Obesity of Populations: Concepts and Terminology

This may be pretty boring for the more scientifically minded/literate, but I’ve found myself arguing and clarifying these concepts before. So I’m going to write it here now, just so I have something to point to when there is a misunderstanding (ie somebody being wrong)


1 Calorie = 1 kilocalorie = 1000 calories. A calorie is an SI unit, the energy required to heat 1 gram of water 1 degree celsius. Food is typically labelled in Calories (capital C), which are equal to 1000 calories. 1 kcal = 4.19 kilojoules (kJ) (scientific literature often uses joules as a unit). I’m going to say kcal to be precisely correct, just understand that these are the same as what food is labelled.

Basal Metabolic Rate (BMR): The amount of energy a person burns while completely at rest. Usually measured in kcal/day. This is most definitely not the number of kcal a person eats or expends in a day, it’s the number a person would use if they stayed in bed not moving. Also called “Resting Metabolic Rate (RMR)”. This number varies among individuals, but measuring it can be done in a lab. Just stick people in a box and measure their output while they’re strapped to a bed.

Total Daily Energy Expenditure (TDEE): The number of calories a person burns in a day, including all activity. Hard to measure, because quantifying normal daily activity is difficult.

Body Mass Index (BMI): Weight divided by height squared. Always presented in kg/m^2. BMI is often used in studies on populations as a measure of obesity, because it’s extremely easy to measure. Some caveats are necessary. BMI correlates strongly with body fat percentage, but the correlation does vary between ethnic groups [1]. Also, while BMI can be used to accurately predict average BF% in populations, it should not be used as anything other than a loose guideline for individuals.

On that note: You lose weight, and your pants become loose. Yes, english sucks, but those are the correct spellings.

Positive / Negative Feedback: People often use “positive feedback” as synonymous with “good” feedback. You see this a lot in economics, in fact. “Negative feedback” being “bad” feedback, something like people being out of work -> people spending less -> more people being out of work, being described as “negative” feedback. In the engineering world, “positive” feedback refers to feedback in the direction of a disturbance, “negative” being the opposite. Imagine having a full glass of water on a table; give it a light tap, and it’ll rattle back and forth. That’s negative feedback. Gravity opposes the change. If you tap the glass hard enough, the center of gravity moves outside the glass, positive feedback begins, and the glass tips over.

Morbidity and Mortality: Mortality is dying, morbidity is a serious and/or permanent disability. These two are not always the same. Overweight people have a higher risk of morbidity (diabetes, heart attacks) but in elderly individuals actually have a lower risk of mortality[2]



Studies typically look on populations of people, and that’s what I’m going to be looking at. Obese populations have a higher incidence of diabetes (among other things) than normal weight populations. That doesn’t mean there aren’t thin diabetics, or fat people without diabetes. But the rates are very much different. Studies often give “Relative Risk” (RR) or “Odds Ratio” (OR). For example, the general population of women have a 12% risk of developing breast cancer, women with a BRCA1 or BRCA2 mutation can have up to an 80% chance[3], which would be an odds ratio of about 6.7. I’m going to present a lot of studies showing increase risk of X or Y, or that some property (sleep for example) is correlated with body weight. That doesn’t mean that if you personally sleep an extra half an hour you will definitely lose a pound (numbers made up). It means that populations which slept a half longer on average weighed a pound less, on average. Personal mileage may vary with these types of effects, but when considering populations they should be effective (assuming you believe the studies in question).

Correlation does not necessarily equal causation

This is a general caveat with social science research. Heart attacks are (probably) correlated with wearing dentures, that doesn’t mean dentures cause heart attacks (or heart attacks destroy teeth). It just means old people have heart attacks and wear dentures.

Genetic is not synonymous with either “didn’t choose” or “can’t change”

“Genetic” here is referring to the base-pair sequence in a persons DNA. Epigenetics is the study of gene expression, and while epigenetic changes are not transmitted through the germ line (ie from parent to child) they are transmitted through cell division. So changes early in life (even in the womb) can have long lasting effects.

Secondly, the way parents raise their kids effects the choices they make later in life. This should not be a controversial idea in general, although the specifics will of course vary. I’m sure there are lots of people which turned out great in spite of bad parenting, and vice versa. But the odds are different.

For example, being abused early in life increases the odds of becoming an abuser [4]. I’m not saying this is the same level of trauma as a parent teaching their kid bad eating habits. Just providing an example

Events have multiple causes

Say an old friend comes in from out of town and you stay up late catching up. You hit the snooze button a few times the next morning, and are still groggy on the drive to work. You stop in at Dunkin Donuts to pick up some coffee to perk up. While turning back on to the road, Jack runs through a red light and hits your car. Both of you sustain injuries which are fairly minor, no permanent damage. Whose fault is the accident?

Most people would say Jack, and he would probably be held legally culpable. That’s fine, from a legal standpoint I don’t disagree. But there are a large number of factors here, and changing any one would have prevented this accident (and perhaps caused a different one). Your friend didn’t have to come in to town, you didn’t need to stay up late, you didn’t need to stop for coffee, if you hadn’t been so tired you may have seen him about to run the light and avoided the accident, if you had hit the snooze button one more or fewer time you wouldn’t have been at that intersection at the same time.

Those little things are what I’m going to be talking about. I’m not trying to lay any legal responsibility on Dunkin Donuts for making the US fat, but objectively speaking, would obesity decrease if there were no fast food restaurants in the country? By how much? These are the types of questions I’m going to look at.

Events have multiple effects

Something can have multiple effects, both healthy an unhealthy. Body weight is a convenient example. A persons risk for heart disease decreases with BMI at all ranges [5]. But heart disease isn’t the only thing that can kill you, mortality risk is higher for the underweight than normal weight (or overweight)[].


  1. [1]Healthy percentage body fat ranges: an approach for developing guidelines based on body mass index

    Am J Clin Nutr  September 2000 vol. 72 no. 3 694-701


  2. [2]Landi F, Onder G, Gambassi G, et al. Body mass index and mortality among hospitalized patients. Arch Intern Med. 2000;160(17):2641-2644


  3. [3]http://www.breastcancer.org/risk/factors/genetics.jsp
  4. [4] The British Journal of Psychiatry (2001)179: 482-494doi:10.1192/bjp.179.6.482


  5. [5]

    Nordestgaard BG, Palmer TM, Benn M, Zacho J, Tybjærg-Hansen A, et al. (2012) The Effect of Elevated Body Mass Index on Ischemic Heart Disease Risk: Causal Estimates from a Mendelian Randomisation Approach. PLoS Med 9(5): e1001212. doi:10.1371/journal.pmed.1001212


This entry was posted in Public Health, Uncategorized and tagged , . Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *