Good News About Noise
How do differ from our ancestors of 200,000 years ago?
I know, we have novels and dental floss and Elon Musk, but what are the fundamental differences?
Here's one: Back then, we weren't responsible for most of what terrified us. Wild animals, natural disasters, and unnamed diseases were just part of the world in which we evolved.
No doubt people victimized each other too, but the scariest things weren't our fault. Nowadays, we have only ourselves to blame for nuclear proliferation, toxic pollutants, AI deepfakes, semi-automatic rifles, etc. Even threats that used to be "natural", like wildfires and viral diseases, are getting worse thanks to the behavior of some.
This newsletter will be a lot more encouraging than it sounds thus far. Briefly, the message is that we've been scaring ourselves needlessly about the health effects of anthropogenic noise (i.e., noise created by humans and our technology).
To illustrate, I'll be discussing a recent New York Times article on the topic. I'll introduce you to a pair of important statistical concepts called mediation and moderation (no math here, I promise). Then I'll explain why the data is less alarming than depicted in the Times article and elsewhere.
We've already made the world scary enough. No need to let statistical misunderstandings make it worse.
Noise about noise
The New York Times has nearly 10 million digital and print subscribers, and a strong impact on what other news organizations cover. It matters when the Times doesn't quite get it right.
Two weeks ago, the Times ran a multimedia piece called "Noise could take years off your life..." In spite of the sensationalistic title, the article is closely reasoned and quite thorough. And, it's a scary read.
For this article, Emily Baumgaertner and colleagues visited a mix of urban, suburban, and rural neighborhoods, measured sound levels, talked with residents, reviewed the scientific literature, and interviewed more than 30 scientists. Their conclusion? Environmental noise causes stress, undermines cardiovascular health, and shortens lives.
What seems especially scary is that the noise in question is much less intense than the kind that's bad for your ears. The WHO, CDC, EPA, and other organizations concur that exposure to 70 decibels or less per day, on average, won't damage your hearing. However, studies linking noise to cardiovascular disease often examine average decibel levels in the 40s and 50s. That's barely more than conversational-level intensity.
(By the way, you can download a free smartphone app from the CDC that measures ambient noise. The app gives you lots of data (averages, maximums, etc.), along with guidelines for interpretation. These guidelines are useful, because links between noise and hearing loss are purely physical. As you'll see, the relationship between noise and cardiovascular disease has a psychological component. Ultimately, this is where we find the good news.)
Why is noise harmful?
Baumgaertner and colleagues skillfully describe the physiological changes that occur in response to noise. These changes, triggered by a small, almond-shaped brain structure called the amygdala, include increased heart rate and blood pressure, more adrenaline in the bloodstream, and other characteristics of the "fight-or-flight" response.
Unfortunately, readers with no background in a field like psychology or medicine may think that what's being described here is specific to noise. In several places, the word choice fosters this misunderstanding ("If the amygdala is chronically overactivated by noise, the reactions begin to produce harmful effects...")
Actually, the writers are describing how people respond to to any stressor. Whether it's a loud sound, a looming deadline, an angry friend, or a scary news article, we exhibit the same, generally useful physiological responses. These responses evolved to help us cope with threats, as they prepare us physically and mentally to either fight or flee.
Stress only becomes harmful when it's prolonged. If you spot a tiger in the bushes, just once, you'll get anxious, you'll calm down later, and you'll be fine. But if you're being chased by tigers every day, then even if they never catch you, your lifespan may be reduced, because the cardiovascular wear and tear increases your risk of a potentially lethal event such as a heart attack or stroke. The same goes for any other prolonged source of stress, whether it's a high-pressure job or environmental noise.
I promised good news, but I haven't gotten there yet. Prolonged noise, steady or intermittent, could indeed stress a person out and increase their chances of heart disease. But it's misleading to simply say it that way.
Statistical mediation
In the language of statistics, stress mediates the relationship between noise and cardiovascular disease.
Crudely speaking, a mediator explains why X causes Y. If noise causes heart disease, it's because noise causes stress, and stress affects the heart. Researchers use statistics to measure these relationships. The results are presented in diagrams like this:
The arrow from noise to heart disease tells you that noisier environments predict higher rates of heart disease. That's what many studies show, and it's the basis for the scary Times article.
However, that arrow doesn't tell you why noise is linked to heart disease. The other arrows do that.
Specifically, the arrow from noise to stress suggests that more noise causes greater stress. The arrow from stress to heart disease suggests that more stress increases the risk of heart disease.
In an actual study, a number accompanying each arrow would tell you the strength of the relationship. This simple diagram merely emphasizes that noise contributes to heart disease by increasing stress. It doesn't contribute in some other way. For example, sound waves don't ordinarily damage your heart by moving through your chest. (If you're a stats person, you'll see that there's no need to consider partial mediation effects.)
Of course, if loud noises regularly wake you up at night, then sleep deprivation may promote heart disease too. But the focus of the Times article, and the pertinent studies, is on the role of stress.
Stress is subjective
There's something missing from the diagram I just showed you. The diagram assumes – as the Times journalists do – that more noise creates greater stress. That's not the case for everyone.
Once upon a time, psychologists proposed that life events do differ in how inherently stressful they are. Each event was assigned a particular score. When you completed a survey, you'd note which events you've experienced, then the researcher would sum up your scores and calculate how stressed you are.
This approach assumes that stressors are objective – i.e., the same stressor will be experienced the same way by different people. For example, on the Holmes-Rahe Stress Inventory (once one of the most widely-used scales), being fired from a job is worth 47 points, while pregnancy is rated slightly lower at 44 points.
Most psychologists now question this approach. Consider a 17-year-old who just learned she's pregnant, versus a 67-year-old man who just got fired because he's planning to retire next month and has already begun tuning out. Do you really suppose the man experiences more stress than the 17-year-old girl?
Nowadays, most psychologists would say: Let's not suppose anything. Stress should be defined subjectively. The same event might cause much stress for one person, but little or no stress for another.
Exactly the same thing can be said for noise. Thus, our original diagram needs another variable:
That new box "Tolerance for noise", plus the arrow sticking out of it, illustrates that in order for noise to stress you out, you have to consider it unpleasant. If it doesn't bother you, you won't get stressed.
The original diagram simply presents an arrow linking noise to stress, meaning that more noise directly creates more stress. This is the assumption underlying the Times article and others like it.
In this new diagram, more noise doesn't necessarily create more stress. It might for some people, but not others, depending on their tolerance for it.
One more statistical concept before I get to the practical implications.
Statistical moderation
In the new diagram, "tolerance for noise" is known in statistics as a moderator variable, because it affects the strength of the relationship between two other variables (noise and stress).
We all know what moderator variables are, even if we don't use the term. For instance, we say that one person eats like a horse but never gains weight. Another person has an extra french fry with dinner and the next morning they're a pound heavier.
In this example, metabolism is the moderator variable. Eating is related to weight, but metabolism moderates the relationship. As a result, for some people that relationship isn't as strong as it is for others.
Analogously, some people don't care much about ambient noise. A slight increase wouldn't bother them either. For others, noisy environments are stressful, and more noise would make things worse.
I'm not just speculating here. There's lots of evidence that subjective perceptions of noise moderate the relationship between noise and stress. I'm calling this moderator "tolerance for noise", because researchers disagree about the extent to which it's purely psychological (noise annoyance) or whether physiology also plays a role (noise sensitivity).
The key point is that people differ. You can't just draw an arrow from noise to stress. Even if that relationship is significantly greater than zero, it's not very strong, because it's based on a glomming together of very different sorts of people.
Why is this important?
Most of the studies reviewed by the Times did not consider subjective response to noise.
The studies rely on highly sophisticated measurements of decibel levels. They look at averages, maximums, and intermittency over long periods of time. But they don't control for, or otherwise take into account, how people feel about noise. And yet, it's clear that individuals differ a lot in how stressed they get in noisy environments. (More than seventy years of stress research tells us that people differ a lot in how stressful they find anything.)
And here's the punch line: The effects of noise on cardiovascular health aren't strong. Those effects are significantly greater than zero, but they could be very easily explained by differences in subjective responses to noise. Meanwhile, cardiovascular disease is much more strongly predicted by the usual suspects (genetics, unhealthy lifestyles, stress from all sources, etc.)
My point is not that anthropogenic noise is harmless. Rather, with respect to cardiovascular health, it's only harmful if it stresses you out, in which case it functions like any other stressor. The actual extent of harm is likely to be small, in spite of the sensationalistic and somewhat misleading statistics that have appeared in the news. In the Appendix, I provide a detailed example.
Conclusion
Noise can damage your hearing. This has nothing to do with how you feel about noise. On the other hand, connections between noise and cardiovascular disease depend a lot on those feelings.
Here are three practical implications for cardiovascular health:
1. If you don't mind environmental noise, then you don't need to worry about living or working in a noisy place. The research covered by the Times article doesn't apply to you.
2. If you're occasionally bothered by environmental noise, you probably shouldn't worry about living or working in a noisy place either. Although the noise will occasionally disturb you, humans have evolved to cope with occasional stress. (Admittedly, it's hard to define "occasional.")
3. If you're often bothered by environmental noise, you should be concerned. However, any source of stress in your life should concern you, and prompt you to seek relief. In the case of noise, you should also remember that the cardiovascular effects are relatively small. Diet, activity levels, and smoking/drinking behavior, not to mention other sources of stress, will almost surely have a much larger impact.
In the end, it seems irresponsible and needlessly scary to tell people that environmental noise is bad for the heart. The message needs to be that noise may be a problem only if you often feel that way about it. This message is particularly important for people who hate noise but can't afford to relocate to a quieter part of town, or install a shrubbery wall outside their residence. For these people, the options include earplugs, noise-cancelling headphones (some highly-rated brands sell for less than $50), heavy curtains, and so on. Some experts recommend deliberate exposure to loud noises for brief periods, accompanied by relaxation strategies, in order to acquire greater tolerance for noise.
And so, in a quiet voice (around 40 decibels), I want to thank you for reading!
Appendix: Needlessly scary news
Some of the most dramatic findings described in the Times article come from a 2022 study focusing on more than 4 million Swiss adults over a 15-year period. I'll use one of the findings to illustrate why the data is considerably less scary than the article makes it sound.
One of the strongest effects of noise in this study was that beginning with 35 decibels (think of a quiet room), every increase in traffic noise of 10 decibels, on average, was associated with a 4.3 percent increase in risk of death by heart attack.
1. That 4.3 percent figure, taken out of context (as in the Times article) is slightly misleading, because the effects on risk level off at around 65 decibels. In other words, from 65 decibels on, there's no increase in risk.
2. The 4.3 percent figure is an estimate of increased risk, over and above whatever risk for heart attacks people already experience. Unfortunately, the risks people already experience could not be adequately measured, much less controlled for. For example, the researchers had no data on lifestyle variables (diet, smoking, drinking, and activity levels) that are strongly predictive of cardiovascular events. Thus, alternative interpretations of the data can't be ruled out. Here's one: Among the urban poor, those who live in the noisiest environments (e.g., between a freeway and a busy road) are relatively inactive, because spending time outside is unpleasant and there aren't many places to walk, jog, or bike. It's not the noise per se that increases their risk of heart attacks, but rather the inactivity imposed by their immediate surroundings.
3. The authors of the study acknowledge that not all prior research demonstrates a relationship between vehicular noise and mortality from cardiovascular events. This discrepancy in the literature is not mentioned in the Times article.