Insulin Resistance and Weight Gain: A Conversation with Dr. Alexander Banks

‍Dr. Alexander Banks is an Assistant Professor of Medicine and the director of the Energy Balance Core Facility at BIDMC and Harvard Medical School. This facility is designed for the measurement of energy intake, nutritional status, and energy expenditure in experimental mice. Dr. Banks also uses molecular genetics to clarify the mechanistic basis of disease pathology and has a keen interest in the relationship between insulin resistance and obesity.

Nutritionist and registered dietitian Heather Davis, MS, RDN, LDN, and Dr. Banks cover a wide variety of fascinating topics in this intimate conversation about metabolism and metabolic health, the relationship between insulin resistance and obesity, stress and glucose, CGM research in mice versus humans, and much more.

‍Heather: Dr. Banks, I want to thank you so much for taking the time to talk with me today. I'd like to start by having you share some of your primary research interests. 

Alexander: I have a deep interest in metabolism and energy expenditure. We all understand how what you eat is going to affect your body weight and also how much energy you burn is also going to affect body weight. And there's this question of how much can you really change your energy expenditure? And human subject research is really hard. It takes a long time, it's expensive. There are limits to what you can ask people to do in order to change their energy expenditure. So we really like to study mice. 

Hacking Metabolic Rate

Heather: Tell me more about mouse models versus human subjects when we think of hacking metabolic rate.

Alexander: A lot of the things about mouse energy expenditure translate pretty well to people, plus you can change their genes, which is something that you don't try in people. And so we've been studying energy expenditure in mice, which is really fascinating. You can get mice to triple or quadruple their metabolic rate in a way that it's not yet clear you can get people to do. 

But of course, those are pretty extreme interventions. And so we've been working pretty hard to understand not just how you change metabolic rate, but what's normal for a mouse, what's a normal metabolic rate? And what we found out was that it actually isn't known. 

Normal metabolic rate is actually really hard to define in people too. There's a lack of standards in mice and there's no standardized method for doing it. We wanted to put data together to say, here's how we analyze it, and here's what's normal, so that we can go ahead and figure out what's abnormal or how we might take normal and move that to a slightly higher energy expenditure. Our goal was to help these mice to have a higher metabolic rate and help them to burn energy, burn fat and potentially lose weight. 

So really, before we can hope to apply this sort of research to people, we have to get our hands around what's happening with the mice. 

Heather: That makes sense. I think there are so many poorly defined terms in this area too. I think a lot of people will confuse metabolic rate with metabolism as a whole. They don’t realize that metabolic rate is just one component of metabolism but metabolism itself includes so much more than this. How might you take your findings from mice and direct them into human clinical trials? 

Alexander: I would love for my research to lead to this. I'm first trying to establish the normative values for mouse metabolic rate. How do you measure it? How do you define it? What are we calling metabolism? And how would you analyze that? And then once we've figured out how you would do that in mice, we can finally say, well, how about people? 

There are all sorts of ways of measuring metabolic rate in people, but if we understand more about what is “normal”, then you could actually, with a little bit more confidence, say, here's where I am in the distribution. There's going to be a bell shaped curve of high and low metabolic rate, and completely normal. There's really not a great way of doing that right now. We can’t answer the question: is my metabolic rate high, low, or normal? 

Heather: Everyone wants to know this as it relates to weight loss—they ask, how can I affect my metabolic rate? What can you say about people, beyond mice right now?

Alexander: There are some tried and true ones that have held up pretty well. Build your lean muscle mass. This boosts metabolic rate. I don't think it's controversial. It's true for mice, it's true for people. You can increase your metabolic rate by having more muscle mass. And so resistance training, cardiovascular training, to a lesser extent. 

Heather: Yes, definitely. The lean muscle mass component is so important. It's something I work on with a lot of people from a nutritional standpoint as well. Your workouts are often only as good as those nutritional foundations. Are you eating to support that lean muscle mass? How are we supporting your lean muscle mass from every angle we can?

Insulin Resistance and Obesity: The Two-Way Street

Heather: Thinking about lean muscle mass, I want to pivot to the topic of insulin resistance with you. I know we see a bi-directional relationship with obesity and insulin resistance, where they both influence each other. It seems like the majority of research examining this two-way street suggests that obesity might be the stronger leading factor. Tell me your thoughts on that two-way relationship. 

Alexander: Yeah, it's definitely a vicious feedback loop. Obesity seems to drive insulin resistance, which then drives more obesity and then drives more insulin resistance. But it seems like even a small amount of increased weight can cause insulin resistance.

On the flip side, if people who are overweight lose about 10 percent of their body weight, they can see tremendous improvements in insulin sensitivity. That brings us back to muscle mass. How do you make sure you hold on to your lean muscle mass while you're losing your fat mass? That's all part of the same conversation of weight loss, insulin sensitivity and health. 

"People who are losing weight successfully with these new medications, or the GLP-1 agonists or the diagonists or even now the triagonists, they lose fat mass, which is what they're hoping to do. But now they're also losing lean muscle mass with these drugs as well. That’s a concern, because it might mean your metabolic rate drops." - Dr. Alexander Banks

Heather: Such a great point. Yes, exercising, eating the right things to support that lean muscle mass. The GIP/GLP-1 weight loss drugs are a hot topic and raise a lot of questions. I've heard from other experts who are concerned about the side effects but they’re also asking the question: are we really correcting the underlying metabolic imbalances with these drugs, or is that something that we're still going to need these diet and lifestyle changes to do alongside? What are your thoughts on that? 

Alexander: I certainly do think that improving lifestyle isn't just going to be one medicine. It's going to be improving diet, improving exercise, as well as perhaps losing weight with these pharmacological agents. 

Can You Be Obese and Metabolically Healthy?

Heather: I agree—it’s all working together! And coming back to this question of insulin resistance and obesity, I'm also curious about how you think of the two groups of outlying folks we see where on one side you have the so-called metabolically healthy obese and then on the other, you have those considered metabolically unhealthy normal weight in terms of BMI.

I know these represent some controversial ideas in metabolic science. What's your take on how this relates to that relationship between insulin resistance and obesity? 

Alexander: So everyone's different, right? If you look into East Asian populations, a slight increase in BMI has much more deleterious effects on insulin sensitivity than a much larger increase in body weight does on other populations. And so I think it depends on your genetic makeup. It's going to depend on whether you could be a metabolically healthy obese person. 

And the other question about that is how long can you be considered metabolically healthy obese? You might have a ticking time bomb where at some point you're no longer a metabolically healthy obese 70 year old, whereas you might have been a fine metabolically healthy, obese 60 year old. So I think everyone depends, and it's going to depend on what your doctor tells you to do. And I'd say listen to your physician. 

Population level health advice is you should have a BMI under 25, you should eat right, you should exercise, but really, it's going to depend on each person and their individual makeup. 

Heather: I think that's a great point about the timeline that we're looking at because one person may be metabolically healthy at a snapshot in time, but are you extending this into the future? Ten years, 20 years? What does that look like over time? 

Alexander: And moreover, people do tend to gain weight over time. So you might be metabolically healthy with a little bit of obesity. You might be a BMI of exactly 30 and officially a person with obesity. But then five years from now, it might creep up to 34, and then you're both older and you have more adipose tissue to contend with. 

Stress and Glucose: A Surprise from the Mice 

Heather: I also have to dig into this one study you did that sparked my interest more. I wanted to bring up what you noticed about fasting stress in mice. Specifically, what you noticed about glycemic variability in mice when you were fasting them prior to a glucose tolerance test. You found some surprising things. Tell me about it. 

Alexander: We did not expect any of the results in that paper, the entire paper! We were able to monitor glucose levels in mice with this brand new technology that we had just acquired. We were able to measure glucose levels every one minute, and we did it in mice for 42 days. One of the first things we thought to do was a glucose tolerance test in these mice. 

The standard glucose tolerance test involves fasting the mice beforehand. And this has been done for at least 40 years since people had these handheld glucometers to do glucose tolerance testing in mice. And that's just been the assumption and received wisdom that their glucose levels will be more consistent heading into the glucose tolerance test if you fast them before you do the test.

And what we found, first of all, just monitoring the mice without any fasting intervention at all was that their glucose was already really good. Their glucose variability was extraordinarily low when they were eating and going about their regular routine. It stayed within a very narrow window, even when they were theoretically sleeping during our daytime nocturnal mice, their glucose barely changed by more than 15 mg/dL. All day.

So that was weird because we believed we had to fast the mice to make their glucose more consistent. But it was already consistent without the fasting. And all the papers I've ever published involve a fasting glucose tolerance test. So we fasted the mice, and we saw this huge spike in glucose. 

I was really extraordinary. It was very quick. And it wasn't just the glucose. We were monitoring all these other metabolic parameters. The mice sprint around the cage, their metabolic rate goes way up with all this exercise. Their oxygen consumption goes up. Also, their fatty acid oxidation drops because they're fasting and they're now exercising. And so this was only increasing the variability of their glucose, of their metabolism, of their body temperature. The fasting approach was the opposite of creating a stable glucose baseline.

And then when we did the glucose tolerance test after fasting, we had a far more variable glucose tolerance test in the fasted mice than in the non-fasted mice. 

And we saw something else that we weren't quite expecting. Every mouse that we fasted went into torpor. It's a hibernation-like state where their body temperature drops by about four degrees. Their metabolic rate goes down 80 percent. And that has interesting implications for people who are doing time-restricted feeding or calorie-restriction studies in mice, where you'll feed them their allotted amount of food, the mice will eat all the food, and then they're basically fasting for 23 hours, much like in our study. 

So you can extend the lifespan of a mouse when you feed them only 70 percent of their normal calories, but then their metabolic rate drops 80 percent for those 23 hours. So I'm interested in longevity, but I'm not sure that calorie-restriction under those experimental conditions is really telling us what we think it is. We found all kinds of things that we weren't expecting. 

Heather: It sounds like the ability to monitor glucose continuously allowed you to see a much richer picture of what was going on for these mice compared to the snapshot approach of a glucometer reading.

Heather: What we are also talking about here is stress and specifically how fasting stress may be impacting glucose metabolism—one of my favorite topics!

There are of course limitations in terms of how much we can extrapolate from mouse data and apply it to human cases. But it does raise some really interesting questions, such as: are there certain fasting protocols that might be inducing a stress response that we're taking for granted in terms of how that might be impacting metabolic function, including glucose regulation? 

I think we need to be asking this a lot more. Have studies like this changed how you plan to do future studies or perhaps even planted a seed for what you might want to focus on in future research questions? 

Alexander: Yes. The first part is the need for standardization. Everyone is doing their experiments in different ways. And if you look at the review articles, people are suggesting different amounts of fasting. And now we're saying actually you get better results if you just don't fast the mice. If we could at least standardize the field, we could all get more consistent results. We could make our research more reproducible if we can standardize our protocol. So that's the first part. 

The next part is about identifying unknown sources of stress in our experiments, both in mice and in people. We really don't want to stress out our mice. We want the mice to be happy and healthy and as stress-free as possible. The same is also true for our human patients and friends. Can we reduce the stress in their lives?

"We know the hormone cortisol induces glucose production from the liver. That's been known for a long time. Chronic high stress is going to contribute to insulin resistance" - Dr. Alexander Banks

Heather: I completely agree with that. There's a sweet spot for stress, including fasting stress. I’m always asking: how can we assess that sweet spot from an individual level? So Alexander, tell me—what are you most excited about exploring more in your upcoming research projects?

Alexander: There are two parts of my research lab. One part is working on standards and methods, and that's related to the paper we just talked about where we were using continuous glucose monitoring and indirect calorimetry to get a comprehensive physiological readout of mice. 

When we did this first experiment, it had limitations. It was done in young, healthy mice that were on a great diet, high fiber, low fat diet. We put mice on a high fat diet for a short period of time, but they were still pretty healthy mice.

We really don't know what happens in mice that are much more overweight, that have had a much longer history of bad diet and lack of exercise. We haven't looked at other tests like the insulin tolerance test or the pyruvate tolerance test. There are a lot of other things that we can do from there. 

And then, the bulk of my research is related to factors that come out of adipose tissue that contribute to insulin resistance. And so I'm really excited about some of that work. We're putting it together for publication soon, but there's a lot that we still don't understand. What are the factors in humans which you can model in mice that are contributing to insulin resistance, that are caused by increased adipose tissue? And I'm really excited about a lot of that work. 

Heather: I think it's interesting from not only a pharmaceutical standpoint for eventual possible drug development that could target some of these molecular elements, but just in general, being able to see the mechanism outlined in the ground detail is key. I'm excited to see what's in store as your work evolves. I’d love to stay in touch!

Alexander: That would be fun. I very much enjoyed talking to you today, Heather. Thank you for reaching out to me. 

Heather: It's been great. Thank you so much for your time. 

Learn more about the connection between insulin levels and weight gain here.