Some significant expenses decline as we age: Most mortgages are eventually paid off, and ideally children grow up and become self-supporting.

But health care is one area in which costs are almost certain to rise. After all, one of the original justifications for Medicare — which kicks in at age 65 — is that older people have much higher health care needs and expenses.

But there are a few common misunderstandings about health costs when people are older, including the idea that money can easily be saved by reducing wasteful end-of-life spending.

Medicare will not cover it all

Half our lifetime spending on health care is in retirement, even though that represents only about 20 percent of a typical life span. Total health care spending for Americans 65 and older is about $15,000 per year, on average, nearly three times that of working-age Americans.

Don’t expect Medicare to provide complete protection from these expenses.

Traditional Medicare has substantial gaps, leaving Americans on the hook for a lot more than they might expect. It has no cap on how much you can pay out of pocket, for example. Such coverage gaps can be filled — at least in part — by other types of insurance. But some alternatives, such as Medicare Advantage, are not accepted by as many doctors or hospitals as accept traditional Medicare.

On average, retirees directly pay for about one-fifth of their total health care spending. Some spend much more. One huge expense no Medicare plans cover is long-term care in a nursing home.

Over half of retirement-age adults will eventually need long-term care, which can cost as much as $90,000 per year at a nursing home. Although most who enter a nursing home do not stay long, 5 percent of the population stays for more than four years. You can buy separate coverage outside the Medicare program for this, but the premiums can be high, especially if you wait until near retirement to buy.

Medicaid plays a bigger role

Although Medicare is thought of as the source of health care coverage for retirees, Medicaid plays a crucial role.

Medicaid, the joint federal-state heath financing program for low-income people, has long been the nation’s main financial backstop for long-term care. Over 60 percent of nursing home residents have Medicaid coverage, and over half of the nation’s long-term care is funded by the program.

That is not because most people who require long-term care have low incomes. It is because long-term care is so expensive that those needing it can frequently deplete their financial resources and then must turn to Medicaid.

A recent working paper from the National Bureau of Economic found that, on average, Medicaid covers 20 percent of retiree health spending. The figure is larger for lower-income retirees, who are more likely to qualify for Medicaid for more of their retirement years.

Hard to predict death

A widely held view is that much spending is wasted on “heroic” measures taken at the end of life. Are all the resources devoted to Medicare and Medicaid really necessary?

First, let’s get one misunderstanding out of the way. The proportion of health spending at the end of life in the United States is lower than in many other wealthy countries.

Still, it is a tempting area to look for savings. Only 5 percent of Medicare beneficiaries die each year, but 25 percent of all Medicare spending is on individuals within one year of death. However, the big challenge in reducing end-of-life spending, highlighted by a recent study in Science, is that it is hard to know which patients are in their final year.

The study used all the data available from Medicare records to make predictions: For each beneficiary, it assigned a probability of death within a year. Of those with the very highest probability of dying — the top 1 percent — fewer than half actually died.

“This shows that it’s just very hard to know in advance who will die soon with much certainty,” said Amy Finkelstein, a Massachusetts Institute of Technology economist and an author of the study. “That makes it infeasible to make a big dent in health care spending by cutting spending on patients who are almost certain to die soon.”

That does not mean that all the care provided to dying patients — or to any patient — is valuable. Another study finds that high end-of-life spending in a region is closely related to the proportion of doctors in that region who use treatments not supported by evidence — in other words, waste.

“People at high risk of dying certainly require more health care,” said Jonathan Skinner, an author of the study and a professor of economics at Dartmouth College. “But why should some regions be hospitalizing otherwise similar high-risk patients at much higher rates than other regions?”

In 2014, for example, chronically ill Medicare beneficiaries in Manhattan, N.Y., spent 73 percent more days in the hospital in their last two years of life than comparable beneficiaries in Rochester, N.Y.

“There absolutely is waste in the system,” said Ashish Jha, director of the Harvard Global Health Institute. But, he argues, waste is present throughout the life span, not just at the end of life: “We have confused that spending as end-of-life spending is somehow wasteful. But that’s not right because we are terrible at predicting who is going to die.”

Of course, beyond any statistical analysis, there are actual people involved and wrenching individual decisions that need to be made.

“We should do all we can to push waste out of the system,” Jha said. “But spending more money on people who are suffering from an illness is appropriate, even if they die.”