Another diet study, frying is okay?

The BMJ has published a paper of 40,000 Spanish patients followed for 12 years. They compared people who ate the most fried food versus those who ate the least and concluded no increase in the risk of heart attack. The wrinkle is that the foods were fried in olive or sunflower oil. Before you celebrate, MedPage Today points out a couple of limitations. First, frying foods increases the calorie content, which could be bad for your health. Deep frying versus pan frying may have an effect and so could the number of re-uses of the oil.

As with the vast majority of these studies, this is observational, ie: someone tracked their habits and then ran the statistics to guess about risks. This is different from the scientific "gold standard", a randomized trial, but you can imagine how hard it would be to randomly assign some people to eat olive oil fried foods and others to eat, say, peanut oil fried foods. Besides, no one eats the same thing every day.

My take home point would be: if you are going to fry, use a healthy oil, but try to avoid it and bake or broil the majority of your food.

See the accompanying editorial for more opinion on the study.

Smoking ban in NC associated with fewer heart attacks?

In the beginning of 2010, North Carolina instituted a ban on smoking in bars and restaurants. The NC Division of Public Health has just announced data indicating that in the time period right after starting the ban, ER visits for heart attacks dropped by 21%. Many headlines only picked up on that top talking point from the press release, but some news outlets delved a little deeper into the complexities of the statistics.

Taking a look at the report, it is interesting to note that ER visits were already on the decline in NC for the two years prior to the ban. That trend could be part of the reason for the continued decline. By raw numbers, the number of heart attacks in women actually went up. Dr. Siegel from Boston University goes into analysis in greater detail and has concerns about the conclusions drawn from this particular statistical analysis.

From NC Tobacco Prevention and Control Report

So while perhaps this evidence from NC is not the strongest with regard to smoke-free legislation, a large, recently published, meta-analysis found a pooled risk reduction of about 10%, but they also admit that the data is heterogeneous. Regardless, seeing as how cardiovascular disease is still the number one killer in the US,  a 10% reduction is strong evidence in favor of widespread adoption of smoke-free legislation.

Ed note (11/22/11): another link about tobacco control legislation and its effects from the National Conference of State Legislatures including a US map of where you can find smoking bans.

How Physicians Should Evaluate Dietary Supplements

I found this older article in a pile of stuff "to blog" about and gave it a quick look over. It seems like a nice refresher for those who are unaware of some of the history about hos the US got to where we are presently when it comes to the regulation of dietary supplements. The authors describe a number of limitations to this regulation strategy which leave supplements inherently flawed. For example, products are not tested and held to standards regarding active ingredients, but are instead usually "standardized" based on the amount of ingredients that are put into the mixer. As such you may have varying amounts of ingredients depending on (just to name a few):

• cultivars vs. wild growth
• misidentification of plants
• stage of growth at harvest
• environmental conditions during growth
• methods of harvest and processing
• misidentification of plant parts (leaf vs. berry)
• environmental contaminants (heavy metals, microbes, insects)
• improper storage
• variations in temperature, moisture, or light
• effects of solvents on active ingredients
• binders and fillers in capsules affecting bioavailability

To answer the "What's the harm?" question, the authors briefly discuss ephedra and its aggressive marketing and misuse in the 90's which led to several reported deaths and who knows how many unreported injuries, hospitalizations, and fatalities. I was surprised, then, when the authors took a very soft conclusion on how they suggest clinicians address these issues with patients. They suggest that taking a history which includes supplement use is important, which I agree with. The authors do not, however, encourage the cessation of supplements, not even for those without evidence. Perhaps I misread the conclusions, but I would have thought that it would be generally good advice to avoid taking unproven remedies with no guarantees of safety and the potential for adulteration with unknown contaminants.

Curious, I looked further into the credentials of the senior author, Larry A. Walker PhD. He is on faculty at Ole Miss. He also serves as a member of the National Advisory Council for the NCCAM. His PubMed results are a little hard to weed through since "walker LA" is a relatively common name and many results go to research which appears outside his scope. His google search results were interesting. Of the first couple of dozen, many point to identical text in which Dr. Walker seems to be supportive of a mushroom-based product for immune support, and specifically a compound called "grifolan". So this is where I decided to indulge in a little trip down Alice's rabbit hole of the internet.

The text is quoting him could easily be taking some other reference out of context and I have no idea if he actually supports the product making the claims. So, I added "grifolan" to PubMed for "walker LA", but that does not yield any results. Searching for "grifolan" alone does generate 44 results, with several reviews of its benefits and, perhaps not surprisingly, no actual experiments involving actual humans. What is really amusing is to search google for "grifolan larry walker" which gets you dozens and dozens of websites all selling mushroom extracts and using the exact same quote: 

"Grifolan has been shown to activate macrophages, a type of cell consider the " heavy artillery": of the immune system, explains Larry A. Walker, Ph.D"

This amazing stuff seems to be good for diabetes, hepatitis, and is even sold at naturalpetrx.com. The echo chamber continues on as I found an Amazon reviewer who has three reviews of different herbal products, each of which reads like it is copied right of the back label of the bottle. Fascinating how out of the ether, internet traffic and content that is posted and reposted can eventually build into something appearing to be based in fact, despite a mere handful of actual scientific studies.

Retraction of scientific journal articles

For the most part, it is a pretty big deal for a journal to retract a scientific article after it has been published. In doing so, either the authors of the manuscript or the editors of the journal are admitting significant breakdown in some portion of the scientific method and/or publication process. Sometimes, the retraction is because a manuscript is sent to and accepted by more than one journal at the same time. This is never supposed to be done, but some do it anyway, either as a mistake or as an attempt to get more exposure for the same amount of work. Sometimes, someone realizes that a significant error was made in data collection or processing, rendering the conclusions invalid. The worst though, are the instances where true wrongdoing or fraud has occurred.

In an ideal world, peer review would prevent this. Peer review is a process whereby journal editors provide manuscripts to the author's peers, scientists in the same fields of study, who provide the editors with an unbiased review of the quality and scientific accuracy of the manuscript. Having reviewed manuscripts for about a dozen or so journals, I can say that I have certainly read my fair share of manuscripts that I consider to be flawed in design or interpretation that need to be seriously reconsidered for publication. Unfortunately, however, that opinion only gets shared with the editors of the journal the article was submitted to and with the authors themselves. If the article is not accepted, the authors are free to resubmit somewhere else. Hopefully reviewers for other journals would draw the same conclusions, but they may not have the same experiences as me, and clearly I lack many experiences that others have. Furthermore, journals depend on reviewers doing work gratis, and therefore are limited by those willing to do the work. You can imagine, then, how some substandard manuscripts manage to make it through peer review from time to time.

I digress, but the point of this post is to highlight a blog that covers this topic by scanning journals for retractions and posting them for all to see. RetractionWatch tracks retractions but also provides background info on some of the more egregious offenders. It has received some excellent reviews since its inception only one year ago, and I have added it to my RSS feeds for regular updates.

Coffee, this time it's safe

People accuse doctors and researchers of never being able to make up their mind about certain things, such as how much exercise is healthy and how much is dangerous, should I eat eggs or not, and is coffee good or bad? Part of the reason we "cannot agree" is that each study tests a different group of people with different diseases and looks for different events over different time frames and using different statistical tests.

Well, this post is about coffee and it is a study in the "coffee is okay" category. The study comes from a cohort, which is basically where a whole bunch of people answer a lot of quesitons about their health and their habits, then they get rechecked several times over the following years. So, the people studied were not specially recruited to answer this specific question, and, in fact, the Nurses Health Study has had dozens of publications from it over the years. It started in 1976 and followed over 100,000 nurses for decades.

This is a good method of research because you can ask lots and lots of questions, but the drawback is that it is not conclusive. But, trying to find thousands and thousands of people and making them drink the same amount of coffee every day and then comparing that group to thousands of others not drinking coffee would be nearly impossible, so sometimes we have to settle for this kind of imperfect data.

So, long story short, a study of 12,000 nurses showed that over a 20 year period, regular coffee drinkers were no more likely to die than non drinkers. This was an especially important analysis because it took only the 12,000 nurses who had a history of a stroke or heart attack, suggesting that for people with these diseases, it is likely safe to drink caffeinated coffee. I mentioned these types of studies have limits in what conclusions we can draw and the authors pointed out that since the study was only in nurses, it may not reflect what is true in a general population. While the researchers used statistics to account for the effects of age, blood pressure, and diabetes, there could easily be many other factors that were not measured and therefore could not be accounted for.

So while we might not know the absolute truth about the safety of coffee, I will continue to enjoy mine.

Doctors choose differently for themselves than they would for their patients

In an investigation released on April 11 in the Archives of Internal Medicine, several hundred doctors were surveyed and asked questions about treatments that they would accept for themselves and ones they would recommend for their patients. Both hypothetical scenarios were situations where taking a treatment (avian flu vaccine or colon cancer treatment) carried some risk of disability from the treatment, but the treatment would reduce the chance of death. In both cases the doctors were more likely to recommend the treatment for their patients than they would be to take the treatment for themselves.

The authors took a pretty neutral stance about this in their conclusions and basically just said that something is going on in the doctors' minds which is generating this difference in self treatment and recommended treatments. The authors also do a reasonable job of questioning their own conclusions and pointing out limitations in this study. The first thing that came to mind for me is my awareness of bad outcomes. For example, when I see patients in the hospital, it is usually because they have had some complications from some disease they have or some procedure that they had done. (eg: bleeding after a catheterization, atrial fibrillation or heart attack after surgery, etc.) I would imagine that for many doctors, the awareness of these bad outcomes (which are usually rare) loom large in their mind when asked to make decisions about treatment for themselves.

Do you want a doctor that clowns around?

Medgadget reported last month on a study in Fertility and Sterility which randomly assigned women undergoing IVF to an encounter with a "medical clown" and showed an increase in successful births from 20.2% to 36.4%. Now, we might all be able to imagine that you could use clowns to reduce pain scores in children, or possibly to scare the living daylights out of people with coulrophobia, but improving fertility? Perhaps there is something about the hormonal environment that improves fertility when the mother is happy and laughing? The abstract is the only text of the article I can access right now, but the statistics they report suggest that the effect might be real, but it might also be a spurious finding and we will not really know until the results can be replicated. In the meantime, I will have to entertain myself with the idea of a "clown consultation service", and I will also have to refrain from any jokes about which other specialties should be first in line to staff such a service.

Safety and efficacy of glycoprotein inhibitors

Published in the Journal of the American College of Cardiology is, me! I am the first author on a meta-analysis in JACC along with several of my colleagues here at UF. A meta-analysis is a study where we take the results of several other studies and combine the results to create a kind of "super-study" with may more patients than a single study would have. The upside is, more patients, more powerful study. The downside is, each study is done slightly (or significantly) different. We use statistics to minimize the differences and measure how different they are, and if done well, can provide important information for doctors.

In this case, we measured the effect of a class of drugs called glycoprotein IIb/IIIa inhibitors. These are anti-platelet drugs which prevent clotting when you are putting in a stent. Since they prevent clotting, they can also cause bleeding and so we were trying to measure the benefit and the risk with these drugs and help decide if overall they are a good idea to continue using. Are results show that they reduce heart attacks and increase minor bleeding and thus, on balance, are helpful. These are drugs given IV, so they are not anything the public would normally ever know about, but the research is interesting and having it published in JACC is great.

Also, see the news release from UF College of Medicine, and the editorial written in JACC. Press release also picked up by the Chipley Bugle and physorg.com.

New journal does not care how good your manuscript is

For anyone who has ever tried to get anything published, an essay, a book, a research article, you can appreciate the humor of the Journal of Universal Rejection. Sometimes I wonder if my manuscripts have been sent here instead of to the intended journal.

8 year olds, building their resume with science!

A group of hard working children aged 8-10 years (with some help from their teachers) conducted a science experiment. Big deal, right? I mean most of us remember elementary school science fairs with "experiements like and adding vinegar to baking soda and food coloring to make a "volcano". These children are different because they came up with a hypothesis, designed a test to answer their question of interest and generated scientific results which are published in a peer-reviewed journal, Biology Letters, a publication of the Royal Society in Britain.

The manuscript is not published in a fully traditional format (ie: there are no references and the jargon is minimal), but give me a break, they're only 10 years old! The children drew the following conclusion:

Principal finding ‘We discovered that bumble-bees can use a combination of colour and spatial relationships in deciding which colour of flower to forage from. We also discovered that science is cool and fun because you get to do stuff that no one has ever done before. (Children from Blackawton)’.

 Yes, science is cool and fun. Congratulations to you all and please, please, keep doing science!

Another tool to help medical students decide on a career

A quick recap of medical education before I get into the item of discussion. The traditional sequence of educational events for foctors in the U.S. if to first complete a bachelor's degree (at a minimum). This is most commonly done in a scientfic field such as Biology, Chemistry, Microbiology, etc. Two notes are that 1) for the most part, there is not really a "pre-med" major, you have to get your degree in some field and 2) more and more people are going to medical school with degrees in humanties and social sciences which medical schools are interested in matriculating to enhance diversity of the student body.

After this comes four years of medical school. The first two held in lectures/classrooms with a courseload somewhere around 30 credit hours per semester (yes...). Third year is in clerkships experiencing what different specialities of medicine are like. The fourth year is a mishmash of elective, required additional training, interviewing for residencies, and a little bit of goofing off with your classmates.

Many future doctors decide on a specialty sometime during their third year. Some know from the time they start (ie: Mom/Dad did such-and-such and I want to do that too). Some struggle and change their minds frequently, even up until the last minute and might even interview for multiple specialties.

R Scott McCain et al have published in the British Medical Journal, a new strategy for answering this crucial question. They observed all the doctors at their hospital as they were arriving for the morning and parking their cars. The researchers watched to see how many doctors had their parking-deck swipe cards out and ready when then pulled up to the gate and then how long it took them to park and exit the lot. They knew the identity of each doctor and then analyzed the data based on gender and their specialty of medicine.

The results? No differences in gender, men and women parked in the same amount of time, on average. There were differences by specialty, however. Surgeons parked the fastest and internists parked the slowest. The authors postulate that medical students should therefore be monitored on arrival to school and directed towards different medical specialties based on the time it takes them to park.

Sound preposterous? Well, it clearly is. This research was published in the BMJ's Christmas issue, in which they traditionally include some piece of tongue-in-cheek "research" which is good for a laugh. I wanted to present it because I enjoyed it and thought it was funny, but also to illustrate an important point, which is that people who do not know any better frequently make such logical errors and use "statistically significant" findings to justify whatever they want to prove.

In this case, the study is a cross-sectional study. It takes a single measurement of some outcome across several groups and demonstrates a difference between the groups. The study's design does NOT make any measurements over time to be able to predict anything about future performance in parking. Even more absurd then is the author's conclusion that a person's parking time could possibly predict the type of medicine they would go to practice. That would require a completely different study design. Lastly, to actually test their proposed conclusion that parking times could predict future medical specialty, you would need to perform a randomized trial and either let people pick their own specialty (control group) or use their time to select for them (experimental group) and then measure some outcome of whether of not you were right (physician happiness?) years later.

These realities are a big part of why people promoting junk science use shortcuts. Randomized trials are hard and take time and resources. Cross-sectional studies are easy and fast, heck, you can just send an undergrad to a parking lot and gather enough data to generate "significant" results. If you can then fool people with those data, why bother gathering the high quality stuff?

This is a blogpost about a news article about a scientific paper

I love dry humor, and this bit is a little old but an excellent read. Martin Robbins, writing for the Guardian in the UK, posted on Sept 27, 2010 an article titled "This is a news website article about a scientific paper". He goes on with the subtitle:

In the standfirst I will make a fairly obvious pun about the subject matter before posing an inane question I have no intention of really answering: is this an important scientific finding?

In a great lampoon of prototypical media failures when it comes to reporting on medical and scientific breakthroughs, including:

• Scare tactics
• Failing to link to the paper and not providing enough information to find it on your own
• Overstating the inadequacy of existing literature
• Oversimplification of the findings
• Failing to delve into sources beyond Google and Wikipedia
• Attaching the science to some minor celebrity's pet cause
• Meaningless figures and graphics
• False balance of competing opinions to generate controversy

I have friends in media, some of whom do an excellent job with their reporting, so I do not mean to suggest that all reporters use these tactics when reporting on science. However, the parody would not be funny if it were not (unfortunately) true so often.

Can Thanksgiving dinner cause a heart attack?

The Washington Post reported on 11/22 that one should be wary of eating a large meal because the risk of having a heart attack might be four times higher for the 2 hours after eating. They gloss over the fact that the science they are reporting is about 10 years old and was apparently presented in abstract form only at the American Heart Association's Scientific Sessions back in 2000. The first author, Francisco Lopez-Jimenez, has written quite a bit over the last several years, but I can't seem to find the abstract being referenced, or any full-text article about the research. I did find one other reference dated from 2000 in Medscape which describes the study as a case-control study of nearly 2,000 patients suggesting a correlation between large meals and heart attacks.

While this study design is useful, it cannot establish causation, only correlation. Furthermore, this is a classic example of the marked difference between relative risk and absolute risk. The RELATIVE risk of having a heart attack after a large meal may be quadrupled, however the ABSOLUTE risk of having a heart attack in any given 2 hour period is so vanishingly small. Let's say that your risk goes from a risk of one in a 1,000,000 to one in 250,000; this is technically a quadruple risk, but still negligible and not something to worry about.

So unless you are trying to get out of eating some foul dish that a family member is pushing you to eat, enjoy your feast and be thankful to have an over-abundance of food, and family to share it with. That, and go for a walk afterwards since Bloomberg is reporting that one half of Americans will have diabetes or prediabetes by the end of the decade!!!

Counterintuition and the importance of clinical trials

Justifying or adopting things which "just make sense" can get you into a lot of trouble in medicine. This kind of mental shorthand is frequently used by people who are interested in promoting products or ideas without doing the science to confirm that they are effective. Many of the nutritional supplements available for purchase have some degree of science behind them, but just because something works in a lab, doesn't mean it will work when you put it in a human.

This is where the role of the randomized clinical trial becomes critical. I want to give you a few examples of ideas which "make sense" but have been modified or soundly rejected after the appropriate science was done.

• Let's start with an easy one, "inotropic agents". Let's say your heart doesn't squeeze well and you have heart failure. There are drugs (inotropes) that can make it squeeze harder. Sound like a good idea? Turns out that using inotropes shortens the average life span for heart failure patients. This might be for a number of reasons, but it turns out that these drugs increase the risk of fatal arrhythmias and they increase the workload of the heart possibly worsening the disease process.
• On the opposite side of the coin, we have beta blockers. These are drugs hat slow down the heart and "make it weaker". There was a time when beta-blockers were contraindicated for patients with heart failure, after all, who would take a patient with a weak heart and give them a drug that slows down and weakens the heart!? It turns out, though, that in addition to slowing the heart rate, these drugs decrease the demand for oxygen and have other beneficial effects on heart health to the point where beta-blockers are a standard of care for anyone with systolic heart failure (where the heart doesn't squeeze well)
• Coumadin (warfarin), aspirin, Plavix (clopidogrel), heparin, fondaparinux, and several other classes of drugs are all blood thinners. So, if all these drugs thin your blood, why do doctors prescribe warfarin so much? Simple, that's what the evidence says to do. All the drugs thin your blood and increase your risk of having some kind of bleeding, but if warfarin reduces your risk of stroke, and Plavix doesn't, why would you take Plavix?
• After people have heart attacks, the scar makes the heart more prone to arrhythmias. This is a big problem for some people because the arrhythmias can be deadly. Now, let's say you hang around in the hospital for a couple of days recovering from your heart attack and your doctor notices premature heartbeats. We have drugs that can make those early heartbeats go away. Maybe if we do that, we can prevent arrhythmias and prevent people from dying. Well, it's a good thing someone thought to to a trial of that, because it turns out that giving drugs to suppress those early beats actually increases your risk of dying! This has to do with the fact that some anti-arrhythmia drugs can actually slightly increase your risk of arrhythmia. (Source: the CAST trial)
• If you have high blood pressure, how low should you try to make it go? Is less than 130 okay? What about 120? 110? At some point dropping blood pressure becomes a bad thing. Thankfully more and more clinical trials are helping to home in on the solution to this problem, however the same conundrum exists regarding diabetes. How low should your blood sugar be? It turns out that recent trials have shown that while controlling blood sugar is crucial at preventing progression of the disease, trying to get your blood sugar down to where a normal person's is could cause more complications (hypoglycemia) and not be beneficial.
• When you have surgery, one many potential complications is to get an infection at the site of your surgery. Thankfully, we have antibiotics that prevent that, right? It depends. A big trial in the New England Journal of Medicine (Bennett-Guerrero et al. Sept 9, 2010) showed that using an antibiotic coated sponge actually INCREASED the number of infections!
• Last one I'll mention for now is a drug called torceptrapib. This potential blockbuster was the first agent that was thought to directly increase your "good" (HDL) cholesterol. People were very excited because statins, which lower your "bad" (LDL) cholesterol have made billions and billions of dollars for the pharmaceutical industry. Unfortunately, when they tried it out in people, the HDL molecule it "made" didn't seem to work quite as well as the ones you make on your own AND, the drug raised blood pressure. It never made it to market, because despite sounding like a good idea, it didn't pan out when someone actually tested it in real people and measured a meaningful clinical outcome.

Anyway, these are just a handful of examples of how your intuition can lead you astray and just because something is PLAUSIBLE does not make it fact.

Lab tests are not all the same

Cross-posted from http://www.nevernotrunning.com

TIME Healthland posted an article on Oct 25 regurgitating the results of yet another study which "discovered" that after marathons, the same lab test that goes up in a heart attack (troponin) goes up after running 26.2. While I am sure the scientists reporting the study results did a fine job, the reporting could use some work. First, there are literally dozens of studies that have already shown this. Second, no one that I am aware of has shown that the increased troponin in marathon runners means a damn thing. For example, if your troponin goes up after a marathon, are you going to die sooner? Have a heart attack or suddenly die during a race? Those are the important things, not that your lab test is abnormal.

Lots of things can make your troponin go up, like getting in a car wreck and crushing your chest against the steering wheel. Or getting an infection on your heart valves. Or having a viral infection. Why aren't we comparing marathon running to these things? There is a lab test called the GGT. It is a marker of liver function. If you go out on a bender and have a half a bottle of tequila, your GGT will go up. Your GGT will also go up if you catch hepatitis, or have liver flukes, or eat magic mushrooms, or OD on acetaminophen. Now, is anyone out there saying that drinking tequila is the same as all these things?

It is certainly possible that endurance sports are doing something unsavory to the cardiovascular system, but at the present time, I am not aware of any such evidence and until you have something convincing, I wish the media would stop trying to scare people away from running, when all the available evidence points to it being healthy!

TEDMED 2010: Medtronic unveils pacemaker without wires

Engadget posted today that Medtronic has unveiled a pacemaker that can be inserted via a catheter (basically the same way a stent goes in) as opposed to the current method which requires minor surgery, cutting through the skin to implant a small box under the skin. They also demo'ed an interface which shows the pacer being monitored via smartphone. This goes along with a defibrillator recently published in NEJM where the leads go under the skin instead of into the heart.

This kind of tech has some great advantages such as easier insertion, less complications during and after implantation. I wonder how they are dealing with battery life and how these could be replaced.

Winchester et al. Evidence for Preprocedural Statin Use

I'm very excited to announce that after a year of hard work, one of my research projects is coming to fruition with e-publication in the Journal of the American College of Cardiology. Our meta-analysis of 21 trials and over 4,800 patients showed that giving a statin prior to an invasive procedure results in a 40% reduction in postprocedural MI after cardiac catheterization and non-CABG surgery, as well as a 40% reduction in post-CABG atrial fibrillation. The article is accompanied by an editorial from Kim Eagle, and will be published in the September 24 issue of the Journal. Simultaneous with e-publication, we presented our research at the European Society of Cardiology with some coverage being picked up by Medpage Today. Enjoy the science!