Short version: It’s not just the idea, but the anxiety produced by the idea, which gives rise to an increase in thoughts of death. If I tell you your worldview is stupid and you don’t care, you don’t have increased thoughts of death. But if you take my “you’re stupid and so’s your worldview” to heart and feel anxious, you’ll have increased thoughts of death. If you have increased thoughts of death, you try to defend your worldview from attack.
Longer version: Terror Management Theory proposes that when we are confronted with thoughts of death, we seek to (1) shore up our self-esteem, and (2) our worldview. For example, an atheist confronted with death can say, I won’t know I’m dead so there is no reason to fear death. A Muslim can say, I will be resurrected to Paradise, so I have no need to fear death. When I thinks about death, they can think about their response to death.

When confronted with some information which undercuts their worldview, (say, there is a god, or Muhammed was not a true prophet), research shows that the victim (or test-subject, depending upon your point-of-view) has more thoughts about death (DTA death-thought accessibility).

Since thoughts of death produce anxiety, human beings seek for ways to relieve that anxiety (anxiety being unpleasant). Researchers have noted two basic mechanisms, first were used to relieve the anxiety. The immediate response is to distract oneself or otherwise try to ignore the information). Then, after a passage of time and as thought the immediate thoughts of death fade, one begins to various “distal defenses” are brought to bear. The victim seeks to shore-up their symbolic mechanism to deal with death.

The research has primarily dealt with the thoughts of death, not the emotion of anxiety. A study published 2014 sought to examine the emotive functions.

The study sought to produce anxiety in Protestant Christian undergraduate students. They were told that the they were testing how a drink effected memory. Some of the students were told the drink contained caffeine and would them “jittery,” others were told it was a vitamin drink.

The reason for the two different drinks has to do with “attribution of arousal manipulation.” The students who drank the “caffeine” might attribute their anxiety to the drink and not to the article challenging their beliefs.

The students were given an article which challenged their religious beliefs (Jesus is the same as Krishna or Mithra or Horus). A control group read an article on the northern lights.

The next phase asked the students to complete words . So they were given coff–. Do they write “coffee” or “coffin”? The reason for this section to was both assess their thoughts of death and to give time for the “distal defenses” to engage.

The final phase as the students to evaluate their article – did it make you angry? How smart was the author?

When the students were given the “caffeine”, there was a marginal tendency to attribute their anxiety to caffeine and to have fewer “death-related” thoughts than the vitamin drink group. The students with the vitamin drink did experience more death related thoughts when having their religious beliefs attacked.

Not surprisingly, the students who read the attacking article had greater emotional response than those who read the article on the northern lights.

But since the researchers had given an introductory questionnaire on death related thoughts, they wanted to make sure that initial questionnaire did not poison their results.

They performed a very similar test. But this time they gave the students an opportunity to set bail for a prostitute. The thinking was that death-related thoughts would lead to more protection for their worldview, which would lead to higher bail amounts.

The surmise was true.

Here is what the researchers believed was significant in these tests: When the student attributed their anxiety to the caffeine they did not seek to protect their world view. It seems that when they blamed the drink for their anxiety it acted to protect them from thinking further about death.

A third test was premised upon this idea: Humans protect ourselves from thoughts of death by distinguishing ourselves from other animals. Therefore, we experience disgust when someone eats strange food, defects on the living room floor or commits incest, because it reminds us that we are animals; reminding ourselves that we are animals, reminds that we can die like animals. Therefore, we feel disgust in those circumstances.

You don’t need to take that explanation for why we experience disgust when someone decides to imitate a dog in your apartment.

The third experiment sought to determine the extent to which misattribution could apply to disgust.

And so we come to a test which I am glad I did not have to experience. The students were going to be subjected to viewing a number of gross pictures, someone vomiting, urine, feces, snot, a dirty toilet, a bloody finger. These apparently makes us think we are animals.

All the students were given an essay to read. One essay said, you’re just animal. The other essay had nothing to do with animals.

All the students were given instructions on viewing the pictures. Some were told to view the pictures carefully. Others were given specific instructions to take a “detached and unemotional attitude.” They were to be clinical and unfeeling as they examined the pictures.

After looking at the pictures, they were examined for disgust.

The students were instructed to have clinical detachment when viewing the pictures had fewer death-related thoughts after viewing them.

And so again, an increase a serious negative emotion increased one’s thought of death.

Here was the upshot:

Our findings suggest that threatening material will only increase DTA when that material is experienced as emotionally unsettling.

Webber, D., Schimel, J., Faucher, E.H. et al. “Emotion as a necessary component of threat-induced death thought accessibility and defensive compensation.” Motiv Emot 39, 142–155 (2015). https://doi.org/10.1007/s11031-014-9426-1.

What precisely takes place is unclear.

This research reminds me of some research I did in college on the grotesque in literature. There is a theory that we are attracted to disturbing things in art because it allows us to focus our existing anxiety on a point and attribute our anxiety on that artwork (rather than on some other matter which may be disturbing me.

There is an important consideration here for persuasion study. Persuasion functions by creating some sort of dis-ease, some anxiety and a proffered means of resolution. You see the car, you want the car: anxiety. You can buy the car: resolution.

If the creation of anxiety generally has a tendency to increase thoughts of death – and thus thoughts of protection of my worldview – this creates a certain complication. The research showed only a “marginal” decrease in death related thoughts when the anxiety could be attributed to the caffeine drink.

If we seek to create a powerful persuasive movement, we have the potential for creating greater anxiety and thus increased death related thoughts. An increase in death related thoughts comes along with protection of one’s worldview.

Thus, a powerful persuasive move have the wind at its back if the persuasion accords with one’s worldview. But, an attempt to make a strong persuasive move (by generating a great deal of anxiety at first) will have a headwind if that persuasive move is contrary to the worldview.

This does not mean that the issue under persuasive pressure is distinctly a facet of the worldview; only that it can be concordant or discordant with a worldview.