The heavier someone is, the less likely they are to have what many people might call a "good death" with hospice care or the chance to die at home, a new study finds. And that difference comes at a financial, as well as a personal, cost.

MORE FROM THE LAB: Subscribe to our weekly newsletter

A study by researchers at the University of Michigan Institute for Healthcare Policy and Innovation, published online today in the Annals of Internal Medicine, draws this conclusion from an in-depth analysis of records from more than 5,600 seniors taking part in a long-term health study.

The researchers looked at how the seniors' body mass index, a measure of obesity also known as BMI, related to end-of-life measures such as their use of hospice services, which provide supportive care to people in their final months of life.

The higher a person's BMI, the less likely he or she was to enter hospice. Someone who had a BMI of 40 had less than a 23 percent chance of having hospice services at the end of life, while those in the "normal" weight, with a BMI of 20, had a 38 percent chance, the researchers found.

For those seniors who did enter hospice, people with obesity spent fewer days there than their leaner counterparts.

About 60 percent of the seniors in the study died at home — an experience that most Americans say they would choose for their own deaths. That percentage dropped as BMI rose.

Elevated among people with high BMIs: the cost to the Medicare system for the care they received in their last six months of life.

"Obesity is a complex disease, and this study has shown another side of its effect on patients and the health care system. Patients with obesity are less likely to receive high-quality end-of-life care," says John Harris, M.D., M.Sc., who led the study during a fellowship at U-M and is now an assistant professor at the University of Pittsburgh School of Medicine.

A population in need

Although the study didn't examine the reasons people with obesity received less hospice care, senior author Jennifer Griggs, M.D., M.P.H., a professor at the U-M Medical School, suspects several factors are involved.

SEE ALSO: Why Does Dying Cost More for People of Color?

"For hospice teams, as for hospital and home-health teams, it can require more staff to take care of people who are obese, but Medicare hospice reimbursement is capped no matter what a person's BMI," she says. "Also, obesity may mask the signposts and changes that physicians might use to determine when it is time for a person to enter hospice."

Griggs, who treats patients with cancer at the U-M Comprehensive Cancer Center, says that even in overweight and obese people, it's possible to see the signs of cachexia, or the natural "wasting away" that occurs toward the end of life. This can include more pronounced cheekbones, collarbones and pelvic bones.

Still, physicians may not notice the gradual changes in a heavier person. A major change in body weight is also more noticeable on a leaner person.

"To refer a patient for hospice, you have to believe that they are in the dying process," Griggs says.

Harris and Griggs also say previous studies have shown a bias against people with obesity by health care professionals, especially physicians. The researchers suspect, but cannot show with the new data, that this may also play a role in decisions about end-of-life care. Griggs notes that obesity, and its role as a barrier to optimal care, isn't even mentioned in Dying in America, a landmark 2014 Institute of Medicine report.

The new results suggest that more policy attention should be focused on people with more severe obesity, Harris says.

"These patients' voices, and those of their caregivers, need to be heard," he says. "They may need extra help. They deserve the best health care and better health outcomes."

More about the study

The team compiled its results using data from in-depth surveys and Medicare records from 5,677 people enrolled in the long-running, nationally representative Health and Retirement Study. All of the participants whose records they studied died between 1998 and 2012 and were not living long term in a nursing home.

The researchers focused on Medicare claims in the last six months of the participants' lives, looking at use of hospice services and controlling for gender, race, marital status, multiple medical conditions and household assets. They also had information from surveys of the participants' loved ones about their end-of-life care. And they had all participants' last measured BMI, which is weight in kilograms divided by height in square meters.

Fifteen percent of participants were obese, with BMIs over 30; 2 percent were morbidly obese, with BMIs over 40. An additional 31 percent were overweight, with BMIs between 25 and 29.9.

In all, 38 percent of the group used hospice services.

The differences in end-of-life care translated into differences in cost. In the last six months of life for a person with a BMI of 20, the Medicare system spent about $43,000 for all types of care. For a person with a BMI of 30, the costs for that same period were about $3,500 higher.

Harris says the findings point to a larger issue that will become more urgent as the increasingly overweight American population ages.

"Many policies focus on preventing or reducing obesity in the U.S., but we will also need policies to encourage the provision of high-quality care for people with obesity," he says. "More attention should be paid to payment structures that recognize the challenges involved with health care for men and women with obesity." Harris is an assistant professor of obstetrics, gynecology and reproductive sciences at Pitt who participated in the Robert Wood Johnson Clinical Scholars Program at U-M.

Griggs, a professor of hematology/oncology in the U-M Medical School who also holds a joint appointment at the U-M School of Public Health, is leading another research effort on factors affecting end-of-life care. It's based on a partnership with the state of Michigan and Hospice of Michigan designed to look at more aspects of hospice care, including race and caregiver factors.