For nearly 200 years, doctors have relied on a formula called the “Body Mass Index” (BMI) to determine whether their patients are within a healthy weight range or beyond.

A new study suggests that this approach may be insufficiently accurate for measuring obesity. Researchers found that roughly half of women and a fifth of men who are considered normal-weight according to their BMI numbers should in fact be classified as obese when tested for percentage of body fat in relation to lean muscle mass.

The authors of the study report, Dr. Nirav R. Shah, Commissioner of Public Health in New York City and Dr. Eric Braverman, a professor at Weill Cornell Medical School, said that the common reliance on BMI readings was misleading both the experts and the public about the true proportions of the current obesity crisis.

Over a third of adult Americans are classified as obese based on their BMI measures, according to the Centers for Disease Control and Prevention (CDC). But because BMI does not distinguish between fat and muscle, many people with a normal BMI may carry dangerously high amounts of body fat.

“Fat causes heart disease, cancer, menstrual problems, depression, anxiety and a host of other medical problems,” said Dr. Braverman. “Without an accurate measurement of body fat, millions of people don’t know whether they are at risk for obesity-related diseases.”

Instead of relying solely on the BMI formula, which is weight divided by height squared in kilograms and centimeters, or weight divided by height squared in pounds and inches multiplied by 703, Shah and Braverman propose to take a test called dual-energy X-ray absorptiometry (DEXA), which is commonly used to evaluate bone density in women to check for signs of osteoporosis. Unfortunately, that kind of test is costly and may not be feasible for use as a routine procedure.

As an alternative, the doctors suggested measuring leptin levels in the body. Leptin is a hormone secreted by fat cells. Testing leptin levels is simpler, less expensive and focuses on the core of weight problems, which is body fat.

Other easy measures that can be done by anyone at home are called waist and hip circumference and waist-to-hip ratio. But both of these only concern a person’s weight, not the amount of fat he or she may be carrying.

“Without knowing how much fat [people] have, you can’t really save them from illness. It is the number one predictor of who’s going to live or die,” said Dr. Braverman.

Based on their findings, Shah and Braverman suggested to lower the BMI markers for obesity from 30 to 24 for women and 28 for men. Under the current standards, BMI classifies millions of Americans as healthy when they actually should be concerned about their weight.

Commenting on the study, Dr. William O’Neill, a professor of cardiology at the University of Miami Miller School of Medicine who was not involved in the research, called the findings “alarming.” “Traditionally, we have used BMI. But this study tells you that BMI really underestimates how many are obese.”

Shah and Braverman agree that this may be a game-changer. “We may be further behind than we thought in addressing the nation’s crisis of obesity,” they said.

Timi Gustafson R.D. is a clinical dietitian and author of the book “The Healthy Diner – How to Eat Right and Still Have Fun”®, which is available on her blog, “Food and Health with Timi Gustafson R.D.” and at amazon.com. You can follow Timi on Twitter and on Facebook.