Arguing with Lasers

When I am thinking about various pseudo-scientific treatments it frequently takes the form of a dialogue, or to be more blunt, argument in my head. Yes I talk to myself.

So here’s a dialogue about Low Level Laser Therapy (LLLT).

But first, I work professionally as an independent physical therapist and strength coach. If you are interested in my services you can email me at: hirsch.sm@gmail.com. Or if you just want to support my blog you can buy me a KoFi.

Actual Low Level Laser Therapy application – credit

In brief, LLLT is a passive modality used in PT that involves the therapist moving a wand over the injured area and the wand shoots a laser at your skin.

Me: The thing that strikes me about LLLT in the peer-reviewed, published literature is the fact that the articles don’t even contain a plausible hypothesis.

Laser Enthusiast: Well, the laser delivers energy to the tissue.

Me: Yes, but so does a hot pack.

LE: A laser is different!

Me: Yes, and . . .

LE: It’s a different kind of energy than a hot pack.

Me: It’s photons. The same thing as a lightbulb. Why can’t I just use a lightbulb?

LE: It’s focused.

Me: A lightbulb can have mirrors in front of it so that it’s just as focused as a laser. And not all lasers are focused. A laser is defined by being EITHER collimated or monochromatic, it doesn’t have to be both and the most widespread kinds aren’t collimated, so they have focusing lenses too.

LE: But it’s also monochromatic, you just said that.

Me: Sure. It’s all a narrow range of frequencies of light. What does that do?

LE: Well, the tissue reacts in a particular way, or something . . .

Me: When I point out that the published research lacks a clear hypothesis this is specifically what I am talking about. The paper will say we used such and such frequency of light but won’t explain why.

This illustrates the importance of doing bench research before doing any animal or clinical trials.

The first laser studies should have involved pointing lasers at petri dishes of cells and seeing what happened. Does a particular frequency stimulate a particular molecule? Does stimulating that molecule mean anything relevant? Does a specific frequency stimulate a particular cell type to do something relevant?

And these studies needed to compare these measures to a good control condition like the same watt-minutes of a different frequency or of white light.

These are the types of questions that should have been answered first. These are the types of hypotheses that should have been tested in the research. But this is absent from the write-ups.

LE: So what if we don’t know the mechanism of action. There are plenty of approved drugs that we don’t know the mechanism for.

Me: Oh sure, that’s a valid point. But the standards for approving a drug require much more clear evidence of benefit than a device. I’ve talked about the low standard of evidence for medical devices before.

LE: But the studies do show positive results.

Me: But are they well designed studies? And do they show clinically relevant results?

What’s the control group in these RCT’s? Usually it’s the same laser wand without the laser turned on. While the control group may not realize they are in the control group because they feel nothing, they may suspect they are in the control group because they feel nothing. But the experimental group KNOWS they are the experimental group because they feel it get hot. So hot that it can cause burns if not used properly.

When both the control and experimental group have a pretty good idea which group they are in then the study isn’t particularly well blinded. This influences reported results. Enough to produce false positive results.

A good control group would be standard of care. In the same way that approval for a new pain-killer would require comparison not merely to a sugar pill but also to another standard pain-killer. If a drug company wants to sell a new pain-killer and it’s not better than Tylenol but it’s also more expensive and dangerous then it doesn’t matter if it’s better than a sugar pill, it’s not getting approved.

LE: Well, if patients think it’s working isn’t that good enough?

Me: No.

This is what’s important about the comparison to a hot pack. The hot pack can cause burns, it’s true, but preventing that is incredibly easy. Like, you let the undergrad clinic slaves apply hot packs (true story).

They also don’t require a physical therapist to sit there and apply the treatment. You just put the hot pack on and walk away. But the LLLT requires my doctorate educated time to sit there and constantly wiggle the wand to keep it from burning the patient. My time is expensive.

Plus, the hot pack is something the patient can do for themself. There are a wide variety of kinds they can purchase. There are long lasting ones using the same kinds of chemistry as a chemical hand warmer. There are ones that can be microwaved for instant relief. There are electric heating pads. And they are all cheaper than a PT clinic Co-Pay.

Hot packs are really convenient for the patient. They improve the patient’s sense of ability to control and manage their own condition. They improve their self-efficacy. These are all good things for the patient long-term.

While the laser (and any other passive treatment) creates the idea that the patient needs an appointment with the therapist to feel better. It may be profitable business but it’s not in the best interests of the patient or the cost of our healthcare system.

LE: But you haven’t really addressed whether or not the lasers work.

Me: How well they work needs to be put in perspective. Before we can conclude that the laser is a better treatment for the patient it has to demonstrate that it is SO MUCH better than a hot pack to justify the higher cost, higher risk and patient’s loss of control. That means big effect sizes.

Statistically significant effect sizes don’t even matter to me. As I discussed in the VR for Back Pain article what matters to a patient are clinically significant effect sizes. And these effect sizes would need to show a clinically significant difference to hot packs, not merely to an inert control.

Any researcher can get a statistically significant result from the “gee whiz!” factor from using lasers but that doesn’t prove the effects are real.

So, no, lasers are for tag, not physical therapy.

The good kind of arguing with lasers – credit

VR for Back Pain!

Isn’t the future amazing. Just think how excited Mark Zuckerberg was to see that the Metaverse could treat real ailments.

But let’s back up.

By ESA, CC BY-SA 3.0 igo, https://commons.wikimedia.org/w/index.php?curid=62840930

But before that

I work professionally as a Physical Therapist and Strength Coach. If you are interested in my services you can contact me at: hirsch.sm@gmail.com or thru Trainerize.

FDA Regulation

If I survey a bunch of folks I expect that most would agree that we want the FDA to regulate medical “stuff” for the following things:

Efficacy – does it actually treat the condition
Safety – does it hurt you
- And what side effects does it have
Content – does the product actually contain what it says
Purity – does it NOT contain other stuff

(Except for Libertarians they’d disagree with the whole premise because the “market” will sort it out better than the Government. But recorded history shows them to be factually incorrect.)

And what medical “stuff” do folks want regulated this way? Well, again, I suspect that most folks would want this list of stuff covered:

Pharmaceuticals i.e. what’s usually meant by drugs in this context
Homeopathic preparations*
Natural and herbal remedies*
Medical devices
Surgical procedures
Other medical treatments not covered by the above.

* Sellers of homeopathy and herbal remedies would claim that FDA regulation is unnecessary because they “know” it’s safe and effective etc.

So what on that list is regulated by the FDA the way most folks want it to be? Item #1 only. That’s it. Formally #2 is also but the FDA only recently started to and they still devote minimal resources to it.

Medical Devices

Leaving aside the lack of FDA regulation for safety and efficacy of surgery or the things that PT’s like myself do, let’s talk about medical devices.

The FDA does not “approve” medical devices. They authorize them. This is a key distinction in their lingo. Approve means that it’s tested for all the characteristics at top. Authorized means the device is safe and it may even work. They apply a lower standard of evidence for what works.

Drugs require multiple double-blind studies showing a clinically significant effect size. As well as a good balance of adverse events and side effects against the dangers of the condition being treated.

Devices need to be safe. So an e-stim or TENS device for relief of chronic low back pain must only shock you in the specific way that it is intended to do and not randomly shock you in a way that causes injury. They are regulated for content in the sense that the TENS unit has to actually deliver electricity. And for purity in the sense that it can’t be secretly a drug delivery method.

Now e-stim actually works for temporary relief of back pain and is appropriate in very limited cases. I’m not knocking on the authorization for it. But it is not approved and not subjected to the same level of testing as pharmaceuticals.

On to the Metaverse!

The FDA authorized marketing for a VR program to treat chronic low back pain. Note a key word in there: marketing. The FDA auth is for selling and advertising. It is not for FDA approval. While the company is likely to tout the “FDA AUTHORIZATION” all over it’s marketing, without breaking any rules, you can safely bet that third party promoters will muddy the waters by saying “approved”.

The FDA’s own description of the single study of this product’s efficacy leaves a lot to be desired. Such as evidence that it works.

They tout percent decreases in pain such as 30% and 50% and for how long those effects lasted. The numbers look impressive. But pain relief is not normally reported that way in the literature I’m familiar with.

Pain relief is normally reported the way the next paragraph reports it.

“To evaluate the effectiveness of EaseVRx, at the end of the eight-week program, participants were asked to rate the following outcomes on a 10-point scale, with 10 being the greatest value: pain intensity, pain interference on activity, pain interference on mood, pain interference on sleep and pain interference on stress. On average, participants experienced a decrease in pain intensity of 1.31 points over the eight weeks of treatment. Participants also reported a decrease in pain interference for all measured outcomes that ranged from .95 points to 1.27 points down from their respective scores at the start of treatment.”
https://www.fda.gov/news-events/press-announcements/fda-authorizes-marketing-virtual-reality-system-chronic-pain-reduction

But those results are not clinically significant. It is standard for me to ask patients to rate their pain on a scale from 0 to 10. It is normal for them to reply with two numbers, such as “1 or 2” or “5 or 6”. Humans don’t really differentiate pain on an 11 point scale. A five or six point scale is probably closer to our actual precision.

This is formally known as Minimum Detectable Change (MDC). Imagine that two businesses are selling sticks of various lengths. And one business claims that, on average, their sticks are 1/4” longer than the competitors. BUT it turns out they did their measurements with a yardstick that was only marked in 1 inch increments, with none of the smaller lines between inches. Would you still trust their claim to be 1/4” longer?

No, you wouldn’t trust the claim. Their instrument for measuring lacks the precision necessary to make that claim. Well, we have just that data for the Numeric Pain Rating Scale (NPRS) – and it’s the same for the Visual Analog Scale (VAS) (just multiply by 10). That data shows a MDC of about 1.5. So any result below 1.5 cannot be trusted as a real difference.

But I said clinically significant. This term refers to whether or not the difference matters to the patient and clinical outcomes. We measure this characteristic in part with Global Rating of Change (GRC). As well as other methodologies. And from this we generate a Minimum Clinically Important Difference (MCID). The MCID for that pain scale is about the same as the MDC, which isn’t surprising given the subjective aspect of pain. If the patient can tell that the difference is real then it matters to them.

Does this new device work?

Short answer: no.

Longer answer: not according to the evidence so far presented.

Why do I summarize this evidence as, “no, does not work”? Because the burden of proof is on them to show their device is efficacious. The correct assumption for any novel treatment is that it does not work.

Why does it seem like it worked in the study?

The control group was poorly designed. Or well designed if your objective was a false positive result. The study is described as double-blinded but the control group got a 2D experience with no therapeutic aspect while wearing a VR headset. They knew they were in the control group. No doubt about it.

The basic reason for the experimental group, with the 3D therapeutic experience, to report a small improvement is an Expectancy Effect. There are a lot of other factors that go into why treatments that don’t work will lead to positive outcomes in poorly designed study. Enough that it would be it’s own long, multi-part series of posts. (if you want that I’ll need a Patreon first).

Conclusion

I would love a more accessible approach to improving chronic low back pain. But this is not even that since a typical VR set up requires more space than most people’s homes.

It was an interesting experiment. But the real results were negative. Which is important to know. However, the product should not be authorized by the FDA but the FDA is following the laws that they have been given.

No, I don’t need to Just Try It

One of the common responses I get when I provide critical* feedback of a diet or exercise idea, is that I should, “just try it”. But no, I shouldn’t. And the suggestion that I should indicates a lack of understanding of science. Either the science underlying the specific topic or of scientific methodology in general and why it is the way it is.

Science underlying the topic

For the topics of both nutrition and exercise science we have 70+ years of published research. The research is not all confined to ivory tower laboratory experiments and deals with people eating in the real world and athletes competing in their sport.

We have enormous amounts of research to back our conclusions. Many landmark studies that exist have been running for decades.

The basic findings of these research programs have all been repeatedly replicated with enough variations to demonstrate broad generalizability. Outcomes are clearly predictable.

As such, no, some new (or old) exercise tool is not going to be some exception to these patterns.

And, no, some new diet is not going to show that actually, this, that or the other was actually the key all along.

Nutrition is NOT always changing

The obvious comeback to me pointing out that there is more than half a century of research backing nutrition science would be the claim that it’s always changing. I even had a nutrition professor say that during the unit on the history of nutrition science. But if you look at the basics and broad eating patterns instead of focusing on details, there has been little change.

Every US food guideline has said that Americans should eat more vegetables and fruits and fewer calories added with sugar and oil/fat.
Eating fewer calories than you burn is the key to weight loss – everything else is single digit percentage fiddling with details.

There are more consistencies than just this and there is more consistency than changes.

Scientific Methodology

My personal experience with an exercise or diet is basically irrelevant. When the published, replicated results and my personal experience disagree then the correct conclusion is that I am wrong.

Personal experience with a topic is not an experiment. You are not controlling for confounding factors in any way.

With dieting this usually just a matter of a person liking whatever diet was the one that they could best live with. When my friend uses a low carb diet with periodic fasting to get back down to a healthy weight then more power to them. But if they say I should do that they are going to run into the brick wall that I am miserable when fasting and really like carbs.

Or the person has bought into bogus health claims. Your diet is not making your blood acidic. So if the low-acid diet** gets you to eat more vegetables and less high calorie density foods then your improvements are from the fiber and better calorie balance. You didn’t control for confounding factors and you are attributing success to the wrong thing. This is exactly why understanding the basic science and doing controlled experiments are so necessary.

Ye Olde exercise tool is not great for everything. No tool is. But if adding that tool got you to actually do more than you were doing before you will see some improvements. It is basically true (for most things) that more exercise will produce some improvements even if they are modest or inefficient.

If your shoulders are getting tired from holding your sword up then sure, gada exercises will help with that, but that doesn’t prove they are good for anything else. Or that they are cost-effective. Or safe.

I will, however, reserve the right to get grumpy about specifically dangerous diets and exercises.

The history of this

The history of what was learned by humans who “just tried it” is well illustrated by the history of medicine. Because there are a staggering number of medical treatments documented in historical sources that just don’t work. But some humans tried them and became convinced that they worked. And then it ended up being written down by somebody called a doctor (or equivalent in their language).

In short, humans have hyper-active pattern recognition; it’s hard-wired into our brains. This is useful in many basic life situations but can also go badly wrong. We think that A caused B when they are in fact unrelated.

Good reasons to do things

There are of course plenty of good reasons to do something that don’t have to do with optimal outcomes. Where I get grumpy is when a person is really just describing their personal preference as if it were the best idea for everyone.

Exercise plans that fit your circumstances, equipment and motivation are great.

Eating patterns that are sustainable for you and improve nutrition are great.

But don’t insist that I need to just try it.

*critical here in a technical sense, not just destructive criticism

** the low-acid diet I’ve seen actually recommended eating oranges which are the most acidic food that humans eat 😀

Medical Marijuana and “What’s the Harm

Recently I posted this article on my Facebook: Marijuana Beliefs Outstrip Evidence. I summarized a key point thus, “The only use for which marijuana has reasonably good evidence of effectiveness is for nausea.”

Note, I am all for removing marijuana from schedule 1 of FDA regulation and allowing for comprehensive research of the ways in which marijuana derived substances can be used medically. However I am firmly of the belief that they should go through the exact same FDA process as all other drugs.

The responses I got were predictable. One of the key points made by many was that if it helps people then what’s harm? I am going to set aside the point that it’s not helping most users and focus on the old “what’s the harm?” gambit. It is a standard part of the responses to criticism of alternative medicine.

So, What is the Harm?

If a treatment is just a placebo, and does not provide actual benefits then these are the harms that may result:

Money. These are hardly ever free. And especially in a country where medical expenses are the most common cause of personal bankruptcy the cost of a treatment that doesn’t work is harm.
Time. The time spent going and getting a treatment done, transportation etc. are all costs for the patients and these are things that have a tangible monetary value and therefore refer back to the above point. Is the patient giving up time at work spending, money on transportation and so on? If so that is a very real harm.
Delayed treatment. If a person delays getting effective treatment because they mistakenly believe that this placebo is actually treating their condition then that is an unambiguous harm which has demonstrable negative effects in a wide variety of diseases.
Side effects. Marijuana very clearly has side effects if we are viewing it as a medicinal substance. It is used specifically for its mind-altering effects. Remember the point of this drug originally was that it got you high, that you are stoned while you’re taking it. This is incompatible with most jobs. And even when it’s not it’s still constitutes a set of risks. FDA-approved drugs go through a risk assessment process comparing the benefits to the risks. And as detailed in the linked article above there are very clearly side effects from marijuana use.
Purity and content testing. FDA approved drugs are rigorously tested for their purity, that is their absence of toxic substances and adulteration. As well as that the medication actually contains what it says on the package. Marijuana for medical purposes lacks all of these protections.
Addiction potential. While marijuana does not have the same addictive properties that opioids do it still possesses a risk of becoming habit forming, as can any behavior, especially when it is a mind-altering one such as marijuana use.

This list is not intended to be comprehensive. But does at least hit the highlights of the potential harms of medical marijuana.

An additional range of harms from the current approach to medical marijuana also includes the fact that it increases the danger to children for accidental ingestion and poisoning, especially with the prevalence of edibles and related products that are in similar packaging and have similar names to candy, as detailed here: More Marijuana More Problems for Young Children.

Conclusion

Ultimately, my position on this topic is that all medical treatments should be subject to a uniform study and approval process. The FDA is not perfect but they have a very good track record. The objective should be to lobby for marijuana to be removed from Schedule I, not to create a carve out from regulation where marijuana gets a free pass written by legislators.

“The Dirty Dozen” – Brief Response to News

(We’ll see if this becomes a regular thing for this blog)

The Environmental Working Group released its annual Dirty Dozen list again. They are a non-scientific group of scaremongers. This list of theirs epitomizes this fact. They take the publicly available USDA data on pesticide residue and then report characteristics like the number of different residues found and the amount of those residues. Then they rank grocery store produce based on these numbers.

What they never do is compare those values to the actual amounts necessary to harm a person. Because if they did, they’d look silly. For instance you’d have to eat 10 kilograms a day of strawberries just to reach the low end of minimum appreciable risk (RfD). (I got the math from here.)

The techniques used to detect pesticide residue are incredibly sensitive. Far more sensitive than necessary for human health. The tiniest amounts can be detected. But those tiny amounts need to be compared to results from toxicology studies to have any meaning.

The Environmental Working Group also asserts, without evidence, that a wider variety of pesticides on a plant is somehow intrinsically more dangerous than just one pesticide. This statement might be true – different chemicals may interact to be more dangerous than toxicology research on a single chemical indicates. But no evidence suggests that it is. And human health is carefully tracked by the CDC.

And even if more chemicals on the plant is more dangerous they fail to take into account the fact that the plant already has thousands of different chemicals inside. At at much higher concentrations than the pesticides. And those plant chemicals include substances evolved to kill other life forms to protect the plant i.e. pesticides. There is more of those pesticides than the ones applied by farmers.

In short, their list has nothing to do with known toxicology information. It is just an attempt to scare people. And CNN and other media swallowed it whole.