Metabolic Rates in the American South in the 1930s: Did Researchers Measure the Origins of the Obesity Epidemic?

There are great tales in the medical literature if you look for them.  In the 1930’s – the early days of metabolic research – a sort of North-South challenge presented itself.  An SEC-Big 10 challenge of metabolic rates (For those who are not American college football fans, the SEC is a Southern Conference and the Big 10 a Northern one and they play each other each year) amongst young college women. 

The issue was that basal metabolic rates (BMR) in college women in the South were consistently 10-15% lower than of college women in the North.  

Several equations for predicting basal metabolic rate based on height, weight, gender and age were published before 1920, most prominently the Harris-Benedict equation^​1​ – developed at Wellesley college in Massachusetts – and the DuBois equation^​2​ – developed at Weill Medical college in New York City.  In 1922, Walter Boothby and Irene Sandiford published, “Summary of the basal metabolism data on 8,614 subjects with especial reference to the normal standards for the estimation of the basal metabolic rate.”^​3​  This was a large study done in Minnesota showing that indeed, the DuBois and Harris-Benedict equations correctly predicted the basal metabolic rates of normal, healthy adults to within plus or minus 10 percent.

I particularly enjoy the handmade graphs from this paper.  The middle line represents people whose BMR are perfectly predicted by the specified equation.  The top dark line is people whose BMR is 10% higher than the equation predicts and the lower dark line is people whose BMR is 10% lower than the equation predicts.  As you can see, 90% of all people fall within the dark lines and those who fall outside of it don’t do so by much.

Look at the PENMANSHIP!!!!

A Discrepency In The South

Then in 1930, Jennie Tilt wrote this in “The basal metabolism of young college women in Florida.”^​4​

This is a deviation from the prediction of -9.9 per cent (HarrisBenedict) and – 10.6 per cent (Aub-Du Bois). If it is accepted that the standards for women should be lowered by about 5 per cent, the average metabolism of this group still is lower than the normal limits. By the Harris and Benedict prediction there are only two whose metabolism is above the predicted and three by the Aub and Du Bois prediction. All the others are below. It is interesting to note that by the Harris and Benedict prediction thirty, or 58 per cent, of the group are below -10 per cent; twenty-eight, or 54 per cent, are below by the Aub and Du Bois prediction. All the very low results were repeated several times so that we feel quite confident that the figures represent basal values. It would thus seem that there is some evidence that the basal metabolism of southern college women tends to be lower than that predicted for young women of the same age and living in the north.

More than half of the Florida women had metabolisms below the dark lower line – 10% below the prediction of the Harris-Benedict or DoBois equations.  This is opposed to less than 5% of the women in Minnesota.

In 1931 Callie Mae Coons, studying young woman in Oklahoma, reported:^​5​

Deviation from the Du Bois (1927) standard varies from – 15.5 for the youngest group to -10.6 per cent for the 25-29 age group. Although some groups are small, it will be noted that four out of the seven groups average a deviation very close to – 14 per cent, – 13.2 per cent being the average for all ages. 86 per cent of the 18 year old group and 71 per cent of the entire number are below the – 10 per cent limits of the Du Bois predictions, and 37 per cent below the – 15 per cent limits. These facts emphasize the inapplicability of present standards to Southern women. 

So 37 percent of the women in Oklahoma had a metabolic rate more than 15% below the standard as opposed to none of the women in Minnesota.

South Carolina, 1931:^​6​

Determinations of basal metabolism on ninety-three student nurses and forty male medical students at Charleston, S. C, showed results averaging about 10 per cent lower than the Aub-Du Bois standard. Climate, the relative amount of iodine in human environment, dietary habits, and the state of nutrition are not considered responsible for the low values obtained.

Tilt published another study of Florida women in 1935:^​7​

1. Basal metabolic rates were determined on a group of thirty normal young women who have lived in Florida 8 years or longer. The average deviations calculated from pre diction standards were: Aub-DuBois, modified by Boothby and Sandiford, -14.1 per cent; Harris-Benedict, -13.0 per cent. 

2. Basal metabolic rates were determined on a group of 8 young women who had recently come from northern states. The average deviations were – 7.7 per cent from the modified Aub-DuBois standard, – 7.8 per cent from the Harris-Benedict (extrapolated). 

3. A comparison of the average basal metabolism of young women living in Florida with those from northern states shows a decidedly lower rate for Florida women. 

4. A 4-day dietary study on eighteen young women of the group living in Florida showed an average intake of 1.23 gm. protein and 39.9 calories per kilogram of body weight per day. The protein was of good quality. 

5. No consistent relationship was noted between basal metabolism and protein or calorie intake.

New Orleans, 1939:

 A comparison of observations on 160 men and women with the usual standards suggests that a statistically valid standard for New Orleans, and possibly other southern regions, would be obtained by deducting 10 per cent. from the Aub-Du Bois values. 

Then in 1939, a paper from the University of Connecticut^​8​ (a Northern state) was published, reaffirming that yes, indeed women in the South had lower metabolic rates:

The average basal rate of the Connecticut State College women was found to be -2.3 Aub-DuBois; -1.9 HarrisBenedict.  The excellent quality and quantity of food may be a factor in producing a relatively high average.

So metabolic rates in New Orleans, South Carolina, Florida and Oklahoma ranged from 10-14% below the DuBois and Harris-Benedict standards, whereas women from Connecticut were only 2% below the standard.  

Factors In Metabolism

The main factors suggested for this difference in metabolic rates at the time were food quality – especially protein, iodine content of the food – affecting the thyroid, and climate.

Th 1935 Florida paper brings into doubt that protein quantity or quality or caloric consumption are affecting metabolic rate.  

From a study of table 4 which summarizes the average daily intake of each individual, it is noted that most of the diets were quite adequate from the standpoint of grams protein and total calories. …  The quality of the protein consumed appears to be quite adequate for meat, milk, and eggs were well distributed throughout the diets. …  There does not appear to be any definite parallelism between protein and calorie intake and basal metabolism. 

The South Carolina paper points out that iodine levels are higher in the South and especially South Carolina than the rest of the country.

The widely quoted studies of Olesen on men drafted for military service in the United States showed that enlarged thyroids are least common in the southeastern part of the country.  Systematic examinations of school children have been made in many places where goiter is known to be endemic, and there are schools in some parts of the country where practically every child has an enlarged thyroid. Examinations usually have not been made in areas where goiter is not a problem, although Hayne 9 reported the results of the examination of 17,600 school children by county health officers in twenty counties of South Carolina.  … practically all standards of metabolic rate have been developed in regions where goiter is more common than it is in South Carolina … From this point of view the results are disappointing, as they show values for both men and women around 10 per cent below those of the Russell Sage Institute or the Mayo Foundation.

Climate

Climate is the thorniest question to disentangle.  The argument is that people in a warmer climate need to do less thermogenesis to stay warm and therefore lower their metabolic rate.  The first obvious problem in this data set is that the metabolic rates in Oklahoma were equally low as those in Florida even though Oklahoma is much colder than Florida. 

Another problem with this idea is that Benedict himself had travelled to the Yucatan Penninsula in 1928^​9​ and found that Mayans had a BMR 5.2% above the rate predicted by his own equation.  A 1931 follow up trip^​10​ found that Mayans had a rate 5.8% higher than his equation.  So people in a tropical climate don’t necessarily have a lower metabolic rate.

A 1941 paper^​11​ found that Europeans (the college students being studies in all of these experiments in the segregated South of the 1930’s are presumably of European descent) living in a tropical climate had a decline in metabolic rate of 5.7% compared to the Harris Benedict standard and 6.3% compared to the DuBois standard.  So there was a decline but nothing like the 13-14% declines seen in Oklahoma and Florida.

The most compelling argument from today’s perspective that the difference in metabolic rates was not due to climatic differences is that today everyone in the US has the metabolic rates of 1931 Oklahomans or 1935 Floridians, regardless of their latitude.  For instance, this 1985 paper^​12​ found that women at the Memorial Sloan-Kettering facility (New York City) had metabolic rates 13.5% below that predicted by the Harris Benedict equation whereas women at the Emory (Georgia) facility were 10.9% below prediction – lower metabolic rates in the North!

Unsaturated Fat

I’ve argued on this blog that unsaturated fat lowers metabolic rate.  My argument is summarized in ROS as a NAD+ Regenerating, Thermogenic Cycle and Obesity as a Global Succinate Dehydrogenase Activity Deficit.  For this to possibly explain the discrepancy between Northern and Southern metabolic rates we’d need to know that there were large regional differences in types of fat consumed in the 1930s.

The USDA 1939 Yearbook of Agriculture presents this table in the article “Present-Day Diets In The US”.  “Other fats, oils and fatty foods” are defined as “lard and lard compounds, vegetable oils, margarines, bacon, and salt pork”.  

Helen Worth’s 1943  “Down-On-The-Farm cookbook”  is matter of fact about the ability to swap out butter for margarine, other hydrogenated vegetable oils or corn-finished lard and mentions liquid vegetable oils.  All of these substitutions would have increased polyunsaturated fat content.  All of these were in wide use and part of the “other fats”.

Combining the data from the USDA in 1939 and data from FAOSTAT, we get a table like this:

Summary 

In the 1930s Southern Women were eating more unsaturated fats than Northern Women and had lower metabolic rates.  By 1985, American women everywhere were eating lots of unsaturated fat and had uniformly low metabolic rates.  At the very least, a theory of obesity should have some predictive value in the real world.  The idea that metabolic rate is partially controlled by the degree of fat saturation has predictive value in this instance.

Metabolic rates have dropped since the 1930s in parallel with penmanship and the rise in vegetable oil consumption.