The Truth in Statistics

I am always thrilled to have patients come to my office with studies or books that they have read. Education is the power to individual health and wellness. The internet has opened up the doorway of medical research and information to all. This is a very good thing. But it also pulls the curtains back on the short-comings of much of the research out there today.

Research is not a perfect science. We all bring prejudices, bias’, and incomplete knowledge to the table. As hard as we can try to keep scientific research pure, we cannot. In many ways, statistics is a game to play to prove your point. So, every study should be viewed through the prism of healthy skepticism.

No better example of this was an article that a client of mine brought to my attention recently.

The Study, “Perimenopausal Depression”, was published in the American Journal Of Psychiatry in January 2008. As Journals go, this is a prestigious journal in the field of Psychiatry. The premise of the study was accurate: Perimenopause is associated with an increased incidence of mood disorders, particularly depression. However, there were several glaring errors in the study.

First, there were several incorrect statements made. The author assumes that an elevated FSH is related to a decrease in Estradiol levels in the perimenopausal period. However, this is an inaccurate assumption. While it is true that an elevated FSH is associated with decreased ovarian function; the elevated FSH is actually due to decreased levels of another hormone–inhibin. The elevated FSH and estrogen levels have no direct link. This is well published throughout the medical literature. In fact, one of the earliest hormonal changes in perimenopausal women is a decrease in progesterone levels due to decreasing effective ovulation.

Second, the author makes the statement that progesterone increased the negative effects of depression, while estrogen alone improved depression. However, the study the author cites for support, used medroxyprogesterone acetate. This drug, a synthetic analogue, is commonly known as provera and is well known to increase depression. However, provera is not Progesterone, it is a synthetic progestin. Actually, progesterone has been shown to decrease depression by 60% when compared to the synthetic progestin, medroxyprogesterone acetate (provera). The author of this study is comparing apples to cumquats.

I want to expand on the second point above. The mistake this author makes is epidemic throughout the medical literature today. The author intermixes progesterone and medroxyprogesterone acetate as if they are the same hormone. They are in fact very different. Progesterone is what is called a “bioidentical” hormone. It is the hormone that a women produces in the later 2 weeks of her menstrual cycle. Medroxyprogesterone acetate, also known as provera, is a synthetic (not made by the body) progestin which was designed to protect the lining of the uterus from excessive Premarin stimulation. This followed the increased incidence of uterine cancer following the introduction of premarin back in the 1950s. Provera does protect the lining of the uterus, just like progesterone, but that is where the similarities end. Just compare the chemical structures below and you can tell they are not the same compound.

A good friend of mine, Eldred Taylor, MD, says he can make you an honorary gynecologist if you can tell the 2 compounds above apart. As you can see from this article in question, this is a fact that many a physician today could not correctly identify.

Last, is the many unsubstantiated statements made by the author in the article. The author references, “other studies suggest…”, “overall, in the studies…”, and “in the studies that used progesterone replacement in addition to estrogen, the progesterone tended to decrease the beneficial effects of estrogen, and in some patients it exacerbated depressive symptoms…”. The problem is the “studies” are not referenced. An unreferenced statement is an opinion. This is why I heavily reference my posts. Opinion is nice, but when it comes to scientific research, opinion represents a potential bias and compromises data.

The goal of this blog is to not call out the author of the above mentioned article, but to have everyone learn to view all information and research articles with a healthy bit of skepticism.

To your health and wellness and to your hormone balance.

Related Posts Plugin for WordPress, Blogger...

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!

SSCP   CAS-002   9L0-066   350-050   642-999   220-801   74-678   642-732   400-051   ICGB   c2010-652   70-413   101-400   220-902   350-080   210-260   70-246   1Z0-144   3002   AWS-SYSOPS   70-347   PEGACPBA71V1   220-901   70-534   LX0-104   070-461   HP0-S42   1Z0-061   000-105   70-486   70-177   N10-006   500-260   640-692   70-980   CISM   VCP550   70-532   200-101   000-080   PR000041   2V0-621   70-411   352-001   70-480   70-461   ICBB   000-089   70-410   350-029   1Z0-060   2V0-620   210-065   70-463   70-483   CRISC   MB6-703   1z0-808   220-802   ITILFND   1Z0-804   LX0-103   MB2-704   210-060   101   200-310   640-911   200-120   EX300   300-209   1Z0-803   350-001   400-201   9L0-012   70-488   JN0-102   640-916   70-270   100-101   MB5-705   JK0-022   350-060   300-320   1z0-434   350-018   400-101   350-030   000-106   ADM-201   300-135   300-208   EX200   PMP   NSE4   1Z0-051   c2010-657   C_TFIN52_66   300-115   70-417   9A0-385   70-243   300-075   70-487   NS0-157   MB2-707   70-533   CAP   OG0-093   M70-101   300-070   102-400   JN0-360   SY0-401   000-017   300-206   CCA-500   70-412   2V0-621D   70-178   810-403   70-462   OG0-091   1V0-601   200-355   000-104   700-501   70-346   CISSP   300-101   1Y0-201   200-125  , 200-125  , 100-105  , 100-105  , 300-209   1Z0-803   AWS-SYSOPS  , 100-105  , JN0-360   CAS-002   350-029   70-243   70-462   70-270   350-029   350-050   LX0-104   350-050   000-017   70-270   c2010-657   1Z0-051  , 74-678   70-533   500-260   300-209  , 350-080   MB2-704   VCP550   PR000041  , 210-065   70-347   70-486   70-980   1z0-808   350-018   000-080   70-417   70-413   70-178   350-029   350-030  , 2V0-621D   70-487   70-461  , N10-006  , 300-115   220-801   CISSP   642-732  , MB2-704   N10-006   200-101   300-070