Sophisticated deep learning methods of analysis are used to predict mortality risk in individuals but can be logistically difficult to perform and are expensive. Coming up with a less intrusive and more cost-effective way to determine biological age is needed.
This is where wearable technology steps in, such as fitness trackers and smartphone apps, and this is what a group of researchers took a look at and discovered that physical activity records are a good predictor of estimating our biological age.
There’s a difference between biological age and chronological age. For example, as pointed out in this study, the life expectancy of an individual with diabetes may be reduced by as much as 8 years, according to some studies.
Other conditions that can reduce life expectancy are hypertension, obesity, and other chronic conditions.
In other words, we may be older (or younger) than what the calendar tells us depending upon our health. That’s not a new discovery.
In a new study, appearing in the journal Scientific Reports, three biological age models were compared against the physical activity records from a 2003-2006 National Health and Nutrition Examination Survey (NHANES). By combining physical activity records with so-called “deep learning” methods, mortality risk prediction was enhanced and outperformed other methods.
The problem with many of the current methods is that they don’t take into consideration physical activity.
A better method may be to use those sophisticated deep learning technologies combined with data from wearable sensors.
Data collected from smartphone apps and fitness trackers can easily be stored in “the cloud” for analysis by artificial intelligence and physicians. Such methods would produce “real time feedback to patients and care providers.”
The most accurate methods involve biochemical or genetic profiling but is “logistically difficult and expensive”.
Using “low-power, compact sensors, based on micro-electromechanical systems” as found in inexpensive fitness trackers is a good alternative because the data obtained is done so in an unobtrusive manner and is easily accessible.
According to the study, by analyzing just one week of an individual’s physical activity data “can be used to produce digital biomarkers of aging and frailty.”
So, by just looking at how physically active I was this past week I can assume that I won’t be breaking any age records if I keep this pace up. I need to step it up!
One thing that popped in my head when reading the study was what was the cart and what was the horse in determining physical fitness and biological age.
I would assume that healthier people would naturally be more physically active and those with chronic health conditions that limit physical activity would be less physically active.
Does health predict physical activity levels or does physical activity levels predict health?
Those that are less physically active may be causing the health conditions that reduce their life expectancy in the first place and some health conditions can be prevented, improved, or reversed with physical activity such as diabetes, hypertension, and obesity.
It appears that the cart and the horse can be switched depending upon lifestyle choices.
But back to the study…
An interesting finding in the study was that age, determined by Principle Components Analysis (PCA), begins to really show a correlation with physical activity in those 40 years in age and older.
That makes sense because health conditions related to physical inactivity usually start to rear their ugly heads in our 40s and 50s, such as high blood pressure and diabetes. Our bad habits catch up with us by mid-life.
The important thing to take away from this study is that introducing physical activity into the equation for mortality prediction improves accuracy. Simply put, physical activity is a good indicator of how long an individual can expect to live.
It’s also important because it introduces a much more cost-effective and convenient way for physicians and individuals to monitor their health by being aware of their physical activity levels determined by data collected from wearable devices.
So, your Fitbit or your Garmin or whatever brand fitness tracker you use to estimate your daily steps and activities can let you know whether you need to increase your physical activity levels to possibly increase your life expectancy.
That 10,000 steps a day goal is still considered a fairly good one to meet. It’s a generic metric but it’s a goal people can understand and meet. Substituting or adding in other types of physical activity is also important.
As wearable technology improves and applications get smarter at determining our true well-being we’ll have a very good picture of predicting what our life expectancy may be, assuming we don’t get run over by a bus or get struck by lightning.
References
Pyrkov TV, Slipensky K, Barg M, Kondrashin A, Zhurov B, Zenin A, Pyatnitskiy M, Menshikov L, Markov S, Fedichev PO. 2018. Extracting biological age from biomedical data via deep learning: too much of a good thing? Scientific Reports [accessed 2018 Mar 31]; 8. doi:10.1038/s41598-018-23534-9
