Obesity and Altitude

Today’s guest post comes from Dr Jameson Voss.  You can find more on Jameson at the bottom of this post.  You can find out how to submit your own Obesity Panacea guest post here.

Obesity is a complex and multifactorial chronic disease that remains a military and public health priority in the United States. Recently, we’ve identified a strong association between obesity prevalence and altitude within the US. Our findings were surprising because they indicated the magnitude of this association was large and the pattern of association exhibited a curvilinear dose response in 500 meter categories of altitude. There was a 4-5 fold increase in obesity prevalence at low altitude as compared with the highest altitude category after controlling for diet, activity level, smoking, demographics, temperature, and urbanization. We published our findings in the International Journal of Obesity (advance online publication doi:10.1038/ijo.2013.5) and presented at the 2013 American College of Preventive Medicine conference.

The process we used is easily reproducible. We combined several publicly available national datasets using statistical software and geographic information systems using the county of residence as a common linkage across datasets. For a basic visualization, Figure 1 shows the Centers for Disease Control and Prevention’s publicly available map with projected obesity prevalence for each county adjusted only for age. This map was created based on similar data as the source for our study, but we used actual self-reported height and weight rather than the modeling shown in Figure 1 and we adjusted for age, sex, race/ethnicity, physical activity compliance, fruit and vegetable consumption, smoking status, employment status, education, urbanization, temperature category and income. By comparison, Figure 2 from http://ned.usgs.gov/ shows the topography of the United States.

Figure 1. Age Adjusted Obesity Prevalence by County.  This image was obtained from cdc.gov/diabetes, but this particular map represents obesity prevalence and not diabetes.

Figure 1. Age Adjusted Obesity Prevalence by County. This image was obtained from cdc.gov/diabetes, but this particular map represents obesity prevalence and not diabetes.

Figure 2. A topographical map of the USA.

Figure 2. A topographical map of the USA (source). Note the similarities with Figure 1.

While it is always important to remember correlation does not prove causation, in this case, we already know hypoxia causes anorexia and weight loss based on well controlled interventional data.  This effect is biologically plausible based on the relationship between hypoxia and leptin signaling, norepinephrine and sympathetic tone, non-erythroid erythropoietin receptor signaling, and the metabolic demands at high altitude.  We hope additional research will help clarify the mechanisms and long term health effect of either high altitude residence or normobaric hypoxia.  These results, showing a large magnitude of association, provide some optimism that this is a worthy line of research.

Jameson Voss

Jameson Voss

About the Author: Jameson Voss is a third year Preventive Medicine Resident at the Uniformed Services University of the Health Sciences.  His research focus is on obesity. 




ResearchBlogging.orgVoss, J., Masuoka, P., Webber, B., Scher, A., & Atkinson, R. (2013). Association of elevation, urbanization and ambient temperature with obesity prevalence in the United States International Journal of Obesity DOI: 10.1038/ijo.2013.5

Related Posts Plugin for WordPress, Blogger...
This entry was posted in Guest Post, Obesity Research and tagged . Bookmark the permalink.

33 Responses to Obesity and Altitude

  1. Is religion lumped under “demographics” (which it shouldn’t be)? I could argue that’s a Mormonism-obesity correlation. I could also argue that because of only partial parallels in shading of the Rockies vs. the Sierras, and little parallels in shading between the Rockies and the Cascades, and no difference in shading between Appalachia and surrounding areas, you’ve ….

    Discovered nothing.

    Add in that relatively high altitude areas of the Colorado Plateau show huge obesity problems even after factoring out whatever demographics you did, and I’m also not impressed.

  2. You’d also have to check, to further try to establish causal correlation, obesity levels in the Andes vs. rest of South America, Himalayas and nearby vs. much of the rest of Asia, and the Alps, etc. vs. rest of Europe. Call me back later.

    • Travis Saunders, MSc, CEP says:

      Religion wasn’t included in demographics, nor is it typically included. I don’t know that Mormons alone can explain the whole trend, especially given the relationship in the Appalachians. Not that that would mean that there is no correlation – it would just explain the mechanism.

      Keep in mind that correlations are never perfect – there will always be individuals/areas that don’t fit with the trend. That doesn’t mean that the correlation does not exist, it just means there’s some nuance. That’s true of all correlations. If you take that as evidence that the study “discovered nothing” then we might as well just get rid of the whole field of epidemiology.

      • I agree that correlations are never perfect. Is the altitude-obesity fit a better one than Mormonism-obesity? I’d say it’s a wash, and that’s why studies of elevated areas in foreign countries are needed before making further judgments.

        • Also, “discovering nothing” was a bit rhetorical. But … per your first graf in response, if the Mormon angle is real, no, it wouldn’t mean this: ” I don’t know that Mormons alone can explain the whole trend, especially given the relationship in the Appalachians. Not that that would mean that there is no correlation – it would just explain the mechanism.”

          Rather, it would mean that there’s a different effect at work, that’s not altitude-affected at all. Altitude would then not apparently be causal, but Mormon dietary strictures would be, instead.

          • Travis Saunders, MSc, CEP says:

            Let me clarify what I meant above – in this sample, altitude and obesity are correlated. It may or may not be causal, but there is a correlation nonetheless. We may find out that the correlation can be explained by some other variable, but that won’t mean that the correlation does not exist. It would just mean that altitude would not be causing the observed fluctuations in obesity rates. The correlation would remain.

  3. C. S. Edwards says:

    That looks like an artifact, possibly of reporting. The granular proof might be a comparison between Kit Carson County, Colorado and Sherman County, Kansas. They share a border and elevation, are both on I-70, have similar economic, employment and education patterns and are similar in size and population. The only difference between the two counties is the state government.

    Another possible line of inquiry would be air quality. The urban and suburban population centers of Utah are in a valley which has a tendency towards inversion. The SLC metro area has a history of high particulate and ozone readings. Denver metro does not have as significant a problem with winter inversion. Both metro areas have similar elevations and populations.

  4. An interesting post. As per the others who have commented on the possible correlates, I just wonder if I could throw in another ‘overlap’ as per the recent paper by Arns and colleagues on solar intensity (SI) and ADHD prevalence (no really…)


    One of their maps of the US with that SI factor in mind (figure 1) looks ‘similar’ (figure 2):


    Obviously I might just as well be talking chocolate consumption and Nobel prize winners…..

  5. I’m imagining the conversation at the doctor’s office . . .

    “I see you’ve gained ten pounds, Mrs. Fitz. Unfortunately, in our corner of the country, oxygen levels are quite high. Have you considered not breathing so much?”

    But I’ll take my excuses where I can get them. :-).

  6. There is a VERY SIMPLE explanation: RESTING METABOLIC RATE is higher at higher altitude and… lower in tropical regions!! This explains this correlation, but also the fact that in tropica regions Obesity rates (with more food available) are increasing faster than in cold regions! http://www.vitasanas.ch/wp/wp-content/uploads/2012/09/BOOK5_Resting_Metabolic_Rate_RMR.pdf

    • Antranik says:

      But then they would eat more to compensate, naturally. Happens every year. Thyroid production goes up during the winter time and we stuff our faces. Goes down in the warmer months and we eat light salads.

  7. JRM says:

    I winder what the correlation would look like if you split the country at about 102° longitude and looked at the eat and west halves separately.

    The solar intensity angle is very intriguing to me.

  8. JRM says:

    Winder should be wonder (obviously)

  9. ben says:

    It’d surely be fun to find all the confounding factors in this one…

  10. I definitely feel this. I used to live in Colorado and Wyoming (college & home)… And used to frequent Santa Fe quite often as well… I would never go to sea level. I felt great and “tight”! Unfortunately (although I do love it here), I moved to Los Angeles two years ago, and I have noticed my body has changed a little bit. Granted if I put on weight it might be a pound or two (I work out religiously and eat a very very healthy diet), but I am a short girl so I can feel it… But no matter what, I just have to accept it. The second I do go home to visit though, I feel lighter on my feet! It’s weird how that works, but it does!

  11. Pingback: 04.22.13 | CrossFit 312

  12. Pingback: Nutrition & Performance at Altitude | Boot Camp & Military Fitness Database

  13. I have heard a NIH researcher make these same claims. This makes me curious – an association does NOT prove causation. Perhaps Coloradans are more active, more outside. Perhaps Colorado draws a healthier population. Perhaps it is ethnic, income, etc. related. Again this is an association, NOT causation.

    If high altitude really was protective against obesity, I’d like to look for an association between the number of frequent flier miles and BMI. I’m pretty sure that those at the highest quartile for frequent flier miles will not be at the lowest quartile for BMI :-) Shouldn’t frequent fliers be toothpicks with this same logic?

    Again, remember an association – Estrogen replacement therapy lowers LDL cholesterol. Therefore it was assumed for decades that this would help prevent heart disease in women. Not true – in fact, HRT increases risk rather than decreasing it. Careful how you interpret this stuff :-)

  14. @Ethan Lazarus:
    You have to distinguish between statistics and physiology!!!
    An increased RESTING METABOLIC RATE, as it is the case at higher altitudes, simply means that the body burns MORE energy (calories) even at REST!!! And since the calories the body burns during any physical activity are RMR-dependent… you can conclude…

  15. hanjwils says:

    Culture also plays a huge role in obesity rates. This has been touched on by suggestions of religion but it also involves diet, transportation and leisure time activities (among so much else). It is very easy to find active things to do in the mountains compared to the plains. Biology probably does play a role but discounting the role of culture in any complex complex human phenotype ignores a large portion of the influences on humans.

  16. Pingback: Does Altitude Fight Obesity? | Report in for duty.

  17. DA says:

    You mentioned that the correlation exists even when controlling for diet. Do you mean diet, or self-reported diet? Those are different things.

    • Travis Saunders, MSc, CEP says:

      Unfortunately you can pretty much always assume that when a study adjusts for diet, it’s for self-reported diet. There’s just no simple way of getting actual food intake outside of a lab-based study.

  18. Ruth Sponsler says:

    Is there any way you could provide a link to a version of the article that is not behind a paywall?

    • Travis says:

      Unfortunately I can’t (not legally anyway), but as the author of the paper, Jameson probably can. His email address is listed on the abstract of the paper, I’d suggest emailing him to see if he is able to send a copy.

  19. Pingback: Obesity and Altitude = Good Science | Dr. Nick Campos

  20. Pingback: The AMA Now Officially Recognizes Obesity as a “Disease” – So is This a Good Thing or a Bad Thing?

  21. Jameson Voss says:

    Thoughtful coverage of this post by Alex Hutchinson from Runner’s World. http://www.runnersworld.com/weight-loss/does-altitude-fight-obesity

  22. Trish says:

    I am obese and live in Georgia. When I visit my son in Colorado (altitude is over 9000 feet), I lose 20 pounds in three weeks. If I stay 1-2 weeks, nothing happens. A similar thing happened when I visited New Mexico, living 5 months at 7000+ feet. I lost 20 pounds within the first month or so and my appetite went down. This is antecdotal, I know, but I have also read research that indicated altitude-based weight loss through direct measurement on experimental subjects. Note that when I return to low altitude, I don’t gain the weight back for at least a year. Eventually I gained it back though.

  23. Jameson Voss says:

    Very interesting, Trish! Thank you for sharing your story.

  24. Pingback: Does Central Heating Contribute to Obesity? | Paleo Diabetic

  25. Pingback: Is Obesity Caused By Lack of Altitude? | Advanced Mediterranean Life