Personalized health recommendations from wearables sound revolutionary, but they're falling short for people with chronic conditions. While fitness trackers and smartwatches promise to tailor advice based on your individual biometric data, the reality is far more complicated. People with conditions like polycystic ovary syndrome (PCOS), recently renamed polyendocrine metabolic ovarian syndrome (PMOS), often find that health tech features ignore the unique ways their bodies work, leaving them to cobble together solutions on their own.

The gap between promise and reality reveals a fundamental challenge in health technology: most algorithms are built for people without complex medical conditions. When you deviate from the "norm" that these systems are designed around, personalized health often becomes a series of workarounds rather than truly tailored guidance.

Why Do Algorithms Struggle With Chronic Conditions?

PMOS affects roughly 170 million people worldwide, or approximately one in eight women, yet most health tech features don't account for how the condition manifests differently in each person. The condition is both hormonal and metabolic, and it can impact multiple organs while being associated with insulin resistance, Type 2 diabetes, obesity, cardiovascular disease, and obstructive sleep apnea. Despite affecting so many people, the medical establishment only recently updated the condition's name to better reflect its complexity.

The problem is that PMOS presents wildly differently from person to person. One person might struggle with insulin resistance while another doesn't. One might experience cystic acne while another deals with excessive facial hair. Two people with the same condition might gain weight through completely different metabolic pathways, yet fitness trackers typically offer the same calorie-burning estimates and weight loss recommendations to both.

Consider the catch-22 that many PMOS sufferers face: doctors often recommend weight loss as the primary treatment, but the condition itself makes weight loss harder. Higher insulin levels trigger excess androgen production, which causes the body to store more abdominal fat. Studies have also shown that PMOS sufferers tend to have lower basal metabolic rates, meaning they burn fewer calories per day than people without the condition, all else being equal. Yet most fitness trackers have no button to adjust these recommendations or account for these metabolic differences.

What Specific Features Fall Short for People With Chronic Conditions?

The limitations extend across multiple health tracking features that claim to be personalized:

• Calorie Tracking: Fitness trackers estimate daily calorie burn without accounting for conditions that lower basal metabolic rate, making their recommendations less accurate for people with PMOS and similar metabolic disorders.
• Reproductive Health Predictions: Algorithms designed to predict fertile windows often cannot handle hormonal birth control, a common PMOS treatment, because they don't account for how oral contraceptives alter body temperature patterns.
• Muscle Building Recommendations: Fitness features don't adjust for conditions that complicate lean muscle development, leaving people with PMOS without guidance tailored to their actual physiological constraints.
• Medication Interactions: Health apps that log meals might identify some food-drug interactions, but they rarely account for how conditions affect medication effectiveness or how different people respond to the same treatment.

The core issue is that each health tech company builds its own proprietary algorithm, and most are optimized for the general population rather than for people whose bodies work differently due to chronic conditions.

How Are Health Tech Companies Approaching Personalization?

During recent industry meetings, health tech companies consistently pitch personalization as the next frontier. The vision is compelling: instead of generic advice, your wearable would analyze your unique biometric data and recommend specific interventions tailored to you. For example, if your heart rate variability shows good recovery but you had poor sleep, a device might suggest a moderate 20-minute yoga session instead of high-intensity interval training. Or if your continuous glucose monitor (CGM) and blood test data show specific patterns, your fitness tracker might recommend particular supplements.

The problem is timing and complexity. Generative artificial intelligence (AI) is still relatively new, and companies are discovering limitations in real time. Additionally, the human body remains largely mysterious to medical science. It's difficult to offer truly personalized health tech when even medical experts struggle to understand certain conditions fully. Most damning of all, good science takes years, while the pressure is for technology to move fast.

What Would True Personalization Actually Require?

Creating genuinely personalized health recommendations for people with chronic conditions would require several advances that don't yet exist at scale:

• Condition-Specific Algorithms: Separate AI models trained specifically on data from people with PMOS, diabetes, sleep apnea, and other conditions, rather than one-size-fits-all algorithms built for the general population.
• Integration With Medical History: Wearables would need seamless access to your complete medical records, current medications, and treatment history, with proper privacy protections, to contextualize what your biometric data actually means.
• Longitudinal Research: Long-term studies tracking how different interventions work for people with specific conditions, so algorithms have reliable data to learn from rather than generalizations.
• User Customization Options: Built-in controls allowing people to adjust recommendations based on their specific condition, rather than forcing everyone through the same algorithmic framework.

Wearable technology is helping researchers discover new correlations between biometric data points, particularly in reproductive health, which offers hope for the future. But as of now, people with chronic conditions often find that "personalized health" means they've cobbled together ad hoc solutions themselves, combining insights from multiple apps, devices, and medical providers.

Why Does This Matter Beyond PMOS?

The PMOS example illustrates a broader truth about health technology: the more complex your health situation, the less useful generic personalization becomes. This affects anyone living with a chronic condition, from Type 2 diabetes to autoimmune disorders to cardiovascular disease. Each of these conditions affects how the body responds to exercise, food, sleep, and stress in ways that standard algorithms don't capture.

The promise of health tech is that it puts power in your hands, giving you agency over your health through data-driven insights. But that promise only works if the insights actually apply to your specific body and condition. Until health tech companies invest in building algorithms that account for the complexity of chronic illness, personalization will remain more marketing concept than medical reality for millions of people.