temperatures, not specifically of pregnant women and in temperate climates in first-world nations, have been reported to vary both daily (typically 0.9oC) and seasonally (around 0.2oC warmer in summer than winter), and thus has a maximum of about 0.6oC above the long-term average.
Studies of pregnant animals suggest that hyperthermia is teratogenic (producing birth defects) with a threshold about 1.5oC above normal body temperature (Graham et al. (1998)). Such effects have frequently been attributed to the production of heat shock proteins.
Extensive studies of women’s temperature (usually measured before rising) have been carried out in connection with identifying when ovulation occurs and to determine if a pregnancy is likely. A temperature increase (around 0.5oC) over a few days may indicate ovulation and continued elevated temperature occurs if a pregnancy has occurred, which continues through the first trimester, slowly decreasing in the last two trimesters ( ). Combined with seasonal and daily variability, early in pregnancy average maximum temperatures may exceed the long-term average by at least 1oC.
In the later part of pregnancy, heat exchange in the uterus may actually keep the fetus temperature 0.5oC below maternal core temperature ( )
Is there any evidence of variation in seasonal variation in losses of pregnancy that might suggest a relationship to ambient temperature?
One problem with all the studies relating miscarriages and 7cv ribs is that they are all retrospective and it is unclear how reflective the studied population is of pregnancy in general. Also, even if they contained data of the time of year the miscarriage occurred the numbers would likely not be large enough to determine whether there was seasonal variation. However, a few very large prospective studies have examined whether miscarriages (referred to as spontaneous abortions) have a seasonal variability.
Before examining those results, it might be useful to see what some problems might exist in relating them to the miscarriages associated with 7cv ribs. 7cv ribs in humans can only be observed in fetuses after about 14 weeks (in 30-50% of miscarriages (second trimester), stillbirths (third trimester), and neo-natal deaths). These are usually thought to be about 15% of total conceptions. Another 10-15% of conceptions are thought to result in miscarriages in the first trimester. If 7cv rib miscarriages did have a seasonal variation, they might be obscured by the larger number of unrelated miscarriages (assuming those to have no seasonal variation), or these unrelated miscarriages have could have a seasonal variation uncorrelated with those associated with 7cv ribs.
In the 7cv vertebrae hypothesis, the events causing the associated deaths occur within a short time after gastrulation (4 weeks after conception). If elevated seasonal temperatures were involved, elevated spontaneous abortions (at least those which could be linked to 7cv ribs), this would be for conceptions in June or July. If lower than average temperatures were involved, there would also be a peak for November or December conceptions. Even if a cause of spontaneous abortion occurred over a short period after conception, if the resulting deaths occurred throughout the second trimester, this would broaden and lower any seasonal peak.
Smith et al. (1980) used a national database of hospital discharge records in the United States covering 1971-1974 of spontaneous and induced abortions and live births. There was a seasonal variation in the number of spontaneous abortions as a function on month of conception with a maximum in November and December and a minimum for conceptions in August with a ratio of 1.16. But there was a similar pattern and size of seasonal variation in total conceptions, so that there was no evidence of a seasonal variation in the rate of spontaneous abortions.
Czeizel et al. (1984) reported on a very large study of all obstetrical institutions in Hungary over a ten-year period. Out of over 3 million pregnancies, eight per cent resulted in spontaneous abortions (loss of pregnancy in the first two trimesters). Still births (fetal death in the last trimester) were less than 1% of pregnancies. It should be noted that 37% of pregnancies were terminated by induced abortions (almost all in the first trimester and less than a tenth of which were for medical reasons).
After correction for seasonality in conceptions (no details given), there was a clear seasonality in the month in which the spontaneous abortions occurred most frequently 0ctober through December and a minimum from April through June. If there was a three-month delay between higher temperatures and abortions, this would correspond to July through September. The monthly variation in spontaneous abortions about the average could be fitted by a sine wave with the peak to trough being about 1/10th the average rate. In this case, the timing of the peak appeared to be consistent with what is predicted from the homeothermic 7cv rib hypothesis if higher temperatures were the cause, but not if lower temperatures could also be a cause.
A recent article (Wessellink 2022) examined the seasonal variability of miscarriages from a large prospective study in which the participants were recruited prior to pregnancy. In the 12 months following enrollment 6104 women reported a pregnancy, of which about 20% reported a spontaneous abortion. The risk was highest in late August with a peak/low ratio of 1.3 (95% confidence interval 1.1-1.6). However, much of the variation was due to abortions within 8 weeks after the last menstrual period, which had peak/low ratio of 1.4. Thus, the peak/low ratio was considerably lower for abortions after eight weeks.
It appears that none of these studies show any evidence that would be strongly supportive of the homeothermic 7cv rib hypothesis. Sometimes it is sometimes said “absence of evidence is not evidence of absence.” I think this is contrary to the Bayesian view of science where the proper use of observations is to change the relative likelihood of the truth of competing hypotheses, which in this case includes the null hypothesis. If the observational studies are of sufficient power that they would be expected to show a given result for a particular hypothesis, then absence of that result would support the null hypothesis.
What about other instances of exposures during pregnancies to abnormal extremes of temperature? Fevers and exposures in hot tubs come to mind.
The largest study (Nybo Amdersen et al. (1984) from the Danish National Birth Cohort, over 24,000 pregnancies) I have been able to find reported no association of fevers (reported in later interviews) in the first 16 weeks and subsequent fetal deaths. Fetal deaths were about 5% of pregnancies, considerably lower than reported in other studies. This is likely related to subjects being recruited at the first physician visit after recognition of the pregnancy at a median gestation time of about 12 weeks (with about a third of the subjects recruited in the second trimester). Thus, miscarriages which occurred without a prior visit are not included in this study.
Only 0.3% of pregnancies were associated with fevers and fetal deaths in the last two trimesters. This is much less than the incidences of fetal deaths associated with 7cv ribs in other studies. Even considering that many miscarriages were not included in this study, this study would suggest that fevers had little in any effect on inducing 7cv ribs and the deaths associated.
Ravanelli et al (2018) concluded that pregnant women can use a hot bath at 40oC for ? minutes or a dry sauna at 70oC for 20 minutes and maintain their core temperature within the safe range (<1.5oC increase above normal).
3/7/25