Gero AI: Mobile API for Quantifying Health & Longevity
Predicting Biological Age and Stress Resilience with Smartphone Data
Data gathered from smartphones and wearable devices holds significant potential for accurately predicting an individual’s biological age and their capacity to withstand stress, as indicated by research from Gero AI.
Gero AI’s Approach to Longevity
The ‘longevity’ focused startup, Gero AI, is dedicated to “hacking complex diseases and aging with Gero AI.” They have created an AI model designed to forecast morbidity risk by utilizing digital biomarkers. These biomarkers are identified through analyzing patterns in step-counter sensor data, which monitors the physical activity of mobile users.
Beyond Simple Step Counts
The company argues that a simple step count alone isn’t sufficient for predicting individual health outcomes. Gero’s AI has been trained on extensive biological data to recognize patterns correlated with morbidity risk.
Furthermore, the model assesses the speed of an individual’s recovery from biological stress – a crucial biomarker linked to lifespan. Faster recovery times suggest better overall health and a more favorable prognosis.
Research Validation and Peer Review
A research paper published in the peer-reviewed journal Aging details how Gero trained deep neural networks to predict morbidity risk using mobile device sensor data. The results demonstrated that their biological age acceleration model performed comparably to models based on traditional blood test results.
An additional paper, scheduled for publication in Nature Communications later this month, will provide a detailed analysis of their device-derived measurement of biological resilience.
Funding and Backing
Founded in 2015 by a Russian scientist with a background in theoretical physics, the Singapore-based startup with research roots in Russia, has secured $5 million in seed funding across two rounds.
Investors include both biotech and AI-focused entities. Bulba Ventures, a Belarus-based early-stage AI fund (Yury Melnichek), is among the backers. Support also comes from private investors with ties to Valenta, a Russian drug development firm, though Valenta itself is not an investor.
From Physics to Biotech
Co-founder Peter Fedichev transitioned from a career in theoretical physics – including a PhD and ten years in academia – to biotech, focusing on molecular modelling and machine learning for drug discovery. This experience sparked his interest in the problem of ageing and led to the founding of Gero AI.
Combining Research and AI Modeling
Alongside conducting its own biological research on longevity – studying organisms like mice and nematodes – Gero is concentrating on developing an AI model to predict human biological age and resilience to stress using sensor data from mobile devices.
The Nuances of Health Measurement
“Health is, of course, far more complex than a single number,” Fedichev emphasizes. “However, if one were to condense human health into a single metric, biological age is often the most informative. It essentially indicates the toxicity of one’s lifestyle.”
He explains that a greater difference between biological age and chronological age – known as biological acceleration – correlates with increased risk of chronic diseases, seasonal infections, and complications from those infections.
Introducing GeroSense API
Gero has recently launched GeroSense, a (currently paid) API designed for health and fitness applications. This API allows these apps to integrate Gero’s AI modelling to provide users with personalized assessments of their biological age and resilience (recovery rate from stress).
Early Partnerships and Data Acquisition
Initial partners include longevity-focused companies AgelessRx and Humanity Inc. The broader goal is to integrate the model into a wide range of fitness apps. This will generate a continuous stream of longitudinal activity data, further refining the AI’s predictive capabilities and supporting Gero’s research mission, including anti-ageing drug discovery in collaboration with pharmaceutical companies.
Value Proposition for Fitness Providers
The incentive for fitness providers to incorporate the API is the ability to offer users a unique and potentially valuable feature: a personalized health measurement that allows them to track biological changes – quantifying the benefits of their fitness services.
“Any health and wellness provider – even a gym – can integrate this into their app and rank their classes and systems based on their value for different users,” Fedichev explains.
Leveraging Technology for Deeper Understanding
“We developed these capabilities to understand how ageing works in humans, not just in mice. We are now using this technology in our genetic research to identify genes, testing them in the laboratory. However, measuring ageing from continuous signals like wearable devices is valuable in itself, which is why we announced the GeroSense project,” he adds.
Resilience as a Key Ageing Phenotype
“Ageing is a gradual decline in functional abilities, but this can be potentially improved through exercise. However, resilience is often lost. This means that when stressed, recovery to a normal state is slower. We report this resilience, and a loss indicates vulnerability to future diseases, even before they manifest.”
A Holistic Approach to Longevity
“Everything we do is focused on ageing: measurement and intervention,” Fedichev states. “Our vision is to build an operating system for longevity and wellness.”
Revenue Streams and Pilots
Gero is currently generating revenue through pilot programs with leading insurance companies, which Fedichev describes as a proof-of-concept for their business model. They are also conducting an early pilot with Pepsi Co.
He draws a parallel to Elon Musk’s insurance products for Tesla owners, based on driving data, highlighting how both approaches leverage sensor data to drive insights.
Future Funding and Scaling
The immediate plan involves raising additional funding and potentially offering the API for free to accelerate data capture.
The Growing Field of Longevity Research
The field of ‘longevity’ startups has grown significantly in the decade since Google-backed Calico launched with the ambitious goal of ‘fixing death’. These startups are conducting research to extend human lifespan.
Data as the Fuel for Progress
While death remains inevitable, progress is being made in identifying potential drugs and therapies to delay its arrival. This is fueled by the increasing availability of health and biological data, open research initiatives, and advancements in machine learning for predictive healthcare and drug discovery.
Increased Interest in Health and Wellness
The coronavirus pandemic has further heightened interest in health and wellness, and mortality, driving momentum in the longevity field.
Diverse Approaches to Longevity
The field encompasses diverse approaches, from bioengineering and gene-editing to AI and big data analysis, bringing together experts from various disciplines to identify biomarkers and combat age-related diseases.
Deep Longevity and AI-Powered Longevity Assessment
AI startup Deep Longevity, spun out from Insilico Medicine, offers an AI ‘longevity as a service’ system that claims to predict biological age more accurately than conventional methods, aiding in the identification of biological factors driving ageing-related diseases.
Gero AI’s Unique Approach
Gero AI distinguishes itself by focusing on data generated by activity sensors in everyday mobile devices as a proxy for studying human biology.
Democratizing Access to Health Prediction
This approach eliminates the need for frequent, invasive blood tests, offering a continuous, passive, and scalable measure of health. The promise of Gero’s digital biomarkers is to democratize access to individual health prediction.
Beyond Bespoke Medical Monitoring
While affluent individuals can afford personalized medical monitoring, such services are not scalable to the general population.
Potential for Healthier Lifestyles and Drug Development
If Gero’s digital biomarkers prove accurate, they could guide millions towards healthier lifestyles, generate valuable data for longevity research, and support the development of life-extending drugs.
Insurance Industry Interest
The insurance industry is keenly interested in these tools, recognizing their potential to encourage healthier lifestyles and reduce healthcare costs.
Personalized Health Interventions
Fedichev notes that individuals often struggle to determine which lifestyle changes or interventions are best suited to their unique biology.
For example, while fasting has shown promise in combating biological ageing, its effectiveness varies among individuals. The same may be true for other generally beneficial activities like exercise or dietary choices.
Addressing Knowledge Gaps
Scientific research is often limited by funding and may focus on specific demographics, excluding others. Fedichev believes a cost-effective measure can address these knowledge gaps.
Leveraging Existing Research Infrastructure
Gero has validated its model using longitudinal data from the UK’s biobank, a research partner, and aims to expand its data collection efforts.
The Convergence of Technology and Social Initiatives
“Technically, what we are doing isn’t fundamentally different, but we can now do it because of initiatives like the UK biobank, government funding, and industry sponsorship. For the first time, we have electronic medical records, genetics, and wearable device data from hundreds of thousands of people, making it possible. It’s the convergence of technological and ‘social technologies’,” he explains.
The Future of Personalized Wellness
“Imagine being able to optimize lifestyles for every diet, training routine, and meditation practice. We could determine which interventions work, which don’t, and potentially test experimental drugs that have shown lifespan extension in animals.”
“With data from 1 million individuals, we believe we can solve ageing,” Fedichev adds ambitiously. “Our goal is to achieve this by the end of the year.”
A Symbiotic Relationship
Fitness and health apps are natural partners for longevity researchers, offering a mutually beneficial relationship: access to users and a halo of credibility based on deep tech and scientific rigor.
The Next Frontier: Quantifying Wellness
“We expect these apps to attract a large user base, allowing us to analyze their data and provide personalized assessments. In the background, we will build the best model of human ageing,” Fedichev predicts, stating that scoring the effects of different wellness treatments will be “the next frontier” for the industry. “Wellness and health have to become digital and quantitative.”
Applying Physics to Human Data
“We are bringing physicists into the analysis of human data. With the abundance of biobanks and signals from wearable devices, we have a long window into the human ageing process. It’s a dynamical system, like weather prediction or financial market predictions,” he explains.
“We may not own the treatments, as they are difficult to patent, but we can own the personalization – the AI that tailors those treatments to you.”
Personalization as a Long-Term Strategy
From a startup perspective, personalization is a clear and enduring trend.
