Corby K. Martin
BOSTON — The multitude of weight-loss apps and activity and food trackers available for smart devices continues to grow, but the optimal choices continue to be those that will result in consistent use, according to a speaker at the American Association of Clinical Endocrinologists annual meeting.
Corby K. Martin, PhD, FTOS, director of the Ingestive Behavioral Laboratory at the Pennington Biomedical Research Center in Baton Rouge, Louisiana, stressed that food and activity trackers can often be imprecise measures for a provider trying to get a complete picture of a patient’s daily eating and activity habits; however, just the act of using these apps and monitors tends to incentivize positive behavior changes.
“The technology is less important than the methodology,” Martin said. “The devices will change. The software that analyzes the data will change. But critically, our studies and other studies have shown that it’s the way in which we utilize this objective data that drives behavior change, that drives adherence to diets and increases activity.”
Tracking food intake
The fundamental flaw in obesity research is that there is no method to accurately quantify food intake in free-living individuals, according to Martin. Self-report methods, he said, are highly inaccurate. Many studies show greater than 37% error on self-report food-intake modules, mostly due to the patients’ inability to accurately estimate portion size, Martin said.
“If you’ve ever tried to track dietary intake in free-living individuals, you are quickly humbled, because it is extremely challenging,” Martin said.
Electronic methods for patients to self-report their food intake do not overcome the limits of traditional, pen-and-paper methods, Martin said. However, food reporting is associated with greater weight loss and remains a valuable tool for clinicians.
“Even though patients inaccurately report intake on food records, the more food records they keep, the more weight they lose,” Martin said. “There is some type of accountability factor that occurs that helps people be a bit more adherent to their diet, even if those estimates are grossly inaccurate. So, there is still a benefit of keeping food records.”
However, he cautioned, “never, ever use a food record to see what patients are eating.”
Martin cited several popular self-report apps, including Fat Secret, Calorie King, LoseIt! and MyFitnessPal. Additionally, many activity tracking apps also allow patients to track food intake.
The Remote Food Photography Method and SmartIntake 3 app, developed in part by Martin and Pennington Biomedical Research Center, may help overcome the limitations of self reporting by allowing clinicians to see live-time photos of the foods people are eating, he said. The food images can be captured on any smartphone, with or without an app, and sent to the clinician via email or text in real time. Online platforms are also available to share food images, like www.meallogger.com. Clinicians, in turn, can send feedback to the patients almost immediately.
“Within a couple of seconds of seeing these pictures, you can provide basic behavioral recommendations for what is going south in their meal, and give objective feedback,” Martin said. “Closing that loop between when the behavior occurs and when the feedback occurs is critical to driving behavior change.”
Martin cautioned that remote food photography has the same flaw of other apps, in that a person may forget to take photos of their food or feel embarrassed to photograph certain meals. A text-prompting system has helped to improve uptake, he said.
“It’s important to highlight the clinical utility of these photos,” Martin said. “The clinical utility is there, even if the patient doesn’t use the method all day, every day, forever. That’s not going to happen. Patients can utilize food photography for about 7 to 12 days, and then they fatigue. We typically deploy it on Thursday through Sunday.”
Tracking physical activity
Objective physical activity monitors are now common and, unlike self-reports of food intake, they are more accurate vs. self reports of physical activity, Martin said. Many options are available, ranging in price from $10 for a basic pedometer to more than $350 for smart watches, that track metrics including steps per day, total daily energy expenditure, heart rate and activity energy expenditure. The data are frequently and automatically uploaded to an app or website.
Martin cautioned that providers should not burden patients with tracking confusing metrics, like daily energy expenditure, which can be difficult to validate, as most companies will not provide their algorithms.
“We track steps per day,” Martin said. “There are national guidelines for steps per day — 10,000 steps per day — and patients understand that. They can look at their monitor and see that they walked around the block and gained 500 steps.”
Basic pedometers remain a good and inexpensive option for patients tracking step counts, Martin said. For more sophisticated tracking, FitBit remains a popular brand of activity tracker. Waist-worn devices, like the FitBit Zip, tend to show better accuracy for step counts vs. wrist-band trackers because the algorithms are more fine-tuned, he said.
“We frequently get asked the question, which device is the most accurate?” Martin said. “I would like to argue that reliability or repeatability is probably more important than validity. If a patient is using a FitBit and on day 1 it shows they took 5,000 steps, if in reality they took 6,000 steps, that is no big deal. As long as tomorrow if they take 5,500 steps it will show 6,500 steps. We’re trying to get patients to track changes in their activity. Critically, many of the devices are quite accurate for steps.”
Apps for body
Tracking weight loss via apps may be one of the most powerful things health care providers can do clinically for patients, Martin said. Several wireless, Internet-connected scales, like BodyTrace, FitBit Aria scale and Nokia Body+, are now available to provide real-time weight-loss data.
“Body weight change reflects energy balance, which is what we are trying to quantify anyway,” Martin said. “We know though that with current methods available, we can never do that reliably ... but when you track change in body weight, that is the perfect reflection of change in energy balance.”
Dynamic models overcome fundamental problems in weight management treatment and allow clinicians to quantify dietary adherence based on observed body weight, and objectively quantify energy intake over time based on body weight, Martin said.
Martin said it is key to develop what he called ecological momentary interventions — treatment delivered to patients in their natural environment, usually via communication technology, that provides tailored, automatic feedback. In the SmartLoss app, for example, data from a wireless scale are transferred from the home environment, graphed to illustrate adherence, and sent back to the patient via smartphone with automatic feedback and clinician recommendations.
“You can have these data from the accelerometers, as well as the scale, transmitted to the internet, you can pull it from the internet, plot it on a graph and give that to the patient in real time,” Martin said.
“These approaches allow you to do more with less,” Martin said. “It allows you to intervene and help the participant only when they need it and specifically to address the problem that they have at the time. We’ve demonstrated really nicely that using these weight graphs are very helpful clinically.” – by Regina Schaffer
Martin CK. Apps and Activity Monitors and How to Use Them Clinically. Presented at: AACE Annual Scientific and Clinical Congress; May 16-20, 2018; Boston.
Disclosure: Martin reports his institution has interest in the intellectual property related to the Remote Food Photography Methods and SmartIntake app, and he is an inventor of the technology.