Hey. Well, hello, hello, guys. You're listening to beauty bites with Dr Kay secrets of a plastic surgeon, and today's podcast is all about ovarian aging and menopause, and can we have a model to predict when we're going to enter menopause? This is an amazing concept. We are interviewing Kiran Kumar. She is the founder and CEO of timeless biotech. This is a venture that's focused on bringing AI models to healthcare in a scalable way, and really to women's healthcare. Their first big product is called meno time, and it's the first and only ovarian aging clock, and it's kind of a machine learning model where we're using AI to enable women to track their track and hack their ovarian age. And actually, ovarian aging is so important that menopause is like a cause of precipitously fast aging, and if only you could know when you're going into menopause, how great would that be? Kieran is a biotech bioengineer from UC San Diego. She's contributed to research and machine learning powered diagnostics and longevity businesses, and she has identified the need for an ovarian longevity focus. So welcome to the podcast. Happy to be here. Okay, great. Well, tell us a little bit about you and your background, and what made you think that we need to track and hack our ovarian age? Yeah, of course. Well, the data really led me in this direction, which I believe has it's done for many different physicians and researchers in the space, because we just see this overwhelming need to focus on longevity focused protocols that are backed by data. And we find that lots of the protocols that we currently suggest don't necessarily work for women as well as they work for men. And a huge reason for that is due to the fact that ovarian aging is such a large driver for female aging. So one thing that well in the longevity space I identify was that we actually treat many things as primary hallmarks for female aging that are very reactive to ovarian aging. So like, there is a study that found that the glycan age biological clock experienced a nine year acceleration in biological age in just the six months following the final menstrual period. Yeah, exactly. That's kind of what happens. That's kind of soon as you hit menopause, you stop your collagen production, you feel like crap, your whole entire lifespan changes. Yeah, people describe it as like a second life, and it's really a huge change that comes from that loss of natural ovarian hormones being produced in your system. And it doesn't sound like a very proactive predictive thing to be measuring biological age when it's so reactive to ovarian aging. So that's why the literature is starting to find that ovarian aging and like endogenous hormone production, is actually a very primary driver of overall longevity for women. And what we should be measuring and focusing on when we're studying longevity protocols for this population, interesting. So you have a biotech background, engineering, bioengineering, and then what kind of, well, tell us about what meno time is, because you used your background to develop this AI algorithm that helps to women to predict how they're going to age in terms of their age of onset of menopause and decline of ovarian hormones. Yeah, so hormones are a mess. We all know that hormone tests are really difficult to interpret anything meaningful from, and a huge reason for that is because the static hormone value itself varies so much from individual to individual and throughout the cycle that it's it's can really mislead us, and that's why many clinicians have actually given up on testing hormones at all and like work off of symptoms and the woman's lived experiences. What we found is when we ran a machine learning model on all of these features we had, which includes blood biomarkers and includes all of the symptom information that hormone values actually tell us everything we need to know. They just don't tell it in their static hormone values alone. So the static hormone value doesn't rank high. But looking at the full picture, the interactions between these hormones, the interactions between these hormones and the symptom information, is very, very predictive of your ovarian age and time to menopause. So that's how machine learning is really able to overcome the limitations we see with hormone testing. Like our model, for instance, runs 100,000 conditional pathways per patient and 1 billion mathematical operations to train that can't be eyeballed, right? That's very sophisticated modeling. Okay, that's used to produce these predictions. Amazing. How much patient data went into creating this model? Like, how many patient lives Did you input? Yeah, so there's over 40,000 visits in our data set, and each of those visits have like hundreds of features. So that's like, Disease Information, symptom information, demographic information, and, of course, blood biomarkers. So we did a full exploratory analysis on all of this information to identify what really correlated to that final menstrual period date, because our data set actually has final menstrual period dates recorded down to the day. So it's not theoretical. It's validated on retrospective clinical data, and that allows us to actually speak with high low like high degrees of statistical confidence to our results. Wow, interesting. So you took 40,000 lives and started at the beginning of their laboratory data and try and these are women who already had entered menopause, and so you were able to go backwards in their history, put input, all the data, values and labs and their exact date of menopause, and this is the model that the machine learning is built on. So it's even more controlled than that, because this data set was actually collected over years. So these women came in well before they had entered menopause, and annually, came in and got their blood drawn, answered these questionnaires and provided all of this information, and eventually, when they experienced menopause, they reported their final menstrual period dates down to the day. So that's like 1000s of women, and that's how we overall have over 40,000 visits in the data set. Are these women that were participating in a research study or another longer kind of like, how did we get all these women together? So this is a research study done all across the United States. So that's where the data comes from. It's very geographically diverse. It's very racially diverse. The health profiles are also diverse. We didn't exclude populations on birth control, we didn't exclude populations on hormone therapy. So our model is actually very generalizable to the actual population, to actually what's happening out there. Exactly amazing. So it's almost, I would say, nearly impossible for a clinician to, like, be able to advise you with that kind of knowledge base. I mean, yes, you'll find a 30 year clinician who's been doing hormones and OBGYN For the last 30 years, and their their own brain has machine learned their own population of patients, and they can give you the gist of how they think you'll trend, but this is so much more freaking accurate, isn't it? That's what machines are really good at, right there. It's kind of overkill. They're really good at taking all of this information, identifying these patterns and creating these precise predictions. What we find is that all of these very experienced practitioners have already been solving the problem, and by that I mean promoting ovarian longevity, right, delaying menopause with their protocols. They just haven't had a way to empirically validate what they've been doing, I would say they haven't been doing that frankly, like women have been so underserved in this whole zone of health that it's only in this last year that menopause is having the moment, and that providers are actually aware that they need to be giving hormone replacement therapy, and that 95% of women in menopause are not even on hormones, and it's just kind of shocking now we've let the sector of health just die until this year, literally, yeah, and you know that's a big thing when I say that many of these experienced practitioners, I'm speaking about practitioners like yourself, right? Yes, and many other clinicians in this space who are really, really focused on delivering like, optimal protocols to their patients who have been practicing this with lots of great experience from like, their patients, and now we can find a way to, like, show, like X amount of reduction or slowed ovarian aging during these protocols that can help other clinicians who have been skeptical really adopt these protocols into their practice. Yeah. So it's like really moving the needle slowly and how to get empirical data to support expansion of what's working. So once a woman knows her ovarian age profile, what steps can she take to hack or optimize it? And yeah, what do you think in terms of lifestyle, medication or supplement strategies? Yeah. So this is a huge priority for us. We didn't want this to be like a doomsday clock. We women, yeah. I mean, on the one hand, I definitely want to know the day of menopause. On the other hand, do I want to know the day of menopause? Yeah, yeah. I want to know it if I can control it, right? That's, that's, you can influence it. And I think acceptance of it is all right, and maybe it helps for planning child rearing and also, like understanding, like your sexual desires and fertility changes. I think there's so many things you can do with that. Yeah, absolutely. Like cycle tracking. Why do people want to know when they're going to get their period? Because I need to know how I'm going to plan my life around that. Like, maybe don't. I am, like a crazy like trialathon, the triathlon, the day that I'm going to be menstruating. So, like, people definitely want to plan their lives around things, and there's definitely ways we can have timely management from knowing the date, but the big thing that people want to prioritize is seeing if they can move the needle and delay. And what we found is that our model not only identifies when you're going to experience menopause, what your existing ovarian age is, it actually knows what's uniquely driving your ovarian aging. So every woman is different. The Model recognizes that every woman has a unique ovarian signature. So the model can actually pull and see what is driving your ovarian aging, and what should you be focusing on to move the needle in terms of what like, what would it tell me like, for instance, one woman might not be Eating Enough Healthy Fats, so she's not getting the cholesterol she needs to make pregnan alone, and she doesn't have pregnant alone, she can't make steroid hormones, and that is leading to significant ovarian aging. So that woman needs to be told to have more healthy fats, like include certain things into her diet, whereas another woman might have too many fats in her diet, she might have high blood sugar, and she might want to focus on waist hip circumference instead. So the model knows which direction the woman should be focusing on, and uses that information to provide like these focus areas, and then we link those to clinically validated protocols for those respective areas. I love that. Are you seeing any interesting trends with the GLP one population, because now that is surging, and we see so many patients on GLP ones, I wonder if we're going to find that that influences it very aging, or slows it because it's anti inflammatory, or speeds it because it's almost like starvation. Yeah. So with GLP ones, we definitely have seen a whole fertility boom, like the whole ozempic baby, yeah, yeah, thing that's been going on. So that's definitely very interesting data. But we also want to look at exactly what you said. It might affect some women positively and some women negatively based on what they should be working on. Like, for instance, it's promoting a calorie deficit, right? If you are one of those women who doesn't have enough healthy fats, you're not promoting your ovarian hormones. You're not producing those hormones, because all of those hormones require a lipid raft, then you could actually be stopping ovulation. You might be having an ovulatory cycles now, and you could be accelerating your ovarian aging, whereas another woman, who may have had issues with waste of circumference, you're directly influencing that with GLP ones. So I think the big take that many people in the space have with GLP ones is, if you are on like a very careful, controlled diet, where you're including all the nutrition you need, you're including those healthy fats, you're including protein, then you actually don't experience that, like all those deteriorative effects, like muscle loss, 50% muscle loss and 50% weight loss, right? Or 50% fat loss and 50% muscle loss. You actually do have these improvements of health without lots of the deterioration. But if you're not on that control diet. There can be some risks introduced, I think so. But I'd be, I will be curious to see, I think you guys should incorporate that as one of your big data sets, because it's going to be ubiquitous that I think 70% of our patients right now, that we see in esthetic practices anyway, are on GLP ones. That's a big movement that we have is, like, using our model to empirically test the influence of lots of these different protocols on ovarian aging. So the same way you know if certain nutraceuticals impact your cholesterol, and you have to demonstrate, oh, this is not negatively impacting these standard markers in a person's life for it to be considered safe, we should know if it's influencing ovarian age before and like how it tests against that marker before we provide it to people. Interesting. What other trends have you guys uncovered in the machine learning data? Yes, so in our exploratory analysis, we obviously found some really incredible new biomarkers to predict menopause. So some include like things like masturbation frequency, like pleasure during sex, very, very high ranking predictors, more highly predicting of earlier menopause or later, actually later. So if you're hearing some more active, better longer term, yeah, more active. Better sexual health in general isn't associated with ovarian longevity, and what we all don't use it, you'll lose it. Could be causality. Could also be like, you feel like doing it more if you have like, it's like an indicator right of better ovarian health. We also find that situations that are genetic. Like something that we use in clinics a lot is hereditary predictors, like, What's your mother's age of menopause, what's your sister's age of menopause? And we have those features for our population, they just didn't crack the top 50 in predictive performance. Interesting, yes, so lots of the thinking, yeah, lots of the things that we use that are hereditary give us the sense that we don't have as much control actually don't rank that high, whereas things we can influence, like some of the things I just mentioned, like and wasted circumference and these hormone values, they are things we can influence, meaning we actually have more control than we thought. That's kind of the whole epigenetic argument, that people think that you have disease and cancer and problems because your DNA develops problems, and it's actually the the environmental influence of epigenetics on your DNA that's the number one most important cause of aging. Yeah, absolutely not DNA destruction itself. So that's so it's great, actually, it's good news, because then we have leverage to influence how we age, yeah, and it's considered like the number one lever right for health longevity, health span. Not a single organ system goes unaffected by ovarian aging. So if you can get to that base and influence it like we're not just working with physicians who focus on like hormones. We're not just working with physicians who are like functional medicine or longevity physicians, there are oncologists who are interested in this right? Preventative cardiology, we would see being interested in this fit, like brain health, immune health, all of these different areas are affected, obviously, esthetics, right? So there's just a huge range of where this tool should be, in the traditional medicine space, and like non traditional medicine space true, I think an esthetics itself. If I was to be able to predict age of menopause onset, I would do a work back schedule and make sure that we boosted people's collagen to the maximum ability possible, like when they have all this vibrancy and vitality still in their system, knowing that as soon as hormones change, we're going to lose a little bit control and start having collagen destructions and kind of be banking and building up their collagen preventatively in the few years prior to that 10 clock. So I think that's so cool to know. Yeah, for sure, I've done a lot of skin care, so that's a huge passion of mine, like skin care, like hair as well hair health and some of the data on what how much happens before you notice it, especially in like hair health is incredible, like people lose up to 50% of hair before they realize they're experiencing hair loss. So true. You really need to know when these transitions are happening in advance, like these hormone fluctuations are happening in advance to catch things, like, like you said preventatively, yeah, I think we have to change how we think about things, because I think people feel that they're getting handed this sentence, when actually we have so many options. So what we can do to start reworking how our physiology works way ahead of time. Yeah, absolutely. Like, one of the things I like to mention is we talk about menopause as this, like singular event, or event that takes maybe a few years when, in reality, we what starts first, we start to see a decline in melatonin first. Really, yes, sleep signals turn Yeah, exactly, so when you start, but that's not a commonly checked lab. How are you guys tracking that? So melatonin, this is now in the literature, where we start to see melatonin is what declines before we see that decline in estrogen, before we see that decline in progesterone. So people start facing issues in sleep. Then after that, you start to see that spike in FSH, and that's what disrupts that is a signal of disrupted ovulation. That's when you start to get the progesterone decline, the estrogen decline, and all of the symptoms that we know about, hopefully know about a menopause come there. But if you start earlier with lifestyle practices that can influence melatonin or with even exogenous supplementation of melatonin sooner, there's data to promising data to show that melatonin might be associated with later menopause. So there's a lot we can do preventatively before we lose these very important hormones. So interesting talk a little bit about if women are on birth control, if they're on hormone replacement therapy, or they got put on it, you know, ahead of menopause, can they still do the meno time test? Yes, we are very, very proud of this, because, unfortunately, many models be especially since they rely so heavy on the theoretical side of things, and they don't build off of clinical data. Exclude these populations. Even though 60% of reproductive aged women are on birth control, and many women going through menopause are on hormone therapy, we are able to account for these populations because the model is able to learn. How these hormone values are influenced by the hormones they're on, and use that to create, like, still an accurate prediction. I love that. And the only people that are excluded from end of time testing are if they've had overectomy or hysterectomy. Yes, and why is that? Well, the data set excluded those populations. So we beginning, yeah, technically, if you continue to add data, would you want to include those populations Absolutely? So a huge part of what we're trying to do is create this huge data set like that recognizes how certain interventions impact, like ovarian aging, how procedures impact, to know that like but so many women get their uterus removed, and like, we don't even know does that influence, like the aging pattern of the uterus and the like the uterus gone? Do your ovaries still work the same? We have reason to believe that there's huge changes. I have a lot of anecdotal examples I could provide of the type of experiences women have had, and interesting stories that clinicians have told me as well. So we definitely like to do more of these case studies and research partnerships with physicians who are consistently working with these patients to see how it impacts them. Yeah, I think that would be very, very good knowledge. How does the OBGYN community review this test like something powerful they're going to adopt. Do you feel like they'll have a lot of adopters or a lot of naysayers like that? Just say this is AI algorithms that we can't trust, so we expect it to get a lot of naysayers, which we actually did not. So that just goes to show that the OBGYN community is rightfully wary of like, non science backed and non statistically like significant like, non statistically competent models that are in the space, especially surrounding fertility, like we consistently look at models that don't work off of individual cohort data, that are theoretical entirely and don't have so strong clinical application that are well adopted in the space. So once we communicate the rigor of the model, once we communicate how we're overcoming the limitation of one blood test, because one blood test is considered a really bad thing to say when it comes to hormones, for good reason, once we communicate that they're actually very receptive and very excited to introduce this product because it is such a need in this space. Interesting. Do you think that patients will need special counseling when you sit down to go over the results of this test with them? Because I'm just trying to envision my first patient that I show their data to and what their reaction will be. I almost think it's like it could be a very heavy set of information to deliver. So there's a reason we wanted to start by launching through physicians, and we don't want to circumvent the physician. I think that's a mistake that's being made a lot, where you go direct to consumer, yes, the direct to consumer route, and I see there's room for that in certain applications, for sure, but especially for a test like this, where we're talking about physicians who aren't testing these hormones, not because they don't want to, or not because they they just don't think it's worth it. It's because they genuinely don't see a value add in getting these static hormone values and in a one time blood test is very difficult to interpret without a machine learning model. So once you solve for that problem, then you find that the willingness is there to test and to, like, sit down with your client who now doesn't have to look at hormone values that you have to explain. Like, I know your AMH is low, but that might not mean something. You might still be able to get pregnant naturally, but you could do IVF like, having these long conversations that could backfire. People had so many bad experiences. Now you can use a report that gives you better data to work with so you can have an easier conversation with going to look like, like kind of a pattern of your AMH is here, your Lh is here, your FSH is here. This Where is where you fall in the machine learning pattern of all these lives that we've looked at already. So it actually shows you your ovarian age trajectory. So it tells you, like where you are, when you can expect that final menstrual period, and like your existing what degree of certainty is this in specificity, sensitivity? Yes, all of the things I learned in statistics. So specificity and sensitivity are great for diagnostics, but this is a survival analysis. So because we're predicting time to event, we give an error range, so with no horizon at all for all populations, birth control, not on birth control, our mean average error is 1.5 years for long range predictions. So the further you are out, obviously, the closer you get to your onset, the less control you have over your time to menopause, and the narrower our error window gets. It gets more and more accurate and precise, but we're definitely closing out that window because we see. That's only 79% of women have menopause in a 20 year range, and this model is able to narrow that down so significantly so that people can actually work off of their ovarian age, rather their chronological age when they're planning for fertility, seeing if they want to do IVF, if they don't want to do IVF, if they want to conceive at all, if they are expecting menopause earlier, are they at risk for poi, like all of these different factors, interesting. Explain what that means that what you said about the 20 year range, yes. So our when you look statistically at the United States, only 79% of women have menopause, their final menstrual period date within the ages of 40 to 59 so that 20 year range, meaning there's a interpretation that all women have menopause within the ages of 44 to 54 20% of women have it randomly, but we actually see that only 50, around 50% of women have menopause within a 10 year range, and 79% of women have it in a 20 year range. And like, when you look at all of women, you're talking about a ginormous range. Yeah. So when you're actually telling women, oh, like you're too early for menopause, or you're too late, like you must have already hit menopause, you're working off of numbers that only, like, talk about 50% of the population. That's why we see perimenopause starting to give people problems at such different ages. Like some people enter perimenopause at 3035, and they have this early symptoms. So like that bell curve of where the majority of women hit menopause is not so, so accurate. It's not accurate at all. Like there's a very wide range. And if you take that forward, and I talk about the fact that we're telling women, Oh, you're 44 you definitely don't, like, can't have kids naturally, or, Oh, you're 37 like, you should, like, consider the maternal age, yeah, yeah. Like, you, you are a geriatric pregnancy. That written that that heard that from my fourth daughter? Yeah, like all of these conversations are happening, and sometimes a they don't need to be happening, or sometimes they need to be happening in reverse for someone who has a younger chronological age. So the same way you have this huge range of menopause, you have this huge range of people who may be in a high risk pregnancy, may not be in a high risk pregnancy, that applies to their fertility as well. So ovarian age is like a better proxy than chronological age, even in fertility matters. And I guess naysayers would also tell you, like, Okay, you're gonna tell me that 1.1 to 1.5 year range of when you're gonna hit your exact menopause state. So isn't that apparent? Like, Can't we tell women the average age is 51 mostly people get it 5051 and everybody in that ballpark. Well, so what do I really need to have an expensive feather test to tell me similarly, around 50 I'm not above. So it depends, because when you look at like that range, we discussed like that 20 year range for 79% of women. So that means, if you're lucky enough to be in that standard 80% of women. You have a 20 year range. The amount of uncertainty that's reduced by this one to 1.5 year range is incredible. Already a lot tighter. Yeah. So when you tell someone, oh, you can expect it at 51 and then they hit it at 41 What are you going to tell them now? Right? Well, I told you to expect it at 51 but even looking. We weren't even paying attention in your 40s. Sorry, exactly. And that's like Final menstrual period day. Your perimenopause journey is starting well before that right now, this woman is definitely not equipped, and now she thinks she's not going through menopause, because that's supposed to happen at 51 so when you give these ranges, and then you don't tell them there's a plus or minus 10 year error, and then you don't tell them that they can expect perimenopause to happen a couple of years before that. Yeah, then you are setting them up for failure. Interesting. So with this data, a one to 1.5 year range of when I can tell you you're going to exactly enter menopause, is what we're going to get on the report. Yes, so that's where you have the plus or minus 1.5 years for long range, and then it gets more and more accurate the closer you get, and that's how we narrow down that huge window. Okay, interesting. So in one of the questions that we have, of course, is, you know, with an AI driven model, how do we ensure that we have, you know, looked at all different populations, that we have ethics and trust with the data that's going in there, and that we have, like, good spread of racial variety. You know what it what are your how have you guys shaped this model to make sure that it has good transparency? Yeah, so there are two things there. One is, like, obviously, making sure that things are diverse. We are a model for all women find that there's so much exclusion criteria in some of the models in the system that I'm like, Who is this one perfect person who you're able to run this test on, who's not in your exclusion criteria? So it was a big thing to us to not exclude the populations in our data, and our data is racially diverse. And it's geographically diverse, and it doesn't exclude populations birth control hormones, doesn't include populations with PCOS, doesn't exclude populations on cancer treatment, doesn't exclude populations with endometriosis. The model just has to learn about these different populations. So that's how we really held ourselves to that standard. We didn't try to exclude things to make a fake higher performance model, and then we stratify in our analysis so that we actually report accuracy for each of these different populations, and we make sure that it holds for people that we are actively trying to encourage to use the model. So that's how we were really it was really important for us to be responsible about that. And transparency really is great for any sort of AI models in health care, you don't want things to be a black box, and that's where the interpretability of AI comes in. And we use like Lyme assessments, which basically tells us what's driving each individual person's prediction. So once I run your report, I actually know what drove your prediction uniquely, and that's where I can also start to give some feedback and like insights of what you can do next. That's interesting. Is it all US population women, or is it global? It is all US population? Okay, so perhaps there's a different pattern of aging globally, which we'll find, I'm sure, in future years, right? Yes, we do have some data on that. Like we know that the earliest menopause dates, I believe, are in India. Highest prevalence of poi is in India. We do. We are premature ovarian insufficiency. Sorry, I'm so used to Yeah, I think. And for me, yeah. For me, early menopause, essentially, India, really, yes, that's bad for us, that that is definitely bad for us. And we're trying to figure out, like, why is that? Is it because of, like, toxins, pollution, like other factors that affect epigenetics or environment? Is it how much is hereditary, how much isn't? So that's definitely some of the research we're also going to be doing, because we will be looking at clinics and running retrospective analysis on different clinics globally and how their lifestyles impact ovarian age, interesting, and then using this kind of concept of AI algorithms to predict aging. What other things do you want to do with this technology, like other pathways of longevity you want to look at or address, or do you have plans to capture some other data within this data set? So there are so many different points of data capture that we are interested in, like I said, we really want to build a library of existing protocols that exist and determining how those influence ovarian age. There's a lot of great therapeutic interventions that people are looking into to delay menopause and procedures, but some of them are not as accessible and not things that really can be widely adopted. So there are already, like great supplements that focus on mitochondrial health, which is so important for the ovary, there are great supplements that focus on, like, big adaptogens that really can actually increase your levels of estradiol and reduce FSH, which are associated with promoted ovarian longevity that we can already explore and study. So that's a big, big priority for us. But when it comes to longevity, especially for women, well, solely for women, we're talking about ovarian longevity and ovarian aging. So when it comes to that space, this tool is really going to be the foundation for anything else that we build upon. Interesting, how so how accurate is the ovarian age model? Like, has it gone wrong, or has it been accurate on all predictions? So that's where we get that error range. So basically it tells us a plus or minus 1.5 year prediction for long range. And then that gets more and more accurate the closer you get to the event. So when it comes to clinical accuracy for predicting menopause, obviously it's more clinically relevant to predict it when you're closer to the event. So that's why it's like very encouraging that we see high accuracy for there. But when it comes to promoting ovarian longevity and talking about how much control you have over that date, it's also encouraging to see that in long range predictions, you have some wiggle room so that you haven't seen it totally fail, where, for example, in your when your range is like, okay, 4047, but ends up being 40. So in our population, there's always going to be outliers. So that's one thing that we're very, very clear about in every single model, diagnostic models, predictive models, even the most leading ones, there's never a situation where there aren't outliers. So we're really, like, clear about, like, communicating that if you're still experiencing menopausal symptoms, regardless of what our model says, you need to, like, go in and make sure you're exploring your options. Yeah, yeah. It still requires a human supervision. Always, always, as a clinician, how? Do you think this is going to guide people in omegan practices in terms of their fertility management or their menopause management? I think a big aspect is going to be, especially in fertility, is really confirming protocols, validating existing protocols. All of the research we're working off of infertility really comes down to time to conceive data, which is a big issue, because, like, 50% of the person involved in conception is not being studied. So it's really difficult to control for your outcomes and actually get meaningful results from time to conceive data because you're there's so much you're not able to control. The data set is not very generalizable. The models you produce are not very generalizable, so it's been a huge limitation in the fertility space. However, if you can now test these protocols that physicians have been using based on time to conceive data or observational data that they found promising and identify how they might be reversing or slowing ovarian age, then you can get some more high statistical confidence, high quality data that's actually going to generalize very well to your actual population that comes into your clinic. So I think a huge part is for OBGYN to start like these pilot studies to see how this is like, these protocols are influencing their population, and then they can actually get better data to like, iterate and improve their protocols, interesting. So for a woman who wants to go down this path and better understand their ovarian age and health, where can they reach you or get more information? Well, they can go to our website, timelessbiotech.com, we're on socials. They can definitely check us out everywhere there, if they want to use the test, we are going to be launching in January with some select early partners, including Dr Kay, who are going to be able to offer it and really walk the patient through their experience and how they can start tracking and hacking their ovarian age to promote longevity, better health outcomes and all fronts. Well, this is going to be incredible. And I can't wait to watch this technology evolve, and to see actually how I use this tool in practice. Because I think women will be fascinated to know this information of their age of menopause way ahead of time, hopefully, and give them that range of where it's most likely going to happen. It's going to be fascinating. Yeah, absolutely. And helping physicians validate protocols, letting them use existing blood work, like we don't run blood work. We're like software, light application layer on existing blood work that they run. So it makes things super easy for the physician, super scalable, and just puts data behind all of these things we've been doing in women's health. I love it. Well, you are absolutely brilliant, and you're so innovative. And I can't wait to check out more info. You guys should look up meno time and timeless biotech, and stay tuned. Come visit me in my office in LA. We'll be doing this in January, and I'll kind of like showcase little things prior to the launch, so we can all learn more together. That's it for now. Guys, don't forget, you got to find me on my instagram. It's Beauty by Dr Kay, D, R, K, a, y, and doing amazing things with people's faces and collagen stimulation, all the things that help prevent this precipitous decline with skin aging as we reach menopause, and our website is the same. It's by Dr kay.com I'll have to send you some of our skin longevity launch products. We have a new peptide based skin longevity line. You guys are gonna love it. That's it for now. Guys, stay beautiful. You you.