Emerging technologies have enabled patients, clinicians, and researchers to track important behavioral markers for patients with diabetes, such as sleep, diet, and activity levels, more accurately and in real time. However, each of these self-assessment tools has unique advantages and limitations.

During Behavioral Markers of Diabetes Self-Management—State of the Science and Measurement Controversies—Honoring the Legacy of Mark Peyrot, experts discussed the various self-assessment tools available for patients with diabetes and paid tribute to the late Mark Peyrot, PhD, who left a 40-year legacy in diabetes research and care, and is considered one of the founding fathers of behavioral diabetes.

The session can be viewed by registered meeting participants at ADA2023.org. If you haven’t registered for the 83rd Scientific Sessions, register today to access the valuable meeting content through August 28.

Ecological Momentary Assessments

Ecological momentary assessments (EMA) can aid patients with diabetes and their care teams by providing insights into a person’s glucose monitoring, insulin administration, and other patterned behaviors, according to Soohyun Nam, PhD, APRN, ANP-BC, FAHA, FAAN, Associate Professor of Nursing, Yale University.

EMAs encourage participants to track their daily feelings and behaviors while capturing data about the individual’s environment. This allows patients, clinicians, and researchers to analyze how a person’s social setting affects their actions. EMAs often require participants to log responses several times each day through their mobile phones.

“It can reduce recall bias compared to other conventional retrospective measures,” Dr. Nam said.

Instead of asking the patient about the past two weeks or the past year, EMA can ask about how the patient is doing at the moment or how much physical activity they engaged in in the past two hours.

Studies that employed EMA found that environmental factors can affect a patient’s adherence to tracking their glucose levels. One study found that when adolescents with type 1 diabetes felt a strong desire to blend in with their peers, it correlated with an increased likelihood of glucose checks. When they felt a strong desire to impress the people around them, occurrences of glucose checks decreased.

EMA data can also be used to track an individual’s sleep patterns, dietary management, and physical activity levels. Dr. Nam’s research discovered insights into how racial discrimination affects social behaviors. For example, Black people who experienced more self-reported instances of racial discrimination became more sedentary afterward.

The most significant limitation of EMAs is adherence. People can miss alerts or purposely decide not to log their data. Dr. Nam reported that adherence in her experience ranged from 59%-96.5%.

Sleep Monitoring

Sleep is a vital protective and risk factor for diabetes outcomes as it can lead to physiological, behavioral, cognitive, and mental health effects. Sarah S. Jaser, PhD, Associate Professor of Pediatrics, Vanderbilt University, reviewed the efficacy of various sleep-monitoring tools.

The merits of actigraphy include that it’s highly correlated with polysomnography and enables a detailed assessment of a person’s natural environment. It shares data regarding sleep duration, latency, and efficiency. However, actimetry sensors are often expensive, and missing data can be common based on an individual’s consistency in wearing it.

The advantages of consumer-grade wearables are that they’re already worn by many people and provide valuable data about daytime and nighttime activities in a person’s natural environment. But consumer-grade wearables often need settings adjustments to achieve an accuracy level similar to actigraphy devices, and many overestimate sleep duration. However, Dr. Jaser clarified that these devices continue to improve rapidly.

“Just a few years ago, these weren’t really recommended for research because they weren’t as highly correlated with polysomnography, but the newer generations of these devices seem to be more accurate,” she said.

Sleep diaries are cheap, convenient, and allow for within-person analysis (such as variability in sleep timing). Since this method of reporting is less automated, it can often lead to retrospective reporting when people forget to log entries, which can compromise the reliability of the data.

Looking ahead, Dr. Jaser said EMAs could potentially be leveraged for sleep monitoring. Linking sleep data with continuous glucose monitoring (CGM) and automated insulin delivery systems could lead to more comprehensive, accurate data.

Physical Activity and Sedentary Behavior with Actigraphy

There is strong evidence to demonstrate an inverse association between aerobic or muscle-strengthening activities with the risk of progression among adults with type 2 diabetes, explained John M. Jakicic, PhD, Professor of Physical Activity & Weight Management, University of Kansas.

“Even the intermediates are actually impacted by physical activity, things like A1C, blood pressure, body mass index, lipids, and so on. And many of these outcomes are independent of things like body weight, are independent of medications people take. So, even when you layer stuff on top of this, activity has its own effect,” he said.

Actigraphy devices are designed to detect motion, but none can capture all activity. For example, wearables on the wrist will likely fail to detect if a person is biking. Actimetry sensors wrapped around an individual’s thigh will likely not detect if a person is lifting weights.

To improve the accuracy of the data, Dr. Jakicic recommended combining actigraphy data with inclinometer and heart measurements, then applying bioinformatics and pattern recognition analytics to the combined data.

Dietary Intake and Dietary Quality

Dietary plans for patients with diabetes should be individualized based on the person’s medical history, comorbidities, treatment goals, and other factors, according to Carla K. Miller, PhD, RD, Professor of Human Nutrition, Ohio State University. Measuring dietary intake can be particularly challenging, though.

“Selecting a dietary assessment method involves compromise,” Dr. Miller said.

Historical approaches to dietary assessment include self-reporting methods such as written food records, food frequency questionnaires, and brief behavioral checklists or screeners. However, the validity of these methods often comes into question due to inaccurate recall or other variabilities associated with self-reporting.

Emerging methods like digital self-monitoring apps, image-assisted dietary assessments, and EMAs harness technology to reduce demands on a patient monitoring their diet. Yet few of these methods have been validated. Small samples are used for pilot testing, and few studies have been conducted among children, teens, or elderly populations. Some of this technology can also lead to privacy concerns.

To identify the best method for dietary assessments, Dr. Miller recommended that researchers define what they want to measure from the beginning. Differing methods may be employed based on if one is measuring energy intake versus nutrient intake. They should also consider the respondent burden and balance it with the accuracy of the nutrient database used for analysis.