Discover more from The Arrow
The Arrow #134
Greetings from Montecito.
Okay, you will no doubt be pleased to learn that after all my weeks of whining about cold weather here, it has finally turned hot. Not Dallas hot, but hot nonetheless. So, if you’ve been lying awake nights worried about my being cold, you can sleep well now.
The Arrow is a reader-supported publication. It costs a coffee per week to subscribe. And I’m not talking some fancy, high-end, high-carb, Venti Caramel Frappuccino with Extra Caramel Drizzle and Whipped Cream Starbucks crap; I’m talking a Joes’ Diner drip coffee. To receive new posts and support my work, consider becoming a free or paid subscriber.
Fauci Under Fire
A new email has emerged as a consequence of all the FOIA requests bombarding various ranks of the government. One of them caught Anthony Fauci with his metaphorical pants down.
As you’ve probably seen, Dr. Fauci and Senator Rand Paul have gone at it numerous times in various senate hearings. From the outset, Paul has accused Fauci of both approving and funding the Wuhan Institute of Virology (WIV) pursuit of the kind of gain of function research that probably foisted SARS-CoV-2 on the world. Fauci vehemently denied it at first, but as more information started seeping in, his denials began taking a different form. He claimed “gain of function” was a “nebulous term,” and that what he funded was something totally different.
You can go to YouTube and find multiple videos of their exchanges. Here is a short, but intense one in which Dr. Fauci out and out denies that he had anything to do with either approving or funding gain of function research. And he tells Rand Paul that he (Rand Paul) doesn’t know what he’s talking about. Fauci brandishes a paper saying what WIV did “was judged by qualified staff up and down the chain as not being gain of function.”
It’s funny that Fauci waves this paper around (that I could not read the title of, so couldn’t look it up) saying it was “judged by qualified staff…” when it is his staff who judged it. He’s the head guy. More about another such paper later.
In the multitude of emails coming to light was this one from Fauci to a group working on a paper about the origins of Covid.
In case you can’t read the part in yellow, I enlarged it below.
It is known that Fauci’s organization the National Institute of Allergy and Infectious Diseases (NIAID) funded the Wuhan lab through the Eco Health Alliance. But Fauci always denied that the Wuhan lab was involved in gain of function research. The Obama administration put a moratorium on funding such research in 2014, but the NIH lifted the ban in late 2017. Fauci would have been in the clear legally thereafter to fund gain of function, which asks the question, Why did he deny it?
From my reading of the email, it’s pretty clear that Fauci was aware that Wuhan was doing gain of function research. You may have a different view. Let me know in the comments.
The hoard of emails released show that the group of scientists Fauci consulted had grave doubts as to whether the virus had really jumped from animal to man. Some had noticed non-zoonotic sequences on or around the furin cleavage site, which led them to believe the virus was created in the lab. But after much discussion, Fauci basically wrangled them into writing a paper denying the lab leak hypothesis while throwing their collective weight into the idea that the virus came from an animal.
I’m not inside Fauci’s head, so I obviously can’t tell you what he was thinking, but it’s pretty clear that he was dead set on doing everything possible to promote the zoonotic hypothesis. It was looking early on like the pandemic could be disastrous, and it would have been equally disastrous to Fauci’s reputation were it shown the virus came from a lab he had funded. Even if the mandate prohibiting funding of gain of function research had been lifted.
The end result of all his cajoling was that the group of scientists published a paper in The Lancet titled “The proximal origin of SARS-CoV-2,” describing how the virus hopped from an animal to humans. The article is highly technical, but here is a paragraph that shows the crux of the argument that SARS-CoV-2 could not have resulted from a lab leak.
It is improbable that SARS-CoV-2 emerged through laboratory manipulation of a related SARS-CoV-like coronavirus. As noted above, the RBD of SARS-CoV-2 is optimized for binding to human ACE2 with an efficient solution different from those previously predicted. Furthermore, if genetic manipulation had been performed, one of the several reverse-genetic systems available for betacoronaviruses would probably have been used. However, the genetic data irrefutably show that SARS-CoV-2 is not derived from any previously used virus backbone. Instead, we propose two scenarios that can plausibly explain the origin of SARS-CoV-2: (i) natural selection in an animal host before zoonotic transfer; and (ii) natural selection in humans following zoonotic transfer. We also discuss whether selection during passage could have given rise to SARS-CoV-2. [My bold emphasis]
This after numerous emails among those involved discussing how the viral sequence shows a great deal of evidence of being manipulated in the lab.
To give you a sampling, here is a short humorous video of the various comments taken pretty much verbatim from the emails. Here is what the article about the video says:
If this were a movie we’d have to say it was “Based on a true story,” though it’s more like “very closely based.” Most quotes are pulled from leaked correspondence of figures like Dr. Anthony Fauci and the Proximal Origin scientists written about in Public and Racket last week. Some, however, are paraphrased by Orf [the guy who put the video together], who of course assembled the story as well. For example, “I literally cannot think of any possible way this could have come from nature!” is based on “I just can’t figure out how this gets accomplished in nature,” from Dr. Bob Garry on February 2, 2020: [My bold emphasis]
The emails are so revealing that there is now a petition being circulated calling on The Lancet to retract the Proximal Origins article. The reason for the demand for retraction comes not because the article itself has turned out to be inaccurate—many scientific articles written in good faith have been shown by later work to have been inaccurate—but because the emails prove the authors knew it was inaccurate when they submitted it. A big difference.
And while these authors themselves express in the emails grave concerns about the lab leak hypothesis being true, they accused others who raised the issue as being conspiracy theorists.
If you want to read more in depth about the scientific outrage over this article, take a look at this Substack by Roger Pielke, a scientist at the University of Colorado.
You can also read investigative journalist Paul Thacker’s piece on the article, but it’s behind a paywall. Here is a sampling:
Many of the virologists’ internal emails and Slack messages began leaking onto Twitter, followed by a joint Public and Racket investigation. The messages showed scientists were deeply concerned that the COVID virus could have been engineered or leaked from a Wuhan lab, even as they publicly ridiculed such thinking as a “conspiracy theory.”
In one example, Andersen [the lead author of the Proximal Origins article] wrote his colleagues on February 1, 2020, in a private Slack message, “I think the main thing still in my mind is that the lab escape version of this is so friggin’ likely to have happened because they were already doing this type of work and the molecular data is fully consistent with that scenario.” [My bold emphasis]
Fauci must be scrambling.
Mainstream Media AWOL Again
This may sound like a political rant, but I can assure you it is not. It is a rant about the total and complete abdication of the mainstream press in the United States when it comes to reporting anything negative about the connected class, or as Thomas Sowell calls them, The Anointed.
On Tuesday the Daily News, a UK newspaper (not an American newspaper) published an “exclusive” story about the latest on Hunter Biden’s sweetheart deal with the Department of Justice. The facts are pretty clear as he has admitted to them. He filed false tax returns and failed to pay a substantial amount of federal taxes. In fact, the amount was large enough that it called for a minimum of 12 months in a federal prison. Along with those charges, Hunter, by his own admission, falsified an application for a gun purchase while he was in the throes of a cocaine addiction, an infraction than could incur a 10-year prison sentence.
His lawyers had worked out a deal whereby he would plead guilty and get off without prison time.
All this is beyond dispute.
A lot of people were upset at this slap on the wrist, because they knew that if they had done the same thing, they would be headed for the hoosegow. The only possible reason for Hunter to be treated so leniently was that he is the president’s son. At least that’s the assumption. Had Eric or Donald Trump, Jr. done the same thing while their dad was in office, can you imagine the same treatment for them?
Now here’s where it gets interesting.
According to the Daily Mail, one of the committees in the House of Representatives, after hearing testimony from the IRS whistleblowers, decided to file a brief in the case adding this additional testimony in an effort to squelch the sweet deal. I have no doubt that had this been one of Trump’s kids instead of Biden’s, a Democratic House Committee would have done the same. As per how these things work, Hunter Biden’s lawyers were copied on the brief.
Then, supposedly, one of the GOP committee lawyers called the court clerk and said the brief was filed in error and to basically throw it away. But in fact the GOP lawyers had done no such thing, and when they discovered the brief wasn’t available on the court’s site, they called to say WTF?
The people at the court told the GOP lawyers that someone from their office called to have the brief ditched. We did no such thing, the GOP lawyer said.
After a bit of checking, it turned out that someone in Hunter Biden’s lawyers office had called and made the request.
As you might imagine, the court was not happy.
You can read all about it from attorney Jeff Childers’s column yesterday.
When the parties arrived in court yesterday, there were some fireworks. The judge apparently felt she had been flimflammed a bit because as the proposed settlement was structured, Hunter was to get immunity for FARA violations and pretty much everything else. The judge caught on and said No.
The deal as agreed to was pulled off the table. Hunter ended up pleading Not Guilty. And the whole deal went back to the drawing board.
My bet is that the two sides will end up working a deal out that the judge will approve and that may be a bit worse for Hunter. But who knows?
What amazed me about this whole fiasco is that the mainstream media refused to cover it. I knew nothing about it until I saw the Daily Mail article. And the Childers piece.
Can you imagine if…
Why Do Doctors Push Statins?
One of the most common things I come across are tales from people whose doctors are not just strongly recommending their patients take a statin, but are threatening to fire them from their practice if they refuse.
One of the commenters in last week’s Arrow brought it up again. He wrote about his brother from Michigan who
has had series of doctors (PCP and specialists) who tell him to take statins and if he balks they tell him they will not be his doctor. Currently his Dr wants to give him a stronger dose using Crestor. All his heart metrics are good except a recent calcium test was only borderline.
This commenter goes on to ask if it is perhaps something in the water in Michigan that makes physicians there act like this.
I had a similar experience in Michigan about ten years ago when my father was still alive. He was 86 years old at the time and had a demyelinating disease that caused him to become progressively weaker until finally he was totally paralyzed from the neck down. He had to be hospitalized because he had a mild heart attack. His EKG didn’t change much, but he had some mild enzyme changes showing perhaps a bit of heart muscle damage. His labwork showed all his various cholesterol readings to be perfect. I had monitored his labs myself for years, and he always had low cholesterol levels, so nothing changed there.
The doctors at the hospital put him on a pretty large dose of the statin Lipitor. It was simply knee jerk idiocy. He was totally paralyzed, he wasn’t going to get any better—only worse, he required round-the-clock care at home, and he was miserable. Why stick him on a statin? It was insane. As you might imagine, I went ballistic and intervened. I encourage you to read this post I wrote about the entire affair while it was fresh in my mind. As you will discover if you read it, the medical treatment in Michigan is not what I was used to.
So, why are doctors so adamant about pushing statins?
Statins have been around for over 35 years. They weren’t around when I was in medical school, but the vast majority of physicians in practice now went to medical school when statins were available. And since much of medical training is influenced by the pharmaceutical industry, most of them learned that statins are a wonder drug.
As I’ve written many times before in these pages, the amount of material that must be learned, absorbed, and regurgitated on test doesn’t leave a lot of time for critical thinking. Most medical students don’t read the medical literature unless a particular paper is assigned to them to read. Most get their info from lectures and standard textbooks. And swallowing the firehose of info doesn’t leave much room for argument.
So, we have a large group of doctors now practicing who have heard nothing but glowingly positive things about statins up to and including that they are wonder drugs. In fact, I read a comment by someone who should have known better saying that statins have probably saved more lives than antibiotics. Which in an absolutely insane thing to say. But that’s the mindset of most doctors today.
The folks who run Big Pharma are clever. They seek out physicians who are involved with big name medical facilities—The Cleveland Clinic springs to mind—and/or on the faculty of prestigious universities. These are the folks who have all the alphabet of letters behind their names. Big Pharma basically recruits these people and puts them on the payroll. They are sent to wonderful places to give talks on specific drugs and are given grants for studies. They write papers extolling the virtues of statins. And up until not too long ago, they didn’t even have to disclose their relationship with the companies whose drugs they were writing about. Life is great if you play nicely with the drug companies.
Because these same physicians have all the merit badges of academic achievement, they are also the ones who are called upon to help set the practice guidelines for the various medial associations. When these guidelines call for the use of a statin if LDL-cholesterol levels are above some level, then if you’re a doc in practice, you’d better recommend a statin if one of your patients fits the guidelines.
If you do put your patient on a statin, and he/she has a heart attack—believe me plenty of patients have heart attacks while on statins—you’re golden, because you followed the guidelines. If, however, you don’t put your patient on a statin, and that patient later on does have a heart attack, you could be in for a nasty malpractice suit.
Therefore just about all doctors prescribe statins out of self defense, if for no other reason.
If the patient refuses to take a statin, then all you have to do to CYA is to write in the chart “I strongly recommended statins to this patient, but he/she refused to comply.” Done. You’re in the clear.
I don’t get this notion about firing your patients from your practice if they don’t do what you say. As a physician, I always thought of myself as being in a cooperative venture with my patients. I gave my patients the best advice I had to give, but if they didn’t want to take it or had issues with it, then we discussed it as a team and worked something out.
In my view, doctors shouldn’t look upon the doctor-patient relationship as a master-serf relationship. It’s a partnership. The patient, as a partner, has a say.
When I went to my website to get the link to the post about my father, I made a quick run through the comments and found one I had forgotten about. It was a comment from a true statinator of the deepest dye. The very kind of academic with all the merit badges that the drug companies love to see, and which, of course, he included in his comment. Hell, for all I know, he is on one of the committees that comes up with the guidelines.
Here are all the letters behind his name and all his academic qualifications that he included in his comment to my blog post:
MD FACP FNLA NCMP
Prevention of Cardiovascular Disease and Women’s Menopausal Health
Assistant Professor of Medicine
Johns Hopkins University School of Medicine
Diplomate American Board Clinical Lipidology and Fellow, National Lipid Assn.
Certified Menopause Practitioner: North American Menopause Society
Director: Baltimore Lipid Center. Board Member, South East Lipid Assn.
He did not reply.
Cholesterol, Low-Carb, CAC, and Statins
A couple of weeks ago I wrote about how LDL is made and transported from the liver. A few days ago someone sent me a nice video from a physician in Australia who went over much the same information with a little extra thrown in about coronary calcium screening and statin-induced diabetes.
The video is great. It is both succinct and comprehensive. You’ll learn a lot in the 20 minutes it takes you to watch it. Highly recommended.
While we’re on the subject of the effectiveness of low-carb diets, let’s take a look at one for the gents.
Hypogonadism and the Low-Carb Diet
I discussed last week how a low-carb or ketogenic diet could help women who have hot flashes. Now it’s time for one for the men. Before we get to the gents, however, based on the comments I received on the hot flash situation, I must not have done a great job in explaining.
The hot flashes occur when blood sugar drops. When that happens, the brain yells for help by releasing norepinephrine, which floods the body and raises blood sugar. And also causes an increase in body temperature and flushing.
Now, on to the men and their troubles with aging.
I came across a study a few days ago that I found pretty interesting for a lot of reasons. I’ll spend a bit of time going over it because it is illustrative of a number of things people often take at face value in reading these kinds of studies that can lead to misunderstanding.
The study titled “The effects of a low carbohydrate diet on erectile function and serum testosterone levels in hypogonadal men with metabolic syndrome: a randomized clinical trial” was designed to see if a low-carbohydrate diet might help with hypogonadism, erectile dysfunction, and testosterone levels in men with the metabolic syndrome.
First thing to note is that this is a really small study. There were only 18 subjects overall who were randomized into two groups. The control group had 6 subjects while the study group had 12. The 18 subjects were randomized using “the Random App from the Apple Store,” for whatever that’s worth.
All of the subjects in both groups had the metabolic syndrome and were alike in a number of ways. All subjects suffered with hypogonadism, which sounds like small gonads/testicles, but is defined by the authors of the study as a
…total serum testosterone < 300 ng/dL and clinical manifestations such as cognitive deficit, decreased lean mass, lower bone mineral density, erectile dysfunction, decreased libido, etc.
The study group on the low-carbohydrate diet
…was instructed by a nutritionist, member of the study team, to reduce carbohydrate intake and increase protein and fat intake. Their diet could not contain more than 25–30% carbohydrates per day, aiming for 20–30 g carbohydrate per day.
When I went to the reference for the diet to see more specifics, it said it was from a study on polycystic ovary disease and the low-carb diet. That diet
…consist[ed] of fewer than 20 grams of carbohydrate per day, as tolerated throughout the 6-month study period. The diet includes unlimited consumption of animal foods (meat, chicken, turkey, other fowl, fish, shellfish), prepared and fresh cheeses (up to 4 and 2 ounces per day, respectively), unlimited eggs, salad vegetables (2 cupfuls per day), and low carbohydrate vegetables (1 cupful per day). Subjects were strongly encouraged to drink at least six 8-ounce glasses of permitted fluids per day, and discouraged to drink caffeine and alcohol.
So if you can make sense of it, difficult I agree, that is what the 12 men in the study group were instructed to eat in the men’s hypogonadism. As for those in the control group, they were
instructed to continue eating normally but received guidance about healthy eating patterns.
The authors of the men’s study claimed the two diets contained an equal number of calories, but who knows?
How long was the study? Who knows. According to the authors it was conducted between March 2018 and October 2020. But that is the total length of the study as patients were added. So it’s really impossible to tell how long individual subjects stayed on the diet. A decent editor and/or peer reviewer would demand to see the actual data and require it to be listed in the paper.
But, despite the small sample size, there were significant differences in findings.
The researchers looked at three separate, but related, data points. They were looking to see if there was improvement in anthropomorphic measurements (waist circumference, hip circumference, weight, and BMI), increases in serum testosterone levels, and changes for the better in subjective questionnaires.
Let’s look at the last one of these first.
The questionnaires used were the IIEF-5, AMS and ADAM. These stand for the International Index of Erectile Function, the Aging Male Symptoms, and the Androgen Deficiency in Aging Males scales. These are all derived from the answers the subjects gave to the respective questionnaires.
As far as I’m concerned, these are all worthless. They are perfect examples of what statistician William Briggs calls trying to quantify the unquantifiable. Watch just a couple of minutes of the video below as Briggs explains it in his humorous way. Then I’ll show you the scales that were used to make all this sound scientific. I’ve got the video queued to the precise spot where he starts talking about this very issue.
See what I mean?
Now let’s look at a these questionnaires.
First, here’s the International Index of Erectile Function (IIEF) questionnaire.
This isn’t really the questionnaire, but it is a listing of how the people who put it together ranked the various questions in importance. The subjects being questioned gave numerical values for these different questions. It is absurd.
What I might rank high as intercourse frequency or satisfaction, someone else might rank as low. Same for desire level, relationship satisfaction, and all the rest. Other than the questions such as frequency, which have numerical values, the rest is trying to quantify the unquantifiable.
I particularly like the IIEF because the authors of the paper go to such length to burnish the scientific cred of what is a series of questions with totally subjective answers.
Relevant domains of sexual function across various cultures were identified via a literature search of existing questionnaires and interviews of male patients with erectile dysfunction and of their partners. An initial questionnaire was administered to patients with erectile dysfunction, with results reviewed by an international panel of experts. Following linguistic validation in 10 languages, the final 15-item questionnaire, the International Index of Erectile Function (IIEF), was examined for sensitivity, specificity, reliability (internal consistency and test-retest repeatability), and construct (concurrent, convergent, and discriminant) validity.
What a lot of horse shit.
How about the Aging Male Symptoms (AMS) scale? It’s a bunch of questions leading to this summary and ranking.
Here is a part of one of the multiple parts of the questionnaire.
As you can see, these are totally subjective. What you may rate as an increased need to sleep, I might rate as oversleeping. Same with every category.
Finally, the Androgen Deficiency in Aging Males (ADAM). Here again is another questionnaire. Some of the questions.
Once again, all are subjective. If you read the papers these questions come from (I’ve linked them all in their respective places above), you can see how the authors who wrote them are trying to give an air of real science to what are totally subjective responses. And, just like Briggs talks about in the video above, all of them have official sounding names and are doubtless copyrighted, and if so, those who want to use them must pay.
I always ignore these kinds of indices. Even the word index gives them a veneer of science that isn’t deserved.
So, even though the various indices above were improved in the study in question, I’m going to ignore them. Beyond, of course, telling you why you should always ignore them as well.
Let’s take a look at the anthropomorphic and testosterone changes the two groups experienced in this study.
As you can see from the above table, the group on the low-carb diet lost the most weight and had the greatest change in waist and hip sizes. There was also a pretty good bump in testosterone levels in those who were on the low-carb diet.
I like this study primarily because it trips my confirmation bias. It tells me what I want to hear.
And having had considerable experience caring for a lot of patients with different varieties of low-carb diets, I tend to believe the anthropometric figures. Had I been peer reviewing this paper, I would have wanted to see the individual data. There were only 18 subjects, so it would have been easy to include. What you worry about in small studies like this is that one or two subjects may have done really well and account for most of the weight loss and girth change in the entire group. Which gives an imprecise picture of what really happened.
Same with the testosterone figures. And, assuming they are all correct and not biased by one or two hyper-responders, we still don’t know if the change in levels came about as a direct effect of the lowered carbohydrate content of the diet or as a result of the weight loss and improvement of metabolic syndrome.
It would have been nice if one group had lost weight via a low-fat, high-carb, low-calorie diet. It can be done; it’s just not fun. If the other group had lost weight that way and still had their testosterone levels increase, we would know it was from the weight loss. But we don’t.
What we do know is that most people lose weight on some version of the low-carb diet, so either way, if you’re a male, you should see your testosterone go up if you follow a low-carb diet.
I came across a chart from an article in The Lancet titled “Excess mortality attributed to heat and cold: a health impact assessment study in 854 cities in Europe.”
In the ongoing debate about the validity of climate change, which used to be called global warming, one of the sticking points that comes up is that way more people die from the cold than do from the heat. We hear or read tales every hot summer about people perishing from the heat, but the truth is that way, way more die from cold in the winter.
Researchers wanted to quantify this difference in deaths, so they took a look at the statistics in 854 cities throughout Europe. Once they tabulated the numbers, it was apparent that many more people died from cold than from heat.
Here is the graph they published showing the difference. Deaths from cold are on the left in blue while those deaths from heat are shown on the right in orange.
If you look at this chart, you’ll easily be able to see that there are more people dying from cold than from heat. But it doesn’t seem to be that huge of a difference.
Until you look at the scale on the bottom. When you do, you see that it isn’t the same. The first notch on the scale going left is 50 deaths whereas the first one going to the right is only 10 deaths, so there’s a difference in a factor of 5.
Bjorn Lomborg, a climate realist noticed this too, and he redid the graphic using the same scale for both sides and posted it on Twitter…er X, I guess.
Here is the link to Bjorn’s tweet (if that’s what they’re still called):
Here they are side by side, so you can really see the difference.
Seeing the real chart makes you worry a little less about a little more heat. (Don’t get me wrong; high heat is deadly, too. Just not as deadly.)
Once I saw Bjorn’s tweet and had the chart as published in The Lancet in hand, I whipped that sucker up on Twitter (X?). I got the following reply.
As I replied to him, that could be a possibility; I kind of doubt it, but I don’t dismiss it out of hand. To be more transparent, they could have put up the graphics correctly (using the same scale along the x-axis) and then added an insert with the ‘blow up’ so you could discern the different colors of the various age categories. Or make a different graph. Scientists should be able to look at the scales and interpret it appropriately. But these kind of graphics are passed around all over the place, and most people just take them at face value. Which would be big time incorrect in this case.
Mendelian randomization, or MR as I’m going to be calling it from now on to save myself a lot of typing, is a difficult concept to grasp, mainly due to the heavy use of jargon in the field. Here is a sampling of the jargon, just so you’ll get the picture.
The best way to get a grasp on it is probably to start at the beginning and go from there.
What is the goal of science? Believe it or not, the Holy Grail of science is the proof of causality. Does A cause B? That’s pretty much it. Take a run through PubMed, pull down all the papers you want, and you’ll find that most of them will be trying to discover if A causes B.
Let’s look at an early example.
Pretend you’re an academic researcher back in the 1950s and you’re looking for a project. You are not a smoker, which makes you an oddity back then since the vast majority of adults smoked in the 1950s. You’ve noticed that several of your older colleagues have come down with or died of lung cancer over the past few years. All of them were heavy smokers. You reason that tobacco smoke coursing through the lungs might cause a problem. Most of the smokers you know have coughs and suffer more upper respiratory problems than you do. So, it’s a short jump to think smoking might be a driver of lung cancer. Especially since lung cancer was rare before everyone took up smoking, and now it is surging.
You decide to find 200 people who are older than 60, who are in apparent good health, and who smoke at least two packs a day and have done so for over 30 years. Once you’ve got your study population, you go out and round up 200 people who have never smoked, are the same age as the smokers, and are in good health.
Then you wait and watch. Over a period of time, you find that the smokers are developing lung cancer while the non-smokers are not. At the end of the study period, which you defined at the start, you find that substantially more people in the smoking group either have lung cancer or have already died from it than in the non-smoking group.
It seems pretty obvious that smoking causes lung cancer. But you can’t be completely sure.
A couple of reasons.
First, what you could be seeing is what’s called reverse causality. Cancer takes a long time to grow. It only seems to grow swiftly at the end stages when an already large tumor doubles in size. In the early days, the germ of the cancer might be just a few cells, too small to see with the naked eye. It takes a long period of cell replication to get the cancer to the size of the head of a pin. And a long, long time for it to get to the size of a pea. As it gets larger, it grows faster.
You can’t rule out the idea that the growing cancer itself might somehow be driving the person in whom the cancer resides to smoke. That’s reverse causality. Your starting hypothesis was that smoking causes lung cancer, but the opposite may be true. Maybe a growing cancer in the lung causes people to smoke. [I very much doubt that’s the case, but for the sake of argument, it’s possible.]
Based just on the data you have—more smokers get lung cancer than non-smokers—you can’t tell definitively whether you’ve determined causality or not.
There is another issue at play as well. There may be some unknown factor that both drives people to smoke and promotes lung cancer. So instead of A causing B, which appears to be the case given the data, maybe C is causing both A and B.
A classic case of a third factor is insulin. Type 2 diabetes used to develop in people after they had become obese. Which caused most everyone to believe that obesity caused type 2 diabetes. Once researchers started looking at insulin levels, it appeared the insulin resistance and hyperinsulinemia set in and caused people to gain excess weight and, ultimately, to develop type 2 diabetes. That idea is not accepted by everyone, but it is an example of a third factor.
So, what do you do now to confirm what you suspect? How can you eliminate reverse causality or a mysterious third factor?
You have to do a randomized, controlled trial (RCT).
The study on smoking and lung cancer that you’ve already done is called an epidemiological or observational study. These kinds of studies don’t really prove causality. An RCT comes much closer.
In an RCT you would recruit 400 people who have never smoked and randomize them into two similar groups: a study group and a control group. You would require those in the study group to start smoking at least two packs per day, while those in the control group would continue not to smoke. After a period of years—set at the start of the study—you would see how many people in the smoking group developed lung cancer as compared to those in the non-smoking group. If there were a substantial number more cases of lung cancer in the smoking group, you could reasonable say that smoking causes lung cancer.
There would be no reverse causality to contend with because all the people in both groups were non-smokers at the start. There were no growing cancers driving them to start smoking.
And you could pretty safely say there were no third factors causing both the smoking and cancer. Had there been, it would have as likely been present in the group who did not smoke at the same percentage as in those who were randomized to the smoking group.
The randomization would have pretty much eliminated those two confounders.
But such a study could never be done on ethical grounds. If your hypothesis is that you suspect something might cause harm—smoking, in this case—you can’t very well ask subjects who are not already doing it to start.
In general, observational studies can’t really show causality. You need RCTs to do that. At least to the extent to which such studies can be done.
Let’s look at a couple of real examples.
Someone made the observation that beta-carotene levels were low in some people who had lung cancer. Once that observation had been noted, a group of people who had lung cancer were compared to a number of people who did not have lung cancer, but were the same age, sex, weight, etc.
When the two groups were compared, those who did not smoke had higher blood levels of beta-carotene than those who did. Consequently, the hypothesis was floated that beta-carotene was protective against smoking. So, researchers put together an RCT to test the hypothesis.
They divided subjects with lung cancer into two groups. One group got beta-carotene and the other group got a placebo that looked like the beta-carotene pill. And the researchers waited to see the outcome.
In many studies like this one where what may be a life-sparing treatment is being tested, researchers who know the randomization codes take a look part way through the study to see if there is any difference in outcome. If it turned out, for example, that the subjects on beta-carotene were living longer and having their cancer grow less quickly than those in the control group, it would be unethical to deny the treatment to those taking the placebo. When this happens, researchers typically terminate the study and make the drug or supplement or treatment available to all in the study.
When they broke the codes on the beta-carotene study, they found just the opposite. Those subjects taking the beta-carotene were dying at a much higher rate than those not taking it. The outcome was the exact opposite of what the original observational study had shown. They terminated the study immediately and took all those in the study group off of beta-carotene.
Another famous study looked at the nutrient selenium and prostate cancer. Someone had made the observation that patients with prostate cancer had lower than normal selenium levels in their blood.
So, subjects with prostate cancer were compared in an observational study with a group of men of the same age, fitness, etc. without prostate cancer to see if there was a difference in selenium levels. And, sure enough, there was a significant difference in the two groups. Those with prostate cancer had lower selenium in their blood.
Researchers put together an RCT called SELECT to determine if taking selenium supplements would protect against prostate cancer. A $100M later, it turned out that the selenium supplements did nothing to prevent prostate cancer.
Here were two high-profiles cases in which observational studies strongly implied that increasing the blood levels of certain substances would be protective against specific cancers. RCTs showed that in one case it made no difference while in the other it made it worse.
It would be nice if there were a less expensive way to run studies on issues such as the selenium and beta-carotene trials.
That’s where Mendelian randomization (MR hereinafter) comes in.
Before we get started, the first thing I want you to be aware of is that the randomization part of MR has nothing to do with the randomization process in an RCT. I suspect those who came up with it wanted to borrow the precision implied in the word “randomization” as used in RCT. It is not that. MR is typically more like an observational study in which there are many confounders.
Here is how it works.
According to inheritance laws laid out by Gregor Mendel, the so-called father of modern genetics, all genes are inherited in random fashion from both parents. In other words, if one of your parents has a gene that causes A and a gene that causes B, it will be totally random whether you would inherit both of these genes. Which in the strictest sense is not exactly true, but it is close to true. This randomness is where MR gets the randomization part of its name. It has nothing to do with the randomization of an RCT.
We all have genes that code for proteins that end up performing some function. And we all have gene variants that are associated with specific traits. And we get these genes randomly from each of our parents.
Researchers can use these gene variants to test for certain outcomes much in the same way observational studies are done, but with more reliability.
For example, researchers combed the genomes of thousands of men and found variants associated with naturally higher selenium levels. Men with these variants didn’t take selenium supplements, yet had higher blood levels of selenium than average. Researchers then compared the rates of prostate cancer in these men.
When men with these variants were compared against a control group of men who didn’t have the variant, there was no difference in the incidence of prostate cancer. Which was precisely what was found in the SELECT trial.
The nice thing about MR as compared to a traditional observational study is that the tendency to have elevated selenium is coded in the DNA, so the possibility of reverse causation is eliminated along with most of the other potential confounders.
I’ve read about a number of studies done with MR and most of them confirm what has already been determined by RCTs. MR doesn’t have quite the reliability of true RCTs, but MR is much less expensive to do and does eliminate many of the problems with traditional observational studies. But MR is not perfect, as we’ll discuss in a bit.
One of the MR studies I read about was fascinating, but troubling from a societal point of view at the same time.
Some farm workers who have been exposed to organophosphates (insecticides) in sheep dip come down with a variety of health issues. Some, but not all. Which has led to those who do manifest illness being accused of malingering or secondary gain in terms of time off, disability payments, early retirement, etc.
This is obviously not a situation in which anyone would do an RCT. You would have to randomize people, all of whom had never been around sheep dip or organophosphates (potential health hazards) into two groups, then expose those in one group but not the other and see if there is a difference in outcome. Since the chemicals involved may be problematic health-wise, it would be unethical to do such a study.
And it would doubtless be tough to do an observational study by comparing sheep dippers with non-sheep dippers for all kinds of reasons.
But, as it turned out, the situation lent itself well to an MR study.
An enzyme that deactivates a potentially toxic component of sheep dip has variants that have different biological activity. Some detoxify strongly, others not so much. It was then predicted that those who had the variant that coded for (and thus produced) an enzyme that potently detoxified the toxic component in sheep dip would have few, if any, symptoms after being in contact with sheep dip, while those whose genes coded for an enzyme that was a weak detoxifier would be more likely to get sick.
When researchers looked at sheep dip workers, they discovered exactly that. Those who had the variant for the strong detoxification never got sick, while those who had the variant for weak detoxification were the ones who were sick all the time.
Which is a wonderful thing to know. One who owned a sheep dipping operation could do genetic analyses on all potential employees and use it to weed out those who would get sick on exposure to sheep dip. It would be a win for the employer and a win for the potential sheep dipper, who would avoid a lifetime of illness.
But there are greater societal issues at stake.
In my exhaustive reading about MR, I discovered there are arrays of gene variants that predict how well someone will do in school. As we all know, achievement in most anything is a combination of innate ability coupled with grit.
So, if we find a way to predetermine those who have great genetics for intelligence, do we then accept only them into medical school, graduate school, and law school? And in doing so eliminate someone with a ton of grit who would make it through based on hard work and probably out-doctor, out-lawyer, or out-research those who are much smarter.
Will we soon be using genetic analysis as part of the employment process? You can see how troubling this could all become pretty quickly. What if in the climate we’re in today these variants sorted themselves along racial lines? Or along gender lines? Could you even imagine?
The growth of MR as a tool to try to tease out causality is almost exponential. It is easy to use and inexpensive. And researchers are using it to generate papers much in the same way they used to use p-hacking. Consequently, you’ve got to be very careful in taking as valid anything you read about a study done by MR.
There are a number of ways MR can go wrong that probably aren’t worth going into here. But rest assured they exist and can lead to outcomes that are not correct.
George Davey Smith is one of the godfathers of MR, and he is highly skeptical of how the process is being used. Here is his summary in a recent paper of how MR should be interpreted.
The findings from Mendelian randomisation studies, which are less susceptible to confounding and reverse causality bias, sit at the interface between traditional observational epidemiology and interventional trials. A well conducted Mendelian randomisation study that reasonably satisfies the above assumptions often provides more reliable evidence than a conventional observational study. But the findings should be interpreted in the context of existing evidence from other sources, using different study designs, and clinical guidelines should not be rewritten solely on the basis of Mendelian randomisation results.
With that in mind, there have been a multitude of RTCs showing no real causality in terms of LDL levels and heart disease. The idea that LDL is related to heart disease is called the lipid hypothesis, and, as far as I know, it is still called the lipid hypothesis. Though many lipophobes and lipidologists want to pretend it’s the lipid fact. So far, they haven’t succeeded in moving it from the hypothesis column to the fact column.
Now with new tool of MR, they are at it. They are saying MR finally proves there is causality. Which puts MR into conflict with RCTs. So… Make of that what you will. I’m still skeptical. And, I’ll admit to not having read the MR papers on LDL and heart disease. But I will. And I will report when I do.
Okay, on to something more lighthearted.
Video of the Week
Here is a guy who is nothing short of phenomenal. He does all the parts—including percussion and instrumentation with his voice—of Queen’s Somebody to Love. Absolutely outstanding. Enjoy!
There are so many incredibly talented people out there who might never get a platform were it not for social media. If for no other reason, I’m thankful to Big Tech for that.
Okay, that’s about it for today. Keep in good cheer, and I’ll be back next Thursday.
Thanks for reading all the way to the end. This post is public, so feel free to share it as you like.