Finding the chase and cutting to it

Breaking the 80% Ceiling of current diabetes technology.

For decades, the clinical approach to insulin replacement has treated the human body as a predictable, linear machine. Standard “Total Daily Dose” (TDD) and “Basal/Bolus” ratios operate on the assumption that if a specific amount of insulin is delivered, a predictable result will follow. While early insulin pumps introduced distinct profiles for weekdays and weekends, and modern Hybrid Closed Loops (HCL) now react to real-time sensor data, these systems remain fundamentally limited, based on reactive responses to basals that rarely change.

Most current HCL users hit a “plateau” of 80–85% Time in Range (TIR). This is because these systems are reactive; by the time a sensor detects a glucose excursion, the physiological event—whether stress, hormonal shifts, or delayed digestion—is already in progress. They are chasing a moving target using a linear lens to view a chaotic biological system.

Embracing Chaos: The Attractor Model.

Drawing on James Gleick’s Chaos Theory, we must recognize that diabetes is a dynamic system where small variations in initial conditions lead to vastly different outcomes. A “standard” Tuesday basal rate may fail on a Tuesday where the user is stressed, at a different altitude, or at a specific point in a menstrual cycle, or a storm is brewing changing the environment in which the pump is operating.

In 1997, I began conceptualizing a model that moved away from “averages” toward a framework of attractors. In this framework, an attractor is any quantifiable influence, such as altitude, barometric pressure, or site location, that pulls the metabolic system toward a specific state. Instead of viewing these variables as “noise” to be filtered out, my framework treats them as deterministic pivots that shift the system’s equilibrium.

Quantifying the “42 Factors” with 2026 Technology.

The 2018 publication of the “42 Factors That Affect Blood Glucose” by Adam Brown provided a taxonomy for these influencers. However, the real breakthrough lies in our ability to quantify them in 2026 through the ubiquity of wearables and the Internet of Things (IoT).

By utilizing a smartphone as a central processing server and wearables (watches and sensors) as data donors, we can establish individual “biological signatures” tied to the person with the pump. We no longer need to guess the impact of laying in bed, a commute, or an intense exercise session; the hardware provides the high-resolution data streams (accelerometers, temperature sensors, GPS) to establish these values in real-time.

The Core Logic: Predictive Tailoring.

The framework operates by taking a personalized “base” basal rate and modifying it through a cumulative total of active attractor coefficients.

• Environmental Attractors: data such as temperature, altitude, and barometric pressure are pulled from the internet and localized to the user’s environment. These may impact the user or the equipment.
• Physiological Attractors: menstrual cycles, amount of sleep, or activity types, stress levels, and illness are mapped via calendars and biometrics.
• Decay Factors: the model accounts for the “lag” or “tail” of certain events, such as the lingering metabolic effects of long-distance cycling or the degradation of an insulin infusion site over several days. Even the degradation of the insulin in the pump.

By multiplying the base rate by the sum of these active attractors (which may be positive, negative, or zero), the system suggests a predictive basal rate for the next 24 hours. This moves the HCL from a reactive “safety loop” to a proactive, contextual model.

Safety, Autonomy, and the “Human in the Loop”.

A critical component of this non-linear framework is user autonomy. Unlike current “black box” systems that make decisions in secret, this model suggests a “sane” value to the user, who retains the power to accept, modify, or ignore the suggestion.

Furthermore, to prevent the risks of cross-contamination, especially in households where multiple people may require insulin replacement from a primary carer, the framework utilizes a physical cable connection for uploading basal rates to the pump. This ensures that the specific attractor model and pump ID are securely encoded to the correct individual, eliminating the risks associated with wireless interference or signal crossover.

Conclusion: A Unified Path Forward.

By unifying the “42 Factors” into a dynamic, nonlinear framework, we provide the Hybrid Closed Loop with the context it currently lacks. We can finally account for the pre-menopausal woman’s cycle, the athlete’s recovery, and the traveler’s changing environment. This is not just about better mathematics; it is about reclaiming autonomy and achieving results that finally break through the linear ceiling and coping with the mistakes non-contextual hybrid loops make across the world.