Welcome to The Weekly Transfusion, 1.4 for the week of April 6, 2009.

Editorial: Medical journal statistics for autodidacts

You can become a better consumer of health information if you take the time to read the research source material (i.e the scientific publication in which the original claim was made). Of course if the study is technical you can see quite a bit of jargon that you may or may not understand. However many medical terms are widely understood and where you bump up against the odd phrase or name that you don't comprehend, there are usually places on the Internet where you can find simple, easy to understand explanations. Wikipedia is actually pretty good for this type of look-up, as long as the subject at hand is not controversial.

However, methodology and monikers aside, most scientific studies distill down to a simple testable premise which is easily understand by almost anyone. Did the medicine work? Was the association between this gene and that disease valid? Past asking the question, what is needed next is to look at and gauge the value of the answer. Surprisingly, even though this is usually some sort of statistical type of answer (and most laypeople are not well versed in statistics) once you know what to look for, you'd be amazed just how easy it is to evaluate most studies.

Most research studies feature a subsection entitled Results or Conclusions. It is here that the results are most often given. There are many way of calculating statistical significance, but the premise is quite simple: What is the chance that the thing we just observed/ hypothesized was random versus the odds of it being due to the relationship we are studying. This is known as probability and in statistics is usually called the P value. To find out just how significant the results of any study are, just look for the P value. The smaller this number is, the less likely the results occurred by chance. Put another way, the lower the P value the more likely you'll want to view the results as significant or important.

The great Ronald Fisher viewed P values as measures of the evidence against a hypotheses, sort of like how a prosecutor presents a case based on exceeding the jury's sense of 'reasonable doubt.'

Now for the secret (OK, not so secret) key to taking control of the medical facts in your life: The standard level of significance used to justify a claim of a statistically significant effect is when P is equal to or less than 0.05; in essence, a one-in-twenty chance that the result had nothing to do with your hypothesis.

For better or worse, the term 'statistically significant' has become synonymous with P<=0.05.

So when looking at any published results, always look for the P value and if it is greater than five cents on the dollar (0.05) you'd probably want to ignore that results (unless the premise of the article was that the researchers failed to show a relationship, which is of course just another type of observation; however, these types of studies usually don't make it out of the researcher's file cabinet) or take a look at the methodology behind the study (scientists are human; studies can be poorly designed and the conclusions derived may not have been the best test of the hypothesis).

So, P<0.05 means the results are significant, but just barely. Good enough to convict, but also likely to send a few innocent people to jail as well, since there are still strong indications that the hypothesis fails to account for the whole of the facts. Personally I like to see P values of at most 0.01-0.02 before I get excited about anything I'm reading. However I do make exceptions for studies with small numbers of participants, or if the we're dealing with an herb or vitamin where the effects studies may be slight or slow to surface.

Oftentimes you'll see P values with lots of zeros. That means they've found a more statistically reliable result. For example, the P value in the following article is P<0.001. This actually means that there 1 in a 1000 chance of the result being a random occurrence and a 999 in 1000 chance that the result was related to the premise of the study.

Just remember, look for at least a P<0.05. That means the results were statistically significant. Beyond that the more zeros you see in the P value, the better. Try your new-found statistical powers on the articles below. Look for the P values in the studies. What do they signify?

Now that you can evaluate scientific material at its source, you'll be less likely to fall for the 'man bites dog' con-jobs that are all too commonly reported in the news or as what passes for scientific discussion these days.

Resting heart rate as a low tech predictor of heart problems in women

In a large, diverse group of postmenopausal women, resting heart rate was an independent predictor of coronary events, with higher heart rate associated with greater risk. The relation between resting heart rate and risk of coronary events was stronger in younger postmenopausal women than in older ones. Resting heart rate did not independently predict stroke.

In general, age, body mass index, and saturated fat consumption were higher and cardiovascular risk factors such as hypertension, diabetes, smoking, hypercholesterolaemia, and depressive symptoms more prevalent in women with higher resting heart rate, as was self reported nervousness. Physical activity and alcohol use were inversely related to heart rate (both P<0.001), and heart rate was lower in women who used postmenopausal hormone therapy than in those who did not (P<0.001).

Article Link

Comment:

One can't argue that this is about as low tech a predictor of future health problems as one is likely to find. It has already been shown that resting heart rate predicts coronary events in men. For women however, the relation between heart rate and coronary events or stroke has been uncertain. The study broke the participants into groups including a 'high heart rate group' whose heart rate was greater that 76 beats per minute and 'low heart rate group' whose heart rate was greater than 61 beats per minute. The association with 'coronary events' (aka heart attacks and death). This association appears stronger in women aged 50-64 than in those aged 65 or older

Being overweight makes you age faster

Obesity and weight gain in adulthood are associated with an increased risk of several cancers. Telomeres play a critical role in maintaining genomic integrity and may be involved in carcinogenesis. Using data from 647 women ages 35 to 74 years in the United States and Puerto Rico (2003-2004), we examined the association between current and past anthropometric characteristics and telomere length in blood. These findings support the hypothesis that obesity may accelerate aging, and highlight the importance of maintaining a desirable weight in adulthood.

Article Link

Comment:

A telomere is a region of repetitive DNA at the end of chromosomes, which protects the end of the chromosome from destruction. When DNA needs to be read (to replicate itself, or generate RNA so as to begin coding proteins) a problem arises in that the enzymes that duplicate the chromosome and its DNA cannot continue their duplication all the way to the end of the chromosome. They need a blank area to 'park' much like the cassette tapes of days past had white 'leader tape' at their front and the back so that the tape head did not start in the song itself. Unlike cassette tape, every time DNA reproduces, a bit of the white leader tape, the 'telomere' at the end, is frittered off and has to be replaced. Telomeres and replenished by an enzyme, the telomerase reverse transcriptase. Telomeres protect a cell's chromosomes from fusing with each other or rearranging - abnormalities which can lead to cancer - and so cells are normally destroyed when their telomeres are consumed. In the women studies for this article, those having a higher body mass index (BMI) in their 30s were associated with shorter telomere length in their 40s (P < 0.01).

I suspect some of this association is epigenetic, and points again to the fact that the GT5 Warrior epigenotype may well need to get their weight profile optimized early in life and be increasingly calorie conscious as they age.

Vitamin D, adult-onset diabetes and metabolic syndrome

Vitamin D is a potent immunomodulator that also enhances the production and secretion of several hormones, including insulin. Vitamin D deficiency has been associated with increased risk of type 1 diabetes. Glycemic control and insulin resistance are improved when vitamin D deficiency is corrected and calcium supplementation is adequate.

Article Link

Comment:

More and more information is surfacing about vitamin D (actually more of a hormone than a vitamin) and insulin resistance. Studies consistently show that vitamin D levels in both North America and the Pacific are typically lower than optimal. In the USA , most vitamin D intake from foods is provided by fortification. Canada and New Zealand have fewer fortified choices, and intakes are correspondingly lower. The mechanism of action of vitamin D in adult onset (type 2) diabetes is thought to be to its role in the control of plasma calcium levels, which help regulate insulin synthesis, but may also be the result of vitamin D stimulating the insulin secreting (beta) cells of the pancreas directly. If you have a history of metabolic syndrome or adult onset diabetes in close family members you may want to consider adding vitamin D to your supplement regimen. However, make sure that you do it in partnership with a nutrition professional.

One from the vaults: Mom's blood type can influence child's risk of Strep (1978)

In a prospective study of maternal genital colonization with streptococci at the time of delivery, epidemiological data, including blood type (ABO group), were recorded for the 1,062 patients studied. Blood type B was found in a statistically significant (P <.005) higher proportion of patients colonized with streptococci (28%) compared with the total population (16.4%)

Article Link

Comment:

Evidence suggests that probiotic supplementation does change the vagina flora of women. Since it appears that the route of transmission of Streptococcus is from the birth canal, physicians should recommend probiotic supplementation for pregnant women beginning 3-4 weeks prior to expected date of delivery as a way to prevent streptococcus infection in neonates. This should be especially emphasized if the mother is either blood group B or AB.

This study again illustrates the fact that some of the best ABO correlation studies are outside the purview to today's physicians, most of whom would tell you that any research from 1978 is better suited to a history class than to any thing taught in medical school.

Since Mother's Day is fast approaching, also remember that recurring otitis media (ear infections) is strongly associated with the child's mother being blood type A. In fact the correlation here is quite startling. Children of mothers who are blood type A are twenty seven times more likely to get a second ear infection within one year of contracting the first. To give you an idea of just how strong this association is, look at the chart below to compare the RR (relative risks) of a few other disease/ lifestyle links.

Update: IfHI 2009

Just a quick word to the wise about the IfHI Conference, Norwalk Connecticut, June 5-7. We had run out of available rooms at the Dolce Center Campus. However 10 additional rooms have just been made available. Unlike previous conferences, where attendees could book almost to the day of the event, IfHI 2009 looks like it will be completely booked by the middle of May, a full month before the event. If you are planning on attending, either for certification or just personal enrichment, please make your reservations ASAP, especially if you want to stay overnight on campus.

Until next week.