JOSH: Hello and welcome to Clearer Thinking with Spencer Greenberg, the podcast about ideas that matter. I'm Josh Castle, the producer of the podcast, and I'm so glad you've joined us today. In this episode, Spencer speaks with Matthew Smout about categories and modalities, Cognitive Behavioral Therapy (CBT), and CBT in relation to other techniques.

SPENCER: Matt, welcome.

MATTHEW: Hi, Spencer, and thanks for having me.

SPENCER: Many people consider Cognitive Behavioral Therapy, also known as CBT, to be the most evidence-based treatment for many different mental health challenges, including depression and anxiety. So let's start there. What is Cognitive Behavioral Therapy?

MATTHEW: Cognitive Behavioral Therapy is one of those talk therapies, psychological therapies, which focuses on helping people identify the way that they perceive and construe situations and the way that they behave that might be contributing to problems in their life.

SPENCER: Now, CBT is composed of two different parts. There's the cognitive part, the C, and the behavioral part, the B. So what is the cognitive part of Cognitive Behavioral Therapy?

MATTHEW: Cognitive refers to anything to do with thinking, so it refers to the way that we pay attention, the way that we remember, the way that we reason. But in Cognitive Behavioral Therapy, particularly, it refers to beliefs we have about ourselves and the world, and that sort of influences what we pay attention to and remember, and also contributes to the kinds of thoughts that go through our mind. The C really refers to how our thinking affects our mood and how we then act in response to situations.

SPENCER: So getting concrete, what are some cognitive techniques that you might do if you went and got Cognitive Behavioral Therapy?

MATTHEW: Yeah, great question. So if you were anxious, for example, there's a really good chance that you are seeing something as threatening. Perhaps you think that people are talking about you at work and judging you badly, which might mean you feel anxious and awkward around other people or don't want to go into those social situations. So, a concrete thing would be for a therapist to help you monitor and write down some thoughts that go through your mind when you are about to enter a social situation or after a social situation, if you're thinking back over how you performed, and sort of look for thoughts that might have gone through your mind that might explain why you felt so anxious in that situation. Like, "I think they're judging me. I think they think I'm stupid. I think they think I'm ugly. I think they think I'm incompetent," those sorts of things.

SPENCER: Okay, so let's say you wrote down those thoughts. What would the therapist have you do with those thoughts?

MATTHEW: From there, I guess the classic sorts of things might be to look at the evidence for and against those thoughts to see whether we're misperceiving the situation or overvaluing some kinds of information and undervaluing others. Often, people who are socially anxious, for example, will take the fact that they're anxious as a sign that they're not performing well, they're not coming across well, or they'll have an image of themselves in their head where they look embarrassed or imagine themselves to look very weak or ineffectual, and they use that information coming from inside themselves already as evidence for their conclusions about themselves. A cognitive therapist might help them shift their attention to what was the evidence outside of their thoughts and feelings. What were other people saying or doing that might give you feedback about how you're actually coming across? The idea is, if we can get people to shift their basis for the evidence they're using, they can start to form healthier thoughts about how they performed. Like, "Well, nobody was really paying attention to me. They were worried about their own performance," or "Actually, people were smiling and nodding at me when I was talking, which probably suggests they thought I was at least as friendly and effective as anybody else in the room," and that kind of thing.

SPENCER: Now, what are negative core beliefs in cognitive therapy, and how does that relate to what you've been discussing?

MATTHEW: Negative core beliefs are beliefs we hold about ourselves usually that are very overgeneralized. The idea is that basically, "I'm incompetent all the time, or I'm unlovable or I'm worthless" — very general statements about negative value — those take some time to shift because they're not specific to a situation where looking at the data in a particular situation would change our minds about ourselves. They're informed by lots of memories and experiences throughout our lives that we hold on to, which make it feel as though these are universal truths about ourselves.

SPENCER: A critique that cognitive therapy sometimes gets, especially from people who are new to it, is they think, "Well, is it just telling me that it's all my fault that I feel bad? I mean, I actually have problems. It's not just my thinking that's the matter. There are objective things in the world that are going badly."

MATTHEW: That's a great point to raise, and it's probably a bit of a caricature of cognitive therapy that we only look at faults in people's thinking. I guess I've started there because that's probably what cognitive therapy is best known for, and perhaps what's a little bit distinct about it compared to other approaches. But in practice, a good cognitive therapist isn't likely to impose or insinuate that you are inherently misreading situations. They should work with you to sort of figure out together what the truth is. We call that Guided Discovery, where it should be an exploration between the two of you to see, "Does the evidence really stack up? Do the things that you're thinking are not at all true? Are they totally true, or are they a little bit true?" That's part of what you go through in cognitive therapy; teasing apart the quality of information and maybe modifying your beliefs. It's not always the case that you're just going to go from having a negative thought about yourself to a completely positive thought about yourself, but looking for a more nuanced view of yourself or the situations can help you feel better about yourself.

SPENCER: One thing that I've observed just in my own mind, and I think it's broadly part of cognitive therapy, is that when you're feeling intense negative emotions, it tends to exaggerate or distort your thinking in certain predictable ways. So maybe there is something objectively going on that's really bad in your life, but the fact that you're feeling anxious, the fact that you're feeling sad, the fact that you're feeling depressed, might make you see it as even worse than it is, or see it in an unhelpful light.

MATTHEW: Yeah, absolutely, that's a great point. And you're right. When you have bad things going on in your life, if you've lost somebody really important to you, or people at work are really giving you a hard time or too much work to do, it's natural in those situations to feel deep sadness, or in a work situation, it's appropriate, perhaps, to feel angry and irritated at times. But as you say, you can certainly take normal disappointment, sadness, loss, and anger and turn it into something much more corrosive by catastrophizing, awfulizing, overgeneralizing, imagining this is going to go on forever when it might be a moment in your life. So, yeah, absolutely.

SPENCER: Now, cognitive therapy has many different techniques. We're just scratching the surface, but I don't want to go too deep into it because we have a lot to talk about. So let's jump to the behavioral part of Cognitive Behavioral Therapy. What is that about?

MATTHEW: Well, it is pretty straightforward, I suppose. It's looking at the things that you do and trying to increase behaviors that are good for you and reduce behaviors that are bad for you, which is a very simple way of putting it. But part of a lot of Cognitive Behavioral Therapy is aimed at trying to identify ways that people cope that may be making things worse for themselves in some way. So I'll go back to my social anxiety example. If you're feeling very anxious in situations and you try to cope with that by slinking off into the corner and avoiding eye contact, and not saying anything so that you don't say something stupid, or over-rehearsing what you're going to say to try and get it exactly right, those behaviors, although they might make you feel a little bit reassured in the short term, might actually end up backfiring. They might mean that you come across as more awkward and standoffish and unreceptive to social contact than you mean to be. Similarly, a lot of people will unwind at the end of the day with a glass of wine, or they might engage in gambling or porn, or all kinds of coping behaviors that might make them feel better in the short term but end up having long-term costs. They spend too much money on it. They spend too much time on it. It starts to interfere with their performance at work the next day or their relationships. The goal is really to try and reduce or get people to do competing behaviors that are healthier. Go do some exercise. Do some meditation, do relaxation exercises, do some planning, get organized. Substituting healthier behaviors for unhealthy behaviors, if you like.

SPENCER: Now, of course, it's true that many people know they could be healthier in different ways, and they know they could replace certain bad behaviors with good behaviors, at least theoretically. But it's challenging to do these things. So what are some of the methods that behavioral therapists, or behavioral therapy as part of Cognitive Behavioral Therapy, use to get people to actually do these things?

MATTHEW: Often, people also have beliefs about whether they can or can't do these things, or should or shouldn't do these healthy behaviors. So that's where the C and B sort of combine a bit. If I know that exercise rationally would be a good thing for me to do, but I have a bunch of beliefs about, "Well, if I exercise, it's going to be too hard, I'm not going to feel good for a long time, or I'm not going to remember to get up in time to exercise," or whatever goes through their mind. So, when it comes to this, this is where the C and B combine. We might spend a lot of time trying to help people identify those task-interfering thoughts and come up with other things they could say to themselves at those moments to guide themselves through and do that. But there might also be little strategies, like breaking tasks down into smaller steps. A bit of a cliched example, but if your goal was to go for a walk before work in the morning, having your shoes right by your bedside or your clothes to change into, so that it's a small step to get into them when you get up in the morning, rather than a bigger step. If your goal was to walk for half an hour a day three times a week, you might start with just getting your shoes on and going down to the end of the driveway to get the mail and those kinds of things.

SPENCER: So it sounds like one technique is reducing the difficulty of a task by doing preparation, lowering the friction. I suppose there are other things you can do, like try to increase the rewardingness of a task. If you know, "Well, okay, maybe you don't like that type of exercise, but you could swap it for another type of exercise that you enjoy more."

MATTHEW: Yeah. And making tasks more pleasant by combining them with other things. If you've got a pile of dishes to do, combine that with listening to a great podcast like this, and make it a more enjoyable task. Or, if you've got a couple of tasks to do, use a more enjoyable task as a reward for doing the first task. Strategies like that are things a lot of people will have ideas about or have heard, maybe on social media from friends. But the therapist is trying to work with that individual to come up with a game plan of a set of steps that they can follow that feels most likely for that person to succeed.

SPENCER: So one way to conceptualize the behavioral piece of Cognitive Behavioral Therapy is trying to get you to do more good behaviors rather than bad behaviors. What makes a behavior good?

MATTHEW: Well, good behavior is one that I think is going to lead to long-term satisfaction if you keep it up. If you're someone who's exercising regularly on balance, you're going to feel healthier in the long run if it doesn't have any adverse side effects. The behaviors that we worry about are things like drinking, gambling, and self-harm, which are likely to damage your body in some way. There's a cost to pay for the temporary relief they give you, whereas healthier behaviors aren't likely to have those sorts of side effects. There's really no adverse effects to sitting quietly, watching your breath until it starts to naturally slow down to a soothing rhythm. There's probably no drawbacks to spending 10 minutes looking at the set of things you think you should be doing for the week, putting them in an order, and scheduling them so that you know they'll get done to relieve your stress. Healthy behaviors are things that are sustainable, that are going to lead to your long-term satisfaction and not have adverse side effects for you.

SPENCER: To some extent, the cognitive piece and the behavioral piece kind of seem stuck together. How did this come to be that they got attached to each other? Do you think conceptually it really makes sense, or is it just that these are two useful sets of techniques, so you might as well apply them together?

MATTHEW: That's a really interesting point. I suppose historically, behavior therapy kind of came first and came out of basic research into Pavlovian conditioning, operant conditioning, and instrumental conditioning. That sounds very jargony, but most people have an understanding of reacting to cues. If you think of Pavlov's dog, no doubt. The human equivalent of that might be walking past the bakery and noticing your tummy start to rumble when you smell the yeast, or feeling hungry at certain times of the day when you can see the clock or the sun going down, and your body responding to those time cues. Those principles involve rewards and punishments, and most people know about that kind of stuff. Early behavior therapy was really applying those principles to psychological problems, and it moved progress forward quite a lot. We were able to solve a lot of problems that weren't recovering very quickly under traditional psychodynamic approaches. The cognitive therapists grew out of the recognition that not everyone is going to be able to follow a set of behavioral instructions and get relief from that. There are lots of reasons why people don't engage in healthy behavior, can't see the point of it, or are afraid of it. Cognitive therapy arose out of the understanding that behavior therapy wasn't sophisticated enough to account for the full range of emotional problems that people have or the nuance in people's thinking. In some ways, I think it is a natural evolution in that journey. There have been periods where scientists fought each other about that, wanting to keep them as very separate approaches and not have them contaminated. Empirically, it turned out that it didn't make much difference whether you focused heavily on the cognitive or the behavioral; you tended to get similar results, but you might get a different response from the people you work with. People might respond a lot better when you took their thought life seriously rather than ignoring that and just focusing on what they did.

SPENCER: My understanding is that some of the hardcore behaviorists didn't even want to talk about thoughts or beliefs at all. They basically wanted to treat humans as black boxes; if it doesn't manifest, the behavior doesn't exist.

MATTHEW: Exactly. Depending on what you're interested in, there are interesting scientific reasons for that, whether thoughts are really just epiphenomena of environmental behavioral associations. But I think most people experience themselves as having some agency over what they think, and when they manipulate their thinking, that can change how they feel and change what they plan to do. So I would suggest it's more useful to think of yourself and your thoughts as being important and as a lever you can pull, rather than just an epiphenomenon, something that follows along with environmental behavior contingencies.

SPENCER: There's also this idea in Cognitive Behavioral Therapy, which connects the cognitions to the behaviors. This idea is that your thoughts influence how you behave, your behaviors influence how you feel, and your feelings influence how you think. All pairs of these thoughts, behaviors, and feelings influence each other, and they end up being inseparable because of that. For example, if you're feeling anxious, you might engage in a behavior where you avoid the thing that's making you anxious. So it's changing the feeling that changes your behavior. Or, if you have the thought, "Nobody here is going to like me," you might leave the party early. So that's the thought influencing behavior and so on. Every pair is influencing every other pair.

MATTHEW: Yeah, that's right, and that's probably the most useful way to think of things most of the time. I suppose it's probably the case that we have more control over overt behavior than we do over our emotions. Our emotions are probably the dependent variable in this network, where it's very hard to just decide to feel differently. It's very hard to trigger an emotional change without doing something differently, whether that be changing the actual things you're doing or changing what you're thinking about. But otherwise, you're quite right that all of these components tend to influence each other. Probably that sort of fight that goes on between therapy developers is whether there is an optimum kind of set of actions to take or sequence of tasks to do that can get the change you want quicker than others. But I think the evidence converges on this being a network, as you're describing, and so most of the time entering that network from one node or the other doesn't really matter too much.

SPENCER: Or it might depend on the individual person where you get low-hanging fruit. For example, I think about people who literally believe that flying is dangerous; they actually think the plane might crash, versus people who know that it's safe. They're fully convinced that it's safer, at least for the distance traveled, than driving, and yet they have this visceral terror. It seems to me that the first person, you kind of have to work on their beliefs before you're going to make any progress, whereas the second person, maybe you just go right to behavioral techniques, because they're already fully convinced. It's just that now their subconscious is still kind of rejecting the safety idea.

MATTHEW: And that's a great point. You're exactly right. That's probably the most likely reason we have this proliferation of different strategies, which is because these make sense differently to different people with different strengths and weaknesses in their understanding. I had a case, for example, of a guy that was afraid of lifts.

SPENCER: Sorry, just to translate for Americans, lift, It is elevator, I assume?

MATTHEW: Yeah, elevator. And I think his understanding of how lifts work was probably based on one of those Die Hard films where there would be an elevator suspended by one fraying cord. And so he was terrified to get into them. When we looked at a diagram of how modern elevators are made, each with five steel cables, each of which is strong enough to support the weight of the lift on its own, that was enough to help him be willing to get on with the exposure. So you're absolutely right.

SPENCER: So, we're going to get into the evidence for different therapies, what do we know about how well they work, and what do we know about what happens when you try to compare them, but before we do, I want to have a little bit of a meta discussion about how we know how well therapy works, and what is good evidence and what is not so good evidence?

MATTHEW: Yeah, look, that's a great question, and I'd probably be interested in your own views, maybe from a consumer perspective, about what kind of evidence is convincing, really. But what I would say is, I guess, with psychological therapies, there are different comparison groups that we can put them up against. We can compare them to, say, a waitlist, where people don't get any treatment while the main therapies are being delivered in the other arm of the trial until the end of the trial. So people are waiting. That's kind of the most common comparison that's done in psychotherapy research, and it's considered the weakest. People actually probably do a bit worse when they're on a waitlist than if they weren't going to get treatment at all; it might even inhibit their natural recovery. So that's a factor in weighing up the evidence. Other comparisons, which might be more convincing, are what we call usual care, or treatment as usual, where you might go into a hospital or community mental health service setting and whatever's going on, whatever people would normally get, people are either randomized to continue to get that, or they get the new treatment to be tested. That's got a lot of advantages to it, but it's also very variable, because the usual care will differ. Wherever you are in the world, there'll be a different kind of usual care. Sometimes the usual care is state of the art and very close to the kind of therapy you're testing, and sometimes it's more like the waitlist control. I guess what people think are probably the strongest comparisons are against other well-developed therapies. That's the most expensive kind of study to do and requires the most cooperation, so there are the fewest of those. You have this difficulty, I suppose, in deciding what bar you want to set. How do you decide what counts as good enough evidence to say something is supported? But I'd be interested in what you think about that issue too.

SPENCER: I don't think I'm the most typical consumer, because I spend a lot of time running studies. I'm opinionated on this stuff, but I do want to point out a few different challenges or issues that come up when running these studies. One is randomization. If a study is studying a therapy and they haven't randomized people to whether they get the treatment, or they're in the waitlist group or in the alternative treatment group, it's basically not reliable. Would you agree with that?

MATTHEW: Yes and no, I agree that randomization is preferable and the state of the art. However, there are some problems, or some groups of people whose difficulties are so rare or complex that there's a practicality about getting people to give informed consent and be randomized. Some of the things that are hardest to treat might involve people that are not in the best position to give informed consent. I wouldn't want us to not be able to learn things from working with them, but given that people can give informed consent, then yeah, I agree. Randomization is the first criterion.

SPENCER: Yeah, fair. And obviously, if you can't randomize, you can't randomize, or if it's unethical to randomize, you can't do it. But surely it gives better evidence. And just to clarify for the listener, why is that important? From my point of view, it's because if you don't randomize, then there could end up being a correlation between who ends up in the different treatments and different facts about what happens to the person. For example, let's suppose you just looked at the people that the therapist decided to use CBT on versus the people the therapist decided to use a different technique on. They may make that decision based on the difficulty of the case, and then suddenly you can't tell if people are doing better because of the CBT or they're doing better because there are different types of people getting the CBT, or easier cases getting CBT, etc. So I just want to clarify, with the randomization, it actually begins to distinguish these cases. And then you can say, "Did this really cause a difference in effect?"

MATTHEW: Yeah, absolutely. Another way of looking at that too is there's an infinite number of ways in which people can differ. When we randomize, we just know the probability that the groups might end up differing by chance. When you don't randomize, you just don't know how likely it is that the differences are due to third variables or chance.

[promo]

SPENCER: Another thing that I think is worth keeping in mind is that while it might seem a really good idea to pit one technique against another state-of-the-art technique, it does present a problem, which is that you need a much bigger sample size to do the study. Because if you're testing an effective treatment against, let's say, a placebo or against a waitlist, then to show that it works, all you have to do is beat something that doesn't really make much difference. Whereas, if you're testing it against an already effective technique, the difference, even if it works and even if it's better, the gap is going to be much, much smaller, which means you may need a much bigger sample with many more people and much more cost. And if you don't invest to get enough people, you might decide that the two therapies are equal, but just because you don't have enough power in your study to tell the difference.

MATTHEW: Yeah, that's right, and I guess there's two ways to look at that. On the one hand, it casts doubts about claims on the existing data about equivalence or superiority. The other way to look at it, I suppose, is clinically, if a difference requires thousands of people to show it, how practically significant is that effect going to be? And sometimes you can get practically significant effects when they're very small. But it depends a bit on what difference you're testing, doesn't it?

SPENCER: Yeah, it's a fair point. If you take thousands of people, maybe it's probably not a very big effect if you need that many people to test it. But on the other hand, if it worked 10% better, that could be a big difference, and that actually might require a pretty big sample size to tell the difference there.

MATTHEW: Yeah, for sure.

SPENCER: The last issue I want to bring up, before we get into the evidence itself, is that any poor implementation may simply not work because it was implemented badly. If you're using a very effective therapy, but you have very poorly trained therapists, then it's not going to work very well. And there's also an issue of dosage and things like that. If you have one study that's giving people one day of Cognitive Behavioral Therapy treatment, obviously, that's not going to show a very large effect compared to one that's three months or six months of treatment. So I think we have to take these failures into account really carefully and say, "What does this failure really mean? Was it really a competent execution? Was the dosage really enough that we'd expect it to work, and so on?"

MATTHEW: Yeah, I think these are so important. I'm glad you're talking about them. One of my concerns about just how many meta-analyses are out there, and how many of these are done by people who perhaps haven't done primary trials themselves, is how well they appreciate these kinds of issues. Because you can get effects sometimes with very brief interventions. But it depends a bit on who you're working with and what your outcome measure is. Those sorts of things, something that might work for somebody we might call rapid responders. There are some people that are just very responsive to interventions because they've probably already got lots of healthy coping strategies in their armament, or a past history of being a reasonably successful person, but they've gotten into crisis. They can use the information they get in a brief consult to kind of remobilize really quickly. If you have a study that's full of those people, you can show effects with small interventions, but those same things aren't likely to work with somebody who's got a 10-year history of being in and out of a mental health hospital. So I'm glad you're talking about this sort of stuff, and you're right. The competence thing does make a big difference. It's tricky because we've got some evidence that suggests competence doesn't always matter, but that's based on poor evidence. I guess we should get into that more in the next discussion.

SPENCER: Well, that's interesting in its own right, but I think we can all agree a sufficiently incompetent implementation will always fail no matter how good the treatment is. But we also don't want that to be a universal way out to say, "Well, it failed, but they probably were incompetent." We have to take it seriously when something fails as well.

MATTHEW: That's right. And again, you can look at it that maybe one of the qualities in a good therapy is that it's relatively idiot-proof versus one that requires such high training, sophistication, and supervision that it's very easy to get a poor outcome with it if you don't do it properly.

SPENCER: It's a good point, because if you're a consumer of the therapy, if it's really hard to find a competent person, how are you really going to find them anyway?

MATTHEW: Well, that's the thing. I think it's a real concern in some ways. If you look at something like behavioral activation, which involves making sure people set themselves small, achievable goals that will lift their mood or give them a sense of achievement. You can still do that poorly, but it's probably not as difficult to learn to do a good job of that as it might be to administer psychodynamic therapy.

SPENCER: Yeah, you also mentioned meta-analyses, and I'm pretty opinionated about them as well, and I'm curious to hear if you agree with me. Often I'll see meta-analyses that combine things that are very different from each other. They'll look at 50 studies on Cognitive Behavioral Therapy for depression. I don't mean to single that out as one example, but they'll do things like that. Then they'll average them all together by converting them into Cohen's d, which is a standardized measure of the size of the effect. Then they'll say, "Look, this is how well Cognitive Behavioral Therapy works for depression." My view is that that's problematic for two reasons. The first is that you really want to get rid of the bad studies. If you average together a bunch of good studies with a bunch of bad studies, you're kind of just screwing up your research. You need to carefully assess the quality of each and discard the bad ones. The second thing is, you need to be very careful about combining things that are fundamentally different. Soon as a meta-analysis will do tests for heterogeneity, basically asking, "Is there evidence that these studies find different sized effects from each other?" That's a good thing, but often I would rather see it broken out. If these studies are really different in nature, like one is a three-week intensive and the others are a six-month program, I would want to see them broken out and not just averaged together.

MATTHEW: Yeah, I wholeheartedly agree with that. I guess the better ones probably do both, in the sense that they do an overall analysis where they put everything together, and they try to do either a subset analysis, where they just have the ones that score really high on their quality index. But as you say, the quality of the meta-analysis all comes down to how the studies are classified, and that's where you need people who really know their stuff doing them, because meta-analyses are so easy and convenient for people to do at lower levels of university training, very convenient student projects. Often, you do get a lot of meta-analyses produced by people who don't understand the subtleties of the studies. They don't tend to spend much time writing about them in the articles.

SPENCER: So now that we've talked about how to think about the evidence, let's actually jump into the evidence itself. It's sometimes said that Cognitive Behavioral Therapy is, quote, the gold standard treatment when it comes to mental health disorders, and that it has far more evidence supporting it than any other therapy. Would you agree with that?

MATTHEW: Yeah, I would say in the sense that there have been the most studies done on Cognitive Behavioral Therapy. I would agree with that. I haven't counted them all, but I would imagine, from what I can tell from my literature searches, that there are close to 3,000 Cognitive Behavioral Therapy studies, and that far exceeds any other kind of school of psychotherapy.

SPENCER: I think there are meta-analyses of meta-analyses, if I'm not mistaken.

MATTHEW: Yeah, there are meta-analyses for literally dozens of problems. And as you say, meta-analyses, meta-regressions. There's a large number of studies; whether that means that that's not quite the same thing as a large amount of evidence saying that Cognitive Behavioral Therapy is superior to another therapy.

SPENCER: Well, where's a question of, "Does it have the most evidence that it is effective?" Which is different than is there evidence that's the most effective. That makes sense, and obviously we don't just count the number of studies. If there were tons of studies, but a bunch of them showed it didn't work, that would be evidence against it being effective. So let's first take that question: is Cognitive Behavioral Therapy effective? Does it work to treat things like depression and anxiety? And can we really be confident of that?

MATTHEW: Yeah, I think the answer to that is yes, it's better than most alternatives. If most alternatives include unstructured treatment, as we talked about before, usual care, if it means a waitlist or no treatment, in some cases medication, or at least placebo pills, it becomes harder to say that if you put it up against another system of therapy that has been as well developed, well supervised, and well implemented. Then it's not really clear that Cognitive Behavioral Therapy is always better than other approaches. I might argue that in certain isolated cases, it seems to be particularly effective in certain anxiety disorders, PTSD, OCD, and maybe even eating disorders. But overall, most of the comparisons between Cognitive Behavioral Therapy and other psychotherapies, when they're matched for all of the quality components of the study, tend to end in a stalemate.

SPENCER: So that's really interesting. A side note on that: if you look at really high-quality studies of Cognitive Behavioral Therapy, where it's pitted against a strong placebo, like either a placebo pill or some kind of care that is not expected to change people's mental health, but people will feel good about, does it pretty squarely win in nearly every single study like that?

MATTHEW: Yeah, I would say so. Cognitive behavioral therapy has been applied to so many problems. There are some areas, particularly in behavioral medicine with certain medical conditions, where it's perhaps very early days. The contribution of psychological factors to physical factors in people's mental health is probably still being established. I wouldn't be as confident in some of those areas, but for the vast majority of areas, it's been researched, and I'd say that it's consistently better than anything that's not as structured and well trained and implemented.

SPENCER: Okay. And now, when we talk about other therapies that could be pitted against it where it may not show that it's better, what are we really talking about? What are some of the competitors?

MATTHEW: When psychodynamic approaches are equally well manualized, supervised, and implemented, there are some trials where there hasn't been a significant difference. The other thing I might just quickly say, Spencer, about Cognitive Behavioral Therapy is that it is a very big umbrella term that probably encapsulates a whole lot of different protocols. If you were to watch a therapist working from their protocols, they could be doing quite different activities following those protocols. But even though they're both called CBT, there's a lot of uncertainty around that as well. Putting that aside, one of the main competitors, which isn't a psychotherapy, is pharmacotherapy, and sometimes cognitive therapy can be superior to that. You might consider that to be a legitimate, non-placebo kind of treatment. But in terms of psychotherapy, what are the alternatives? Acceptance and Commitment Therapy, which, again, depending on how you want to draw the divisions between psychotherapies, you could consider that to be a CBT or an alternative to CBT. There have been some trials that have found them to be similarly effective for things like anxiety disorders or chronic pain.

SPENCER: That sort of goes by the acronym ACT, right?

MATTHEW: That's right, yeah. I'm trying to think of a good psychodynamic ACT. There might be meta-cognitive therapy. Again, is that not a cognitive therapy? I think it is a cognitive therapy or belongs in the CBT branch, but you could distinguish.

SPENCER: It's a subset. It has a distinct flavor to it, right?

MATTHEW: That's right, and that's the thing. So if we're going to call CBT, I guess what CBT looked like up to the end of the 80s, let's say then, yeah, metacognitive therapy is an extension and enhancement on that. In one or two trials, it's been shown to increase the amount of gains that people get compared to CBT. But yeah, there are not as many sort of head-to-head trials, I guess in eating disorders, like you might have, family therapy would be a distinct protocol.

SPENCER: What about interpersonal therapy?

MATTHEW: Of course, sorry. I always forget about that one. I feel that probably just my slight discomfort with interpersonal therapy because it was originally a placebo condition designed in the NIMH trial, and now it's being disseminated like a bona fide hour. That's happened a couple of times. In psychosis research, they deliberately developed a placebo condition called befriending therapy, and then that ended up having no difference between that and the active condition. So now the training, befriending therapy, that sort of stuff just kind of irks me a little bit. So that's why I probably forget about it. But no, you're right. Interpersonal therapy would be a good example of something that's been run head-to-head with CBT, and often with equivalent to no superiority type results for the power of the study. EMDR would be another good example that's quite different to CBT in terms of its in-session activities.

SPENCER: So that's Eye Movement Desensitization and Reprocessing, which is treatment for panic disorder. Is that right?

MATTHEW: Post-Traumatic Stress Disorder, mainly.

SPENCER: And then another one is DBT, or Dialectical Behavioral Therapy, which many people think of as a form of CBT. Would you consider that just part of CBT or a different thing?

MATTHEW: Well, this is the dilemma. It's certainly a recognized brand, so it is researched on its own. It is definitely a distinct protocol. But when it was developed, it really harvested everything that was known about CBT at the time and integrated it with some new contributions. What makes it distinct are very clear guidelines for the sequence in which to do various activities and the structure, but the actual activities themselves really were mostly harvested from existing CBT at the time. So again, it's one of those things where it's kind of like, yeah, it's its own thing. That's definitely its distinct thing. But is it radically different from CBT? I don't know.

SPENCER: The last kind of category I think about is mindfulness-based, kind of third wave therapy. What would you say about those?

MATTHEW: Yeah, so again, I think they are reasonably worth distinguishing from CBT in the sense that, particularly something like mindfulness-based cognitive therapy, which involves quite large amounts, is always delivered in groups and always involves lots of meditative exercises. It's quite a different use of time. So it makes sense to see that as somewhat distinct. And again, it was developed to add to CBT rather than compete with it originally. So it was there to sort of solve the problem of how do we stop people relapsing after they've successfully recovered from CBT, rather than, instead of giving people CBT, let's give them MBCT.

SPENCER: Before we go more into the evidence about comparing CBT to these other therapies, I do want to deep dive on psychodynamic for a second. Could you tell us a little bit about the origin of that and how it evolved?

MATTHEW: I'll be honest, this is not really my area, so I have a very sketchy understanding. But obviously most people have heard of Freud and Jung and the development of the psychoanalytic tradition, and there were lots of authors that put their own spin on psychoanalysis. Psychodynamic therapies started to become manualized and abbreviated following the success of CBT in getting funding for trials in the 80s, and so gradually, over the last 20 or 30 years, there have been more attempts to package psychodynamic therapy in a more structured and time-limited form. But I'm hedging around the fact that, "Do I know exactly what goes on in those sessions?" I'll be honest, I don't really.

SPENCER: And so psychodynamic really has evolved quite a bit from Freud. It started with Freud, but there have been a lot of add-ons and adjustments. They stopped believing a number of Freud's claims.

MATTHEW: I would say so. Again, I'm not really an expert on the content, but what I'm more familiar with is the changes they've made to the format, where analysis was very resistant to empirical research to start with and very skeptical about abbreviating itself. That was a big change in thinking amongst that group, to start to package it more and deliver it in randomized controlled trials, the same way that CBT was doing.

SPENCER: Funnily enough, I got in a debate with a client-based psychodynamic therapist last night at a bar.

MATHEW: Really?

SPENCER: Yeah [laughs]. It was really a weird coincidence. I guess Klein was one of the successors to Freud and took his theories in certain directions. Our debate was kind of entertaining. She claimed it was impossible for an adult to get a trauma, except in a special case. She said that it's possible if they had a childhood fantasy that they didn't really want to come true, and then it did come true, like they secretly wanted to have sex with their mother when they were a child, but then they actually had sex with their mother as an adult. That could traumatize them, but otherwise it's impossible for an adult to become traumatized.

MATTHEW: Really? So she worked with many military personnel? Was war not traumatizing?

SPENCER: I guess she would say, by definition, it's not, but I don't know. I can't steel man her arguments. To be honest, I would really struggle to do that. But I do just want to say something else about psychodynamic therapy, which is that, as it's kind of become more structured over time, in my view, at least — I don't know whether secondary people would necessarily agree — but it's taken on some elements of CBT. Many elements it doesn't take on, but some, I think it does take on, taking on the elements of having a specific focus and specific goals for the therapy. In doing so, it's become much easier to study it, and then you can kind of put it alongside these other therapies, whereas if you take the traditional psychodynamic approach, it might be hard to even compare it because it's structured so differently, and it doesn't have a clear set of goals in the same way and a clear sort of endpoint.

MATTHEW: I agree, and I think that just that aspect of having a clear focus, a treatment goal or target itself across psychotherapy schools is associated with better outcomes. The more generic and heterogeneous the ability for therapists and clients to sort of use that time, the less improvement people are likely to make.

SPENCER: And I think the type of psychodynamic that I've seen most studied is short-term psychodynamic therapy, or STPP. Is that what you've seen as well?

MATTHEW: Yeah.

SPENCER: All right, so let's get into the evidence. So when they pit CBT against these other therapies, what do they find, and what do you find most convincing there?

MATTHEW: I guess most of the time, particularly if you look at meta-analyses, where you combine a number of trials, and for all the disparaging comments I made about meta-analyses earlier, if we restrict it to just good quality trials, and those that are head-to-head trials, usually better quality trials. This is an area where you do want to draw your conclusion from across a number of trials, rather than just a single trial or one or two trials. But yeah, when you meta-analyze CBT versus active comparisons, most of the time there's no significant difference. I do think that you could make a case that, say in social anxiety, for example, and this is a particular protocol, I think within CBT, and they would call themselves cognitive therapy, even though it still involves operating on thinking and behavior. I think that Clark and Wells' approach to social anxiety treatment is especially effective. I think if you look at PTSD research, there are clearly some CBTs that work better than others, although there's no evidence that CBT works better than EMDR consistently that I can find. But if they're trauma-focused, CBTs generally do pretty well. I'm not sure that there's a psychodynamic treatment that's worked as well as those. I'm not sure that there's a psychodynamic therapy that's as effective as CBT or family therapy for eating disorders, although I know that that's definitely a focus of development. My knowledge might be out of date there, but I know in the past there have been comparisons that haven't worked out so well. I think the area of personality disorders too, we don't have many head-to-head trials, but there's probably stronger evidence for maybe DBT. I would like to say schema therapy. I've got a very soft spot for schema therapy, but if I'm being unbiased, there's probably still work in progress to be clear about whether schema therapy is better than CBT yet for that.

SPENCER: Sounds like they're mostly finding no difference when you pit CBT against these other therapies. Occasionally, you'll find cases where maybe there is no other therapy that really offers a solution for the issue, or some cases where maybe on that specific problem, CBT has an edge. But are there therapies where CBT just beats the hell out of them? Or where it's clear that therapy doesn't work well?

MATTHEW: If you don't include any form of exposure — meaning facing your fears in the treatment of anxiety disorders — then they're probably not going to do very well. I think that's an area where CBT, or any therapy that involves exposure, is going to do better than one that discourages exposure. I'm trying to think of any other examples where CBT is vastly superior.

SPENCER: You might think there are all kinds of crackpot therapies that people have developed that just do nothing. Why don't we see a bunch of studies showing that these other therapies just get crushed by CBT? Is it because nobody ever bothers to even test them?

MATTHEW: That's a great point. I say this to students: when I was starting out my career, we would read about all kinds of stuff that just never got tested. All of our essays in undergrad would conclude that more evidence is needed. We should do some studies on this. That's not the case with CBT, which really led the way in testing psychotherapy and producing study after study. Why we don't have lots of crackpot therapies is because CBT now really is something to beat. If you want to read about crackpot therapies, Jeffrey Masson's Against Therapy, if you ever read that one?

SPENCER: No.

MATTHEW: It covers egregious things that were done in the name of psychotherapy, like things that were born out of the '60s, probably psychedelic workshops, with all kinds of stuff that was downright abusive in some ways. But we've kind of weeded out all of that stuff. The availability of things like CBT, which are much more trustworthy and work better, I think has led to it not being attractive to develop these kinds of unfounded alternatives.

SPENCER: It's funny you say that because if you go on a website like Psychology Today, where you can search thousands of therapists, you look at what they say they do, they often will check 10 or 15 boxes about all the therapies they practice. They're kind of all over the place. They're credibly eclectic, and even if they say they do CBT, if you actually go to them, and I've seen this happen, it's clear they're not doing CBT; they're doing something vaguely CBT inspired, but it's just something else. It's maybe pulling in some techniques from CBT. So are you saying that people aren't doing all kinds of wacky stuff, or are you saying just in these big randomized control trials, people aren't doing wacky stuff?

MATTHEW: Yeah, well, I can't be sure what people are doing behind closed doors or in conferences that I would never go to. So that's my disclaimer. But certainly no one's doing randomized control trials on implausible therapies, I would say. Let's take NLP for example, right, which was really popularly disseminated through self-help and management types.

SPENCER: Neuro Linguistic Programming, is that what you referred to?

MATTHEW: Yeah, that's right. NLP was very popular in the 80s, and lots of book sales in mainstream bookshops, but there were only a couple of attempts to experimentally test their hypotheses, and they basically never got confirmed, really, and so it never rose to a point where they would do a proper NLP RCT, so it just really died out.

SPENCER: EMDR support is still popular in self-help workshops, right?

MATTHEW: It might be. I don't see much of that these days. The one that raises my eyebrows a little bit is internal family systems therapy, because as far as I know, there are no randomized controlled trials for that. Even though it's been a long time since it was originally developed and started being trained. It might be good stuff, I don't know, but I'm surprised that something that's being so heavily promoted in workshops hasn't had any empirical evaluation or not serious empirical evaluation.

SPENCER: I think it'd be great to have people evaluate it. I think of internal family systems, which basically has you imagine different parts of yourself with different needs, and you kind of communicate with these parts. I think of it as just a method for exploring your internal feelings, and it's a method that anecdotally a lot of people seem to find useful as a self-exploration tool. Where it makes me uncomfortable is that some people seem to take it so seriously, where they treat it as though they really have these different parts of themselves, like on a more literal basis, which seems false to me. But I think if you keep it as a metaphor, and you kind of use it as a way to explore your feelings, I think it's a way that resonates with a lot of people.

MATTHEW: Yeah, and I'm not inherently opposed to it. I consider myself to be a schema therapist of sorts, and part work is a big part of that as well. But schema therapy researchers have put a lot of energy and time into testing out their ideas, so I'm not sure why IFS hasn't.

SPENCER: Yeah, and honestly, a lot of it might come down to the mentality of the developers. If they think it's very important to test things empirically, then it's probably going to get tested. If that's just not their orientation or focus, then it's much less likely to. I would say that isn't it almost always the case that the first studies are done by the originators or people closely connected to them, rather than some random people?

MATTHEW: Makes sense. But again, where I suspect you might share this too, I think most of my colleagues and I would feel uneasy is when the dissemination of an approach sort of outpaces the research of it. It should be the other way around; we should research more, and then as we get more confident that it works, then we disseminate it more. But that doesn't always happen in our field.

SPENCER: Yeah. I mean, it would be really nice if the more evidence-based stuff rose to the top. Of course, I think it also depends on what you think about the risk-reward profile and how easy it is to tell if something works. If you believe that a lot of stuff is very low risk and someone could tell if it works, then that sort of calls for a more experimental approach, like, try the thing, see if it works for you; if it doesn't, discard it. Whereas, if you think that there's more potential for harm, or you think people can't really tell what works, like, they could do a thing for a long time and think it works when it doesn't, then you want to be more cautious and say, really, only do the stuff that's evidence-based, because you might hurt yourself otherwise, or you might just go down a blind alleyway for a really long time thinking it's helping when it's not.

MATTHEW: Yeah, I think that logic is really sound. It's in practice of, "How well can we make those discriminations? How easily can we make that call?" Particularly when you go into therapies that we might say require a longer period of time to know whether they're going to pay off. So it's always a bit of a leap of faith when you back something like a meditation approach, because people really need to put in that four to eight weeks of practice before we expect them to get the full effects. Whereas there are other things where we can demonstrate in five minutes the difference between how a behavioral strategy might have different effects on a person, or how where they pay attention, or what they've believed to be true, can make a big difference to their mood. It's less of a risk when you can demonstrate an acute effect of the recommendation you're making. So, yeah, it's a really interesting point.

SPENCER: As a kind of little experiment, I was curious what people show on their profiles on Psychology Today. So I just clicked on the first profile I saw to see what treatment approaches this person says they use. I searched my zip code, and they list, I think, about 20 different techniques that they say they use. Here's the list: attachment-based, culturally sensitive, dialectical behavioral therapy, experiential therapy, expressive arts, Gestalt, the Gottman method, integrative, internal family systems, psychodrama, sensory motor, psychodynamic, relational, somatic, and trauma-focused. That's one person on the site. I think this is really typical actually; I've seen lots of people like that. I want to emphasize to people that when you go to an actual therapist, it's very important to figure out what they're actually doing, if anything. Do they have a set of techniques, or are they just kind of building something totally from scratch? Do you really know what you're getting? I think that's a very important set of due diligence to do.

MATTHEW: Yeah, great recommendation. I imagine many therapists will cringe when they see that and maybe see it as a bit of an overcompensatory marketing strategy. But if I were being more generous about it, it's possible these people are working with populations for which there isn't a well-trodden path to recovery, or because not everything is as amenable to any form of intervention as some things are. If you've got anxiety, depression, eating disorders, or trauma, I think your prospects are quite good with psychotherapy. If you've got chronic pain, schizophrenia, or bipolar disorder, it's quite a lot less certain. Or, in the emerging neurodivergent populations, we don't really know what works best for some of these people long-term as they transition to adulthood. If I were being generous, I'd say these people might have a lot of different ideas about things to try in cases where there isn't a clear evidence-based path, and that might be good for some people. But whether those therapists are using their collection of workshop-induced ideas to that effect, I guess we don't know.

SPENCER: I can certainly see the view that says, "Are more tools better? Would you rather have someone fixing your home who has a toolbox full of many tools than one who just has three tools?" That's maybe fairly persuasive. But on the other hand, I tend to view that if you have a specific problem, you should focus on the techniques that are most proven for that problem, ideally. You want someone who is very good at applying that set of tools, not someone who is dabbling in too many different tools, where they're unlikely to have much specific knowledge of how to use the tools really well. That being said, I think some therapists do benefit from pulling in lots of tools from different sets of techniques. So there's surely some trade-off there.

MATTHEW: Yeah, I agree, and I think probably the experience of the therapist, in terms of how many years they've been doing it, might be a factor here too. I have a set of perpetual frustration as a trainer of students still in their degree that qualifies them to become therapists, who get exposed to a different class on a different therapy every week. I see them in the clinic, and they come in saying, "How can I use this new thing I've learned about on the client I've been working with already?" It's the same thing every year. They come in each week with a different set of ideas to deviate from a plan we had already come up with, and that bothers me. People go to a hardware store and say, "Oh, I've got a new chainsaw now. What can I cut?" That's kind of the first way you want people to think. You want them to look at the job and ask, "What does the job require?" But a lot of therapists, it's hard. If you go to workshops, pay a lot of money, and get really excited about a new idea, it's hard for people to sit on those and just wait for the right time to come.

SPENCER: I think also it depends on the skill level, the difference of the different tools being used, and some of the tools actually require quite a lot of skill before you become competent at them. And so that's, I think, where it gets more worrisome as well, whereas other tools are relatively easy to administer.

MATTHEW: I think one of the things that you might sympathize with is that cognitive behavior therapy now is sort of almost considered an establishment type approach, where people are expected that by the time they graduate, they'll know everything they need to know about cognitive therapy, whereas I actually think it takes quite a lot of practice and skill to get good at that. I think people get seduced by the shinier things that are more often offered in workshops than cognitive behavior therapy, and so people are seeking out more exotic things that they haven't developed competence in CBT yet, and going to other stuff.

[promo]

SPENCER: So we think about these results, where these different therapies, in many cases, will test the same as cognitive behavior therapy when pitted head to head. There are a few different theories that you could have about that. One is that CBT is actually better in many ways. It's just that you need a super high-powered study to tell the difference, like maybe it's only 10% better, and unless you have a large enough study, it's just going to look the same. So that would be theory one. Theory two would be a kind of survival of the fittest theory, where it's like, "Look, by the time you're getting to testing something against CBT in a randomized control trial, it's probably a really good therapy." CBT is only being tested against things that have also been honed and are really very good. It's not that everything works; it's just that everything that kind of gets to that point of iteration works well. But there aren't that many things that ever got to that level. We were able to list a handful of them, but not dozens. The third hypothesis would be that all of this is kind of misleading. Really, all you need for therapy is a few common components, and as long as you have those, the therapy will "work," and those common components are things that almost every therapy we listed actually uses. An example would be someone who's very empathetic, who's listening to you carefully, who's asking open-ended questions, who's treating you with unconditional positive regard, who's meeting with you weekly, et cetera. It's not really about the technique or the theory; it's really about these elements of therapy, and this is what produces positive change. So on those three theories, CBT is actually better, but it's just hard to tell, or survival of the fittest, all these work well, or third, that it's really the core components that matter, not the technique or theory. Where would you say you fall?

MATTHEW: Probably three, with maybe a bit of two as well. I think the other thing that's worth mentioning is that a lot of the RCTs that go head to head are focused on a specific problem, which makes it more likely that even just the conceptualization of the problem and what people need to change to solve the problem is more likely that approaches are going to converge. If you've got an effective treatment for panic disorder, irrespective of your brand of therapy you subscribe to, if what really makes a difference is how people perceive and react to their internal sensations when they get them, whatever language you use to describe meaning-making or approach and attention or evaluation, or the strategies that people use to cope, even if you don't use CBT terminology, my guess is that if you study this problem and you're following the data, your approach is going to end up converging regardless of which school you started from. Partly, I think the difficulty in telling these things apart is that by the time you strip away all the language differences and the stuff that doesn't really matter, and you get down to what do clients actually do differently, there isn't as much variation in what's going to be helpful for clients to do. That's my firm belief. In terms of attrition, you're right, CBT is only going to be pitted against in head to head. Well, I guess the other thing is that anything that's not established will end up becoming rebranded in meta-analysis as attention placebos or psychological placebos, and they'll be easily defeated. The only thing that counts as a bona fide comparison is going to be a similarly rigorously developed, implemented, and supervised approach. So that's where I think I probably favor those two.

SPENCER: Got it. So in terms of the core components, this is sometimes related to what they call the dodo bird hypothesis, that all the therapies kind of work equally well. What do you see as those factors that are actually driving good outcomes?

MATTHEW: I'd say the therapeutic relationship is, of course, necessary. People are not going to listen to someone they don't trust or don't think is credible. The working alliance is powerful; if you think about the way it is conceptualized as agreement on tasks and goals as well as bond, I think the agreement on the tasks and goals is the more critical part, and the part that therapists can have more say in. This means that whatever you do, you need to explain the client's problems in a way that makes sense to the client. It turns out there are a lot of ways that you can do that. This is why you can have a lot of different theoretical approaches that can still have an effect. If it's true, as we said earlier, that all our experience is sort of like a network, we influence the thoughts that influence emotions with behaviors that influence emotions, influence emotions, influence behavior, and so on. You can have a lot of different rationales that can explain a critical portion of those interrelations so that a client will go, "Yeah, that makes sense. That makes sense why I should do things differently." We know pretty much you need a good rationale; everyone agrees on that. This is where my idea of convergence comes in. Whichever brand of therapy you start from, if you're following the literature about the psychopathology, that's going to guide you to formulate something that's going to be a pretty good match for reality. Whichever angle you approach the network from, you can come up with something plausible that works. Everyone pretty much agrees that avoidance of a problem either makes it not get better or worse. So pretty much anything effective involves stopping avoidance of a problem that you're dealing with. It's pretty basic, but that's how most people cope when they don't go to therapy. I think you need to change your meaning. However, whatever language your psychotherapy uses, lasting change doesn't seem to happen unless you change your meaning of the evaluations you make of yourself and the evaluations you make of threatening situations, depending on what problem we're talking about. If it's eating disorders, we're talking about the importance of body weight and shape or the importance of comparing yourself to others. Whichever parameters you're working on, meaning change has to happen. There are probably principles, and there are people out there who've tried to come up with taxonomies of this, where you'll find across therapies that people are still targeting these quite decent meta-level, trans-theoretical, transdiagnostic type parameters.

SPENCER: Well, take something like depression. You might think that cognitive therapy, which is just intervening at the level of thoughts people are having, would be very fundamentally different from someone who's using a pure behavioral approach that is just trying to get people to do pleasurable activities to treat their depression. You might think that would be very different from, let's say, a psychodynamic approach, which might be going into your childhood and trying to understand what childhood experiences might have led to how you're feeling right now. What converges about those three approaches fundamentally?

MATTHEW: I agree that even though the activities you might be doing in the session might look different, if you start to do things differently, you're going to start to have different qualities of thought. If you go out for that walk and start noticing the beauty of nature, your attention gets off your thoughts, and your rumination is going to go down. That's the behavioral activation changing cognitions kind of path. If you're starting from cognitive therapy in the room and looking at things differently, you might be more likely to go on that walk. At the end of the day, you'll be both doing more walking and doing less ruminating. Those are the things that make the difference because our understanding of what drives depression is complex. There are lots of roots of depression, but we know that rumination and inactivity are two very robust ones, and both will target that. The psychodynamic approach, again, would be operating on the meaning level. If you are exploring your past, and if that's going to be helpful, it's going to be helpful because it changes your belief about who you are and what you're capable of, which might mean, if I keep my example going, it will go from being ineffective, "I can't do my exercise. I'm not worth taking care of my body," to "I am worth taking care of myself, and I can do these things."

SPENCER: That suggests that one might imagine many different therapies that wouldn't work. Because I think what you're suggesting is that the therapy still has to have some way of getting to that result, whether it's changing the thoughts, changing the behavior, or changing the feelings. You've got to get there. But there may be many things that don't change any of those variables.

MATTHEW: Yeah that's true, and I think that the most common one is supportive counseling where somebody might go in. Now supportive counseling can be careful, because counseling actually can work very well for depression, but it depends. It depends on what you talk about. So I hear a lot from clients who come to me dissatisfied with their previous therapist. Their therapist, in the name of building rapport, would just ask them lots of stuff about what they're interested in, their stories, or what they've been doing, or over-disclosing their own experiences. Ironically, even some of that could still wake up, still remind people of things that they had forgotten, that they liked doing, or that give them ideas for better ways to spend their time than ruminating. So you could still get that outcome even from that. So I'm trying to think of an example where it would be really unlikely to get that effect. You just have to be talking about something really.

SPENCER: And I know people who have been to the therapist for, let's say, sexual assault, and the therapist told them it was their fault for the way they dressed. I mean, that's clearly not gonna work.

MATTHEW: So depression is a bad example if you talk about trauma, yeah, absolutely. Well, anything that involves not dealing with the trauma at all and, as you say, something that would make it worse, which is taking blame. In that case, that's a perfect example. We know that blame is a major driver of some of the constellation of trauma symptoms. To actually increase their self-blame is to take them in the opposite direction. So, yeah, that's a great example.

SPENCER: Sometimes people who talk about common factors in therapy focus more on the relationship with a therapist. We touched on that briefly before, but it sounds like you're placing the effectiveness a little bit less on that and a little bit more on what the therapist is doing for the patient, or where they're helping them get.

MATTHEW: I would say, "Yeah, I agree." I would express it like people overrate how important it is for their clients to like them at the start, before they do anything, and underrate how important it is that they check that the client understands and agrees with their emerging understanding of their problems. So I think a lot of where I see ineffective routine practices is therapists spending lots of sessions going so light and so slowly in the name of not wanting to upset clients and wanting clients to like them, and they're overlooking that the client is still suffering with whatever they came in with. They are not actually doing anything about it and aren't necessarily sure what the therapist is making of them or where this is going. I think the robust literature shows it's probably an interaction. You need to have a certain level of rapport; otherwise, the client is not going to listen to you, but clients are going to like you a whole lot better if they can see that you're helping them.

SPENCER: Yeah, I think it's a really good point. If the client really dislikes you, it's probably just not going to work. It's not going to function. But if they like you enough to trust what you're saying, and then they start feeling better, yeah, that's probably going to raise the amount they like you and the amount of trust they have. So these are definitely dynamic variables. My view is that even the act of meeting someone once a week who is going to listen unconditionally and be kind and empathetic and help you think about what you're trying to change in your life, like that right there probably has some therapeutic benefit, whether that's done by a friend or a therapist, and everything else we're adding on top of that, like the actual techniques, is just building on that base, but you probably already get a benefit just from that. Would you agree with that?

MATTHEW: I certainly think you can. Yes, absolutely. The question is, what is your presenting problem? It's particularly likely, I think, what you're describing to be beneficial if the person has a healthy repertoire already of success in their life, the ability to cope, so that you're sort of reactivating their healthier mode of living, rather than clients who have been through really impoverished histories and maybe just don't have a lot of skills in their repertoire. So I think it does depend a little bit on who you're working with, but you're right; everyone has a need to relate, to be related to, to matter to people, and to care about them.

SPENCER: It also gives you time to reflect, which I think on the margin, people tend to reflect too little on their life and what they're trying to do and what they could do better.

MATTHEW: That's true, and I don't know what your experience is. You might move in really good social circles, but my impression is that a lot of people find it hard to find good listeners in their social worlds who will ask them questions and patiently listen to their answers without changing the conversation back to them. In that sense, I think therapy is a real treat for people.

SPENCER: I would also add that I really think it does matter what technique you use. I do think there are many techniques that don't do anything above and beyond what we're describing because they are out of sync with reality; they don't change the mind. In other words, they don't cause you to think differently. They don't cause you to feel differently. They don't cause you to behave differently. They couldn't possibly add any additional benefit, or they might even make it worse. As an example, the therapy where they treat people who are grieving right after, let's say, a family member died, there is actually evidence that it makes people worse off.

MATTHEW: Yeah, I think that may be the case. I'm less aware of that. I'm more aware of people using therapy techniques that have the potential to be helpful in the right context, but then sort of apply them mindlessly to whatever client is in front of them. Mindfulness is a great one; a lot of people will take their clients through mindfulness exercises, perhaps without much thought as to whether this is something that is going to help the client. They haven't really established whether the client has problems with sitting still and paying attention to calming themselves and focusing on a focal point or redirecting their attention. They might use mindfulness without really knowing why they want that to be used, other than they just like it and think it's healthy for everyone to do. I see more of that, but it's possible there are people out there doing things that are really callously mismatched for what could be helpful for clients.

SPENCER: Would you agree that not all techniques are going to get a person to change? There is something different about techniques that cause people to change. There might be many techniques that cause positive change, but technique does matter to some degree.

MATTHEW: I do agree with that. My own clinical experience is quite persuasive. There are some things that are really robust and consistently help people improve, and other things that are extremely hit and miss. They might help sometimes, but a lot of times fall flat.

SPENCER: Before we wrap up, I'm curious to hear your thoughts on comparing therapies, whether it's Cognitive Behavioral Therapy or other evidence-based therapies, relative to antidepressants, which are often used as an alternative.

MATTHEW: I think the literature is pretty consistent showing that, with the exception of things like maybe bipolar disorder and schizophrenia, in most cases, I don't think I've seen convincing evidence that antidepressants are better than CBT for virtually anything else. In which case, usually, CBT has longer lasting effects at follow-up than the medications if they're withdrawn, which arguably could be due to iatrogenic effects of withdrawing from antidepressants. But I think it's more just that cognitive therapy does aid lasting change.

SPENCER: You mean, if you stop doing CBT, you stop doing the medication, and you follow up a certain amount of time later, the CBT seems longer lasting. That's what you're getting at.

MATTHEW: Exactly.

SPENCER: Someone might argue, "Well, okay, but they could just stay on the antidepressant, right? If it's relatively cheap and it's just a pill, it's easy to take." What do you think about that?

MATTHEW: It's a very personal choice, and I'm sure there are some of your listeners who are possibly on these sorts of medications. I would not disparage anything that people have found helpful. But I do think the necessity of that is being vastly oversold. I think we're only now, in the last five years or so, really starting to see the convincing evidence of just how there is a cost to that. A lot of people do have side effects on these medications that really do compromise their quality of life and find the process of getting off them quite difficult. So I think you want to be really sure that the difficulties you're having haven't responded to psychotherapy before you make that decision.

SPENCER: When someone has a very severe mental illness, let's say extreme depression, where they can't function at all. They can't work. They are currently just at home in bed all day. I generally would say their priorities should be to both get Cognitive Behavioral Therapy or a similar effective treatment and also probably consider medication at the same time, because the priority is getting them feeling better as fast as possible. What do you think about that?

MATTHEW: I would support that too. Again, I'd probably go by their capacity. Not everyone has the capacity to do learning therapy or talking therapy when they're that depressed. If they can't concentrate or understand or remember anything, a conversation-based approach to improvement isn't really going to work. In those cases, biological therapies are really the only option or last resort. But I would say, even if they could still have a conversation and concentrate enough and had enough ability to follow instructions and try things, even if they were just tiny steps, and even if they were quite debilitated, sometimes increasing the dose, as you mentioned earlier, seeing people twice a week can be worth a trial before going to medication in some of those cases.

SPENCER: Is that mainly just to avoid the side effects? Because the main reason that I would recommend that is, first of all, I think it could be additive. The antidepressants work on a very different mechanism. And second, I think it can actually boost therapy in a way where, if they're starting to feel a bit better from the antidepressants, it can make it easier to do the kind of behavior changes and cognition changes, I suspect.

MATTHEW: Yeah, I think that definitely can happen. I definitely experienced that too. The trouble we have at the moment is we can't really predict for whom those different courses are going to occur. And then sometimes, what happens in these situations is that people don't respond necessarily much quicker than the antidepressants, or they have side effects, or they have a difficulty with needing to chop and change them. So, yeah, it just depends, I guess. And sometimes the only way we can tell is just to try it out. So definitely open to that. But again, I kind of think if the clients and the service have the time, it's worth trying, but that's usually the constraint. A lot of people, most services I know of, struggle to provide weekly, let alone twice weekly sessions. Most funders don't give you enough money to do that, and even clients don't always find the idea of twice a week appealing, so antidepressants are definitely an option in those situations.

SPENCER: On the point of antidepressant side effects, if you look at SSRIs or SNRIs, I think a lot of people have sexual side effects. I think it's actually very substantial. It might even be as high as 40% or 50% of patients. Is that your understanding?

MATTHEW: Yeah. And I think people, we don't talk about sex much in therapy anymore, which is one of the losses, perhaps since Freud's time. But it's a huge part of the quality of life to give up.

SPENCER: And then, another symptom that is fairly commonly reported, although I think much less than the sexual side effects, is just a feeling of general dulling, or a little numbness, or maybe they feel less depressed, but it also saps some of the highs from life as well. This kind of mechanism of action, yeah, it's not exactly just countering depression. It's sort of doing something else that's not quite the same as the opposite of depression.

MATTHEW: That's right. And it's ironic, because that's a state that we often are trying to get people out of. If they've got personality disorders, we're often trying to penetrate that kind of state. But yeah, that's often what people have to settle for with depression treatment.

SPENCER: On the other hand, I think when some people find the right medical treatment for them, it's really life changing. I know a bunch of people that get on a certain SSRI or SNRI, and they're just like, "Oh my god, it just feels so much better." It's ridiculous. And they don't need to stay on it their whole life; they could do it for a year and then taper off. It can be kind of incredible. If you look at the effect sizes in meta-analyses of these drugs, what's a little misleading is they're averaging over many patients, both those that responded well and those that didn't have a response. The real truth is that if you don't respond to one, the protocol is,"Okay, we'll try a higher dose, and if that doesn't work, switch you to another one." Oddly, sometimes switching to a different one, even if it's a similar class of drug, like maybe you go from SSRI to an SNRI, or you try two different SSRIs, sometimes people have radically different responses to them. The procedure is looking for something that's not just barely helping you. It's trying to find something that has a meaningful effect for you.

MATTHEW: Yeah. And probably, one of the things that was going along a lot more in psychotherapy studies in the 80s and 90s was really focusing more on recovery and reliable change rates, whereas with meta-analyses, people have gone more for effect sizes, which is sort of moving your symptom count down a bit, sort of in a continuous dimension. But I think if people put more stock in actual remission or reliable improvement rates, that's really what you're talking about there. We want to know what percentage of people are making meaningful change more so than what's the average level of severity reduction across a very heterogeneous group.

SPENCER: Because, yeah, the average sort of assumes that everyone's just going to take this one SSRI indefinitely. And that's not really realistic. We're actually, I think it's really lucky that people respond differently to different compounds, because if they didn't, you'd only really get one shot at it. But in practice, there are at least a few different drugs you can try to see if any of them make a big difference.

MATTHEW: Yeah, that's right. Same with psychotherapy. Just because you don't respond to CBT in the hands of this particular therapist at this time in your life doesn't mean that if you try in six months or three weeks, or with somebody else, you wouldn't get a very different response. So no, I agree, and I just want to also acknowledge that I do have a bias because I'm a psychologist. People who go to the GP and get an antidepressant that changes their life, I'm less likely to see those people. They won't even come to me. So my perspective is probably skewed by my practice.

SPENCER: Yeah, I've had so many people I know have a bad experience with one therapist or one type of therapy and just sort of write it off. I've tried to encourage them. "Look, there are a lot of different therapists out there, different skill levels, different mentalities, different approaches, different personalities, and a lot of different types of therapy. So it's not useful, I think, to think of it as one thing." If you have a bad experience, that's like saying, I had a bad experience at a restaurant, so I'm not going to go to restaurants. There are a lot of other restaurants out there. If you went to one type of therapy and it didn't help you, really look at the spectrum of what's out there, because there might be one that's just much better suited to you and a better practitioner for you.

MATTHEW: That's really solid advice. YeahSome others I think about are solution focused, absolutely.

SPENCER: So the final thing I'm curious about before we end is, you mentioned that you kind of work with schema therapy. But I'm wondering, how do you think about being a practitioner with a certain specific theoretical approach, in light of this whole discussion about a lot of these therapies seeming to be roughly equivalent in head-to-head trials?

MATTHEW: Well, I think the way I'm sort of thinking about it now is that what I see consistently with clinicians is this sort of fear that when they see all of these different brand name therapies being thrust in front of them in workshop programs online, I would say 80% of my colleagues are deeply insecure that they don't know the right acronym-based brand therapy, and they're selling their clients short because they haven't done that training they need to get across it. And I guess I've been around the block long enough now, to go what you've said earlier. It's kind of better to know a couple of things really well and be effective with them. It's okay if you don't have 25 credentialed therapies listed on your Psychology Today website. It's better that you know your tools and do a good job with them. So for me, schema therapy is a great framework that helped me make sense of a whole lot of stuff. I didn't really do much past focused work at all. I was very present moment, very CBT ACT. Let's just look at what's keeping the problem going today in your life. And I found schema therapy for me just opened up an awareness of a whole range of patterns that I hadn't been paying attention to, and once I knew that literature, that was a huge learning curve. Of all the things I've learned, that took me quite a few years to really get my head around and get the knowledge up. But once I saw those patterns, you start seeing those very consistently, and it led to more efficient solutions and plans and better retention. Clients gave me clear feedback that I was understanding their situation much better. So it was pretty hard to ignore that it was working a lot better than what I was doing.

SPENCER: And have studies contested schema therapy against other methods?

MATTHEW: Yeah, they have, and I guess I have a soft spot for both DBT and schema therapy, in that these guys are setting out to deal with some of the hardest cases that can respond to psychotherapy. It's not that hard to run a two-week internet CBT study online these days, but it's quite heroic to get a group of borderline personality disordered clients to go through a randomized controlled trial across sites, across years. Schema therapy has been trialed in courses of 18 months or three years. These are huge studies to do, and it's done well. It's come up against some psychodynamic approaches. And look again, if you're being conservative, the differences aren't strong, but where there are differences, particularly on secondary variables, they do tend to favor schema therapy. Having said that, I just became aware of a study this year where schema therapy went head to head with DBT. There's actually another trial underway, same head to head, but there's this first one that was published. And again, being conservative, no differences, but where there were differences emerging, they did slightly favor DBT. So, I'm aware that schema therapy is not the be-all and end-all; it's overkill for a lot of clients' problems. Not everybody needs to have their current problems linked to experiences in their childhood. Not everyone needs a complex language of different parts of themselves, but it certainly was repertoire expanding for me as a therapist and certainly helped make a lot better sense of a lot of cases that in the past, I think I would have just kind of ham-fistedly given them pretty simplistic behavioral instructions to follow.

SPENCER: Matt, thanks so much for coming on.

MATTHEW: Thanks so much for having me. Pleasure, Spencer.

[outro]

JOSH: A listener asks: "Is the feeling of happiness learned?"

SPENCER: I don't think the feeling of happiness is learned. I think very young babies could feel happy. At least they could feel pleasure, let's say, so I think pleasure is the starting point. If we take happiness broader than pleasure, then we have to ask what do we really mean by happiness beyond pleasure? Certainly, some higher happiness would have to be learned, like the happiness of appreciating an incredible novel has to be learned since you cannot feel it until you've learned to read, until you can understand the novel, and until you can appreciate the different aspects of a novel. So, I would say that certainly there needs to be some pre-conditions to have certain types of happiness. But to serve the basic core of happiness, which maybe starts with pleasure, probably we get right out of the gate when we're first super young babies. It would be interesting to know whether there is a moment in which we don't have that. Is there some point, like when we were inside the womb or outside the womb, when we can't even feel some pleasure. That, I don't know.