The Dismal Science

Is Donald Trump Experiencing Cognitive Decline?

The president’s speech has changed dramatically over the years. One writer investigates whether it’s age and stress or something else.

U.S. President Donald Trump
President Donald Trump waves before departing Newark Liberty International Airport after a weekend at Trump National Golf Club in Bedminster, New Jersey, on Sunday.

Yuri Gripas/Reuters

When inaugurated in January, Donald Trump, who turns 71 on Wednesday, became the oldest elected president. On Monday’s Trumpcast, Jacob Weisberg talks with Sharon Begley, senior science writer at Stat News. Begley has examined the change in Donald Trump’s speech and thinks there’s something happening beyond aging, stress, or rage. This interview has been edited and condensed for clarity.

Jacob Weisberg: You wrote recently about the question of Trump and cognitive decline. The piece pointed out that Trump used to be not so linguistically challenged. His language when he was younger was more complex, more sophisticated, bigger vocabulary. What’s going on here?

Sharon Begley: A number of political reporters had been saying that Trump just sounds different today than he did a number of decades ago. He has been in the public eye since the ’80s, and there’s a great deal of the historical record where you can listen to him. I pulled as many of the clips that I could.

A number of studies recently have looked at what’s called spontaneous speech—it’s not how well you read a written speech but how well you speak unscripted—and have used that as a sort of keyhole into cognitive function. So we thought this might be an interesting exercise.

There was that fascinating study that a team from Arizona State University did about Reagan: They looked at Reagan’s press conferences and compared it to the first President Bush. With Bush, you had a steady state—his language seemed to be pretty consistent over that time. With Reagan, you saw this change, which people take that as a marker of Alzheimer’s, which they knew he developed at some point. What does this point to with Trump? Is it the same thing?

There are a number of things that could be going on here, so let me describe some of the very clear changes in how Trump speaks over the years. One is vocabulary level. Back in the ’80s and ’90s, he used not terribly jargon-y words, words that you would find in crossword puzzles, but they were at a level that is notably higher than the way he speaks today.

Another thing people look at is the concreteness of words: filler words; how many ums and ahs; and the use of thing and related words—anything, something. Those are considered a marker of perhaps diminishing cognitive ability because when you can’t think of the exact, perfect, concrete word, you’ll say thing, right?

What we did here at Stat News, and what I asked the experts to help us with, was to listen to some of Trump’s spontaneous speech over the years and see what they could pick out. There was a long Charlie Rose interview that dates back to 1992, which was another presidential election year. Ross Perot was running as a third-party candidate, and Trump said the following:

Had he not dropped out of the election, had he not made the gaffes about the watchdogs and the guard dogs, and they were real bad days, he could have conceivably won this crazy election.

He began that with “Had he not”—this was referring to Perot. It seems like a little thing, but it’s fairly unusual for people to say “Had he not” rather than “If he hadn’t.” Generally, people speak with the noun before the verb, and Trump did not, so it’s a slightly more elevated syntax than usual. He also had a fairly complicated dependent clause in the middle of that utterance, and if you are going to insert what, in written language, would be set off by commas, you have to know where you’re going. It requires forethought and not losing your place as you’re speaking. The experts say that that’s another marker of cognitive ability and cognitive status.

Could it also just be the difference between being in your 40s and being in your 70s?

Let me first offer two different hypotheses before we go to cognitive decline, whether it’s age-related or pathological. This could be neither of those.

It could be intentional. Trump could be very aware of his base. He could be aware of how they like him to sound, how they like him to speak. So rather than the “had he not” construction of 1992, it could be a much simpler construction. He could totally know what he’s doing; he figured out how to win the White House, so we absolutely have to, at least as a hypothesis, give him credit for knowing how to speak in a way that is effective for him.

Another possibility is—as all of the researchers emphasized to me—if you are under stress, your spontaneous speech can deteriorate. Trump has himself described how he has been frustrated and angry, whether it’s at the courts blocking his immigration ban or at how difficult Washington is. It’s very different from being the CEO of a privately held company that does not even answer to shareholders. During one of his hundred-day interviews, he even talked about how difficult it is and how he actually liked his former life a little bit better.

So now let’s get to the key question. People keep asking me, “How do you tell the difference between Alzheimer’s and the normal, nonpathological, non-disease-related cognitive decline that comes with aging?” The simple answer is it’s either very difficult or impossible, but frankly, I’m not sure why it matters. In some cases, Alzheimer’s-related cognitive decline can be very, very quick. It can be a steep decline compared to the normal cognitive changes that come as we get older, but frankly, cognitive decline is cognitive decline, and it’s something that I think most people would not want in their leader. So, the simple answer is you can’t tell, but I think a more nuanced answer is: Is that really what we need to be focusing on?

I don’t think Trump gets any sleep. He’s up late watching Fox News and yelling at the TV, then he’s up early firing off tweets.

My understanding is that one thing that is not different is his sleep pattern. So yes, it could be that as you get older, the all-nighters that we pulled in college are a lot harder on our brains. It could be that the lack of sleep that he managed to just power through when he was a younger man is taking a toll on him in 2017. He has released only minimal medical records, so a lot of what he does is sort of behind the brick wall. But whether it’s his tweeting habits or some of the other hints that he’s dropped, he’s not getting a full seven or eight hours most nights.

You say it maybe shouldn’t matter to us so much why he’s impaired; the question should be “Is he significantly impaired?” I think people would feel very differently if it was some kind of dementia, partly because of the other things that can come along with that as it accelerates.

The key word there is accelerates, but the cognitive decline that comes with either Alzheimer’s disease or the other forms of dementia is absolutely much faster than if you just go from 70 to 71 to 72, etc. But any cause of cognitive decline in the leader of the free world is something to pay attention to and possibly to worry about. I have no idea how much his supporters focused on last year, or are focusing on this year, the age factor. Both candidates were at the older end of the spectrum, and Trump will be our oldest president. On the other hand, Jacob, there are 90-year-olds who are still clicking on all cylinders cognitively, so the kind of cognitive changes that come with aging are hugely variable.

What’s your guide about what’s affecting him?

You have to keep open a possibility that it’s intentional or emotional stress. But in speaking to the experts that I did for the story, their consensus is that something is going on. It’s not an acute emotional response; it really is an indication that there might be cognitive decline that we are seeing in him. People cannot tell whether it’s a disease, pathological, a form of dementia, or age-related, partly because, although it may seem like we’ve been in this administration forever, it’s only June. If his unscripted speech continues to decline, and if we start to see a significantly accelerated decline, then that really is a red flag that something pathological might be going on here.

Do the people who study this have any way of designating the speed of decline?

As it happens, after my story ran, another research paper was published in a scientific journal looking at football players and their spontaneous speech, mostly during Q-and-As with reporters. It analyzed their spontaneous language then correlated it with which of those players had suffered chronic traumatic encephalopathy, which is the brain damage you get from repeated head injuries and concussions.

With the worst CTE, there was much more accelerated cognitive decline, as shown by how the football players spoke. There is no consensus in which people say that if someone’s vocabulary drops from fifth grade to third grade in six months as opposed to two months, then this means something, nor that if a person stops using dependent clauses completely in some period of time. It’s breaking new ground with every study.

I think some people have qualms about this kind of analysis because it can be exculpatory. But when you pathologize someone’s issues around what they say, you tend to diminish the factors of morality, character, preparation, qualification. If you think people have little control, you can’t ultimately blame them in the same way.

That’s an interesting discussion if it were someone in your family or a friend, but when someone is in a position of such responsibility and power, it seems to me that competence, cognitive function, brain function, regardless of reason, simply have to come to the fore. I’ve also written stories describing what psychiatrists are seeing in Trump’s behavior, and there is a very clear divide between those who think that there is some mental disorder going on and others who are saying, “That’s ridiculous! You are pathologizing.” I think in the case of cognitive decline, whether you’re trying to find something exculpatory or using that in an exculpatory way, that’s a hard sell.

What do you think the reasonable remedy is here?

You remember during the campaign that when he “released his medical records,” they were—I think “skimpy” would be a fair assessment. And there was no psychological assessment, no cognitive assessment. But yes, so this is completely hypothetical, because it does not seem that he’s interested in being that transparent. On the other hand, this is something that policymakers and legislators have grappled with, because there is the 25th Amendment, which allows a way for the Cabinet to make an assessment. Not themselves—because most Cabinet officials are not physicians, let alone neurologists—but with some kind of expert input, to draw conclusions about whether the president is cognitively able to carry out the duties of his office, and there’s a mechanism for addressing that. How that would be carried out, I cannot begin to imagine, but at least it’s there in the 25th Amendment.