I have suggested that increased consumption of the Omega 6 polyunsatured fat (PUFA) linoleic acid throughout the 20th century has led to an American population where the majority of us are in a torpor-like state as defined by high levels of stored PUFA in our bodyfat coupled with high expression of the transcription factor PPAR gamma and it’s target gene SCD1. Symptoms of this torpor state include low body temperature, low metabolic rate and increased fat storage. I have previously shown that the consumption of PUFA drops the body temperature of mice, that the body temperature and metabolic rate of a starch eating South American tribe dropped given access to vegetable oil and that the saturation level of Americans stored bodyfat dropped dramatically between 1961 and 1992. Previously I showed that the body fat of obese adults is highly unsaturated. Furthermore, the last two links demonstrate a parallel rise in both stored PUFA and monounsaturated fats (MUFA). In a non torporous state, PUFA should suppress the enzyme SCD1 and therefore lower MUFA. PUFA and MUFA rising together is a sign of torpor.
If all of this is true then we should expect measured Resting Metabolic Rates (RMR) of women to have decreases as the 20th century progressed.
- Women’s Resting Metabolic Rates (RMR) dropped between 1919 and 1986
- In a woman below 30 weighing 130 lbs, the drop was about 180 calories per day while chilling in a chaise lounge
- That’s 20 lbs of butter over one year
- Metabolic rates were higher in 1919 for all ages and BMIs of women
- Trained athletes in 1986 had lower RMRs than non-athletes on a college campus in 1919
- Plotting metabolic rate per kilogram vs. BMI really showcases the differences between 1919 and 1986
- The difference in RMR is highly statistically significant
Luckily, the tools to measure RMR are relatively simple and we could do it over a century ago. Harris and Benedict published their landmark book on the subject1 in 1919, including the individual heights, weights and metabolic rates of over 103 women performed in the lab at Wellesley college, an elite New England women’s college. The women ranged from 15 to 73 years of age with BMIs from the teens up into the mid 30s. Harris and Benedict made an equation to predict someone’s RMR based on their age, height, weight and gender known as the Harris-Benedict equation, which is still in use today.
In 1986, Owen published A reappraisal of caloric requirements in healthy women2, concluding that “the currently available tables and regression equations (including the Harris-Benedict equation) overestimate the RMR of healthy women by 7-14%”. An equation developed to predict metabolic rate in 1919 overestimated the metabolic rates of women in 1986. A great thing about this paper is that 1) it gives us the heights, weights and metabolic rates of 44 individuals, not just the mean number like most papers and 2) they include 8 trained athletes, so we can’t assume that the lower metabolic rates seen in the paper are due to sloth.
I plotted out all 147 women from both papers. I’ll post the spreadsheet and my Jupyter Notebook at the end if people want to look at the data.
Here is the metabolic rate of all women below age 30, 1919 vs. 1986, plotted against their weight.
A 60 kilogram woman (132 lbs) in 1919 burned on average an additional 180 calories per day than a woman of the same weight in 1986 while lying in a chaise lounge. To put that into perspective, it’s the difference of 20 pounds of butter per year.
Even trained athletes in 1986 had lower RMRs than average University students did in 1919. The differences persisted in older women:
I also plotted RMR per kilogram of body weight, thinking that this looking at BMI would control somewhat for body type given the limited data. This is in women below 30 again. If anything, this accentuates the differences between the groups. When you look at BMI, there are almost no women from 1919 below the trendline from 1986 nor women from 1986 above the trendline from 1919. No matter how you look at it, metabolic rates dropped significantly between 1919 and 1986:
The trends move lower as BMI increases because this stat is per kg and fat tissue is metabolically less active than muscle.
I even did a statistical analysis to show that the measured RMRs from 1919 are significantly higher than those from 1986.
|Mean RMR/Kg 1919||Mean RMR/Kg 1986||P-Value|
|Methods: Cut and paste from Stack Overflow. I got this solution from S Anand. It’s in the Jupyter Notebook posted below.|
This was my first project using a Jupyter Notebook. What amazing tech! Huge shout out to anyone involved in developing Python, pandas, NumPy, SciPy, etc! And of course those involved with the Jupyter Notebook project! And also to all of you who post such great solutions on Stack Overflow. Sometimes I say I’m a coder, which is to say I’m clever enough to cut and paste solutions from Stack Overflow and change the variable names.
Also check out the Fire In a Bottle Jupyter Notebook Github repo although there’s not much there yet.
- 1.Harris J, Benedict F. A Biometric Study of Basal Metabolism in Man. Carnegie Institution of Washington; 1919.
- 2.Owen OE, Kavle E, Owen RS, et al. A reappraisal of caloric requirements in healthy women. The American Journal of Clinical Nutrition. Published online July 1, 1986:1-19. doi:10.1093/ajcn/44.1.1