Three years ago, a doctor told me to lose abdominal fat. Not weight — fat. My BMI was fine, but I had borderline high cholesterol and one mildly elevated liver enzyme. She suggested I invest in a smart scale. I ignored her vegan diet recommendation, but I did start a multi-year journey of frustration with body composition tracking.
Here’s what happened. I started using smart scales, which use bioelectric impedance analysis (BIA) — a weak electrical current sent through your body, with algorithms estimating fat, muscle, and bone based on resistance. Then I got a DEXA scan, the clinical gold standard. Then an InBody scan at my sports clinic. None of these devices ever agreed.
My most recent readings, all within two weeks: Withings Body Smart said 27.4% body fat. Twin Health said 28.8%. Withings BodyFit said 34.1%. DEXA said 39.3%. InBody said 44.4%. That’s a 17-percentage-point spread. My visceral fat estimates ranged from 1.9 to 14 on a 1-to-20 scale.
So which one was right? Honestly, it doesn’t matter that much. The accuracy of these devices matters far less than their consistency. DEXA scans have a clinical margin of error of 1-2%, but that depends on the facility, machine calibration, and your hydration. The first scan takes stock of where you are. Subsequent scans show whether you’re progressing.
What mattered for me was that every device told the same story: body fat and visceral fat went down, lean mass stayed flat or increased slightly, bone density held steady. The trend was the truth, not the number. If you’re tracking health metrics, pick a method, stick with it, and watch the direction — not the digits.
