How does childhood trauma relate to depression and anxiety later in life?
- Amanda Metskas
- 2 days ago
- 10 min read
Updated: 14 hours ago
What is the relationship between childhood trauma and adult anxiety and depression?
Is trauma the key to understanding mental illness?
A popular claim is that childhood trauma is the source of most (or all) mental illness.
As part of our efforts to study and understand depression and anxiety, we investigated questions like these. What we found really surprised us.
Key Takeaways
📊 Population-level patterns show strong links. People with more types of self-reported traumatic experiences in childhood (ACEs) are more likely to have mental health challenges in adulthood when you look at averages across the overall population (e.g., on average, people with high ACEs are 3x–5x more likely to have depression or anxiety as adults).
🔍 But individual prediction is weak. The relationship between ACEs and adult mental health is only weakly connected when you're predicting individual people's mental health (ACEs explain only 5%–10% of the variation in people's mental health scores).
🧠 Mental health is shaped by many factors. This means that for any individual, a lot more goes into explaining their mental health than the number of types of adverse childhood events, even though at a societal level, on average, people with greater ACEs suffer a lot more adult depression and anxiety.
We'll start by explaining what the existing research finds, and then share the results of our recent study, which tells a more nuanced story about childhood trauma and adult mental illness.
Note: If you’re keen to dig even deeper than this article, you can read the full write-up of this study, with plenty of extra details, here.
How is childhood trauma measured?
When scientists are looking to study childhood trauma, they often do so by measuring the number of different types of ‘adverse childhood experiences’ (aka ACEs) their study participants experienced growing up. This is done by surveying participants with questions like “Did a parent or adult in your home often swear at you, insult you, or put you down?” Participants are then given a score that records how many different types of ACEs they said they experienced.
Here’s a table showing different types that are commonly identified by researchers (10 in total):
How common is childhood trauma?
The first major study on childhood trauma and adult health outcomes was conducted by the CDC and Kaiser on more than 17,000 participants between 1995-1997. The chart below shows the percentage of participants reporting 0, 1, 2, 3, or 4 or more such ACEs:
Notice (in blue) that only 36% of people had 0 ACEs, meaning that nearly two-thirds of people experienced at least one of the ACEs in childhood - so it's uncommon that people make it through childhood without at least one potentially traumatic type of event. Also, notice (in red) that 12.5% of people reported experiencing 4 or more ACEs - so the severe end of the spectrum is relatively less common.
What does existing research show?
That same CDC and Kaiser study found that adults who had experienced 4 or more Adverse Childhood Experiences (ACEs) had a 51% chance of experiencing 2 or more weeks of depressed mood in the past year, compared to only a 14% chance for adults who hadn’t had any Adverse Childhood Experiences. People's odds of experiencing depressed mood in the past year as adults were 4.6x higher if they had 4 or more ACEs in childhood, after accounting for age, gender, race, and educational attainment. This suggests a remarkably strong association between childhood trauma and adult depression!
Relationship between ACE score and “Two or more weeks of depressed mood in the past year” from CDC-Kaiser study
Looking at these odds ratios, you can see that the higher the ACE score, the higher the odds ratio for depression, suggesting that the risk of depression increases the more ACEs a person has experienced. There’s a small increase in the odds of depression for 1 ACE compared to 0, somewhat higher odds ratios for 2-3 ACEs compared to 0, and a much higher odds ratio for 4 or more ACEs compared to 0.
Since that initial study, a great deal of additional research has been conducted reporting similar findings, suggesting that there are serious links between ACEs and health outcomes in adulthood.
Based on their research the CDC claims that preventing ACEs would lead to improvements in the rates of the following conditions across the U.S. population:
Source: BRFSS 2015-2017, 25 states, CDC Vital Signs, November 2019.
Although the original CDC-Kaiser study did not assess anxiety disorders, more recent studies have found similar odds ratios for anxiety disorders. Whittaker et al (2021) (full results here) found that the odds of experiencing an anxiety disorder in the last year were 3 times higher for adults with 3-5 ACEs after controlling for age, gender, race, childhood socioeconomic disadvantage, and current socioeconomic disadvantage compared to those with 0 ACEs. Another large study estimated that anxiety disorders would decline by 31% worldwide if Adverse Childhood Experiences were eradicated, based on the strong association between them (Kessler et al., 2018).
So it seems that ACEs cause adult depression, anxiety, and other health problems - right?
But wait, it’s more complicated than that!
Here’s what we found. We included the ACEs questionnaire in a study we ran, examining stressful life events. This meant we were able to attempt to replicate some of the effects mentioned above, as well as to probe the topic further. We wanted to understand what percent of the variation in a person’s depression or anxiety in adulthood can be predicted by self-reported childhood trauma.
If childhood trauma is the cause of most or all mental illness, then it should reflect a high percentage of the variance in adult mental health.
We ran single-variable regression models looking at the relationship between ACE score and three things:
Depression (measured using the PHQ-9 survey)
Generalized anxiety (measured using the GAD-7 survey)
Neuroticism-anxiety (measured using the "anxiety" subscale of the personality trait "neuroticism" from a Big Five personality test)
The results were not what we expected!
Depression: For individuals, on average, only 10% of the variance in current depression symptoms (measured by PHQ-9 score) is explained by differences in ACE score.
Generalized anxiety: Only 7% of the variance is explained by differences in ACE score.
Neuroticism-anxiety: Only 5% of the variance is explained by differences in ACE score.
Given the strong relationship between ACEs and rates of depression and anxiety in the general population found in several large studies, and the emphasis being placed on the role of childhood trauma in adult mental health, it was strange to find that only 10% of a person’s current depression score is explained by variation in the ACEs score (and less than that for the other outcomes). Given that the odds of experiencing depression over the last year is 4.6 times higher for those with 4 or more ACEs compared to those with 0, we expected to see a larger portion of the variation in a person’s current level of depression to be explained by their ACE score.
Why do we find that ACEs scores account for so little of the variability in people's depression? And if it’s the case that ACEs only explain 10% of the variation in a person’s depression score, is it really the case that reducing ACEs in the population would have the large effect on reducing the rate of depression that the CDC claims? Isn't that contradictory to their finding?
The two questions
Our research was essentially asking:
“On average, how much of individual people's current anxiety or depression is explained by their ACEs score?"
But this is subtly different from the question being investigated by the large-scale research on ACEs, which is:
“How much difference is there in a person's odds of experiencing adult depression if they have a high ACEs score rather than a lower one?"
These questions seem similar, and like the answers to them should look the same, but they are importantly different.
The main reason these results can be different is variation between individuals. Looking at the scatterplot below, you can see two main things. First, higher ACE scores do, on average, predict higher depression scores (the blue line and the red line). The blue line is a linear model of the relationship between ACES and adult depression, and the red line is a flexible fit model of the relationship between those same two variables. They both show that, at a full population level, depression scores increase as ACE scores increase.
Second, you can see that, although this relationship exists, there is a lot of variety in the combinations of ACE score and depression survey score that people have (the pattern of the black dots). If ACE scores were an extremely accurate predictor of depression survey score, the black dots would be much more closely clustered around the red or blue line.
In other words, we see that, on average, depression goes up as ACEs do, but at the level of individuals there is tremendous variability.
So, when looking at a population of people, the relationship between ACE score and depression will be clearly present, but when looking at any individual person, the ability to predict their depression from their ACE score will be quite limited because there is so much individual variation.
Thinking critically about population data
The pattern in this graph shows why our data replicates the results of the original CDC-Kaiser study, while also seeming to say something different when looked at another way. Our research findings do not contradict the claim that reducing ACEs would reduce rates of depression or anxiety in the population. However, they do shed light on the limitations of applying those claims to making predictions about the causes of depression or anxiety for any given individual.
Something that has a large average effect (across a whole population) may not explain that much of what is going on for individual people. In this case, there is a strong (average) association between ACEs and worse mental health at a population level, but ACEs scores don't do a very good job predicting which individual people struggle with mental health challenges in adulthood.
To understand this seeming contradiction better, consider the example of education and income. It is well known that more educated people make more money, on average, than people with less education. But knowing someone's education you still wouldn't do a very good job of predicting their income, because there is a ton of variability due to other factors. For instance, some people with no college degree work in lucrative trade jobs, while some people with the highest level of education take on low-paid positions at nonprofits. Similarly, ACEs are quite strongly linked to mental health challenges across the population, but they don't let you predict very well who has mental health challenges at the individual level because of so much person-to-person variability (at each level of ACEs score).
Why might ACEs not be great predictors at the individual level?
Our data also suggest that there is a lot more that goes into adult mental health than the types of adversity we experienced in childhood. This suggests that those who blame most or all adult mental health problems on childhood trauma are probably off-base, and are giving the types of childhood experiences we've had far too much credit for how we feel as adults.
On the other hand, though, measuring childhood trauma is hard, and the ACEs measure is far from perfect.
There are many factors that aren’t captured by the ACEs questionnaire. The ACE score is a count of types of adverse experiences, but it doesn’t capture differences in frequency, duration, or severity of any of those experiences. What's more, the same event might be more traumatic for one person than it is for another, based on differences in their personalities or their other past experiences. How ACEs impact a child later in life could vary due to these factors, and others. If ACEs were adjusted to take into account the frequency, duration and severity of symptoms, it’s possible that studies would find a greater link between ACEs and mental health.
It is also important to point out that it’s possible that the correlation between childhood trauma and anxiety and depression rates in adulthood is due at least in part to other variables. While studies typically control for some demographic factors, there may be other factors that are related to both higher ACEs and higher risk of depression or anxiety in adulthood that aren’t being properly controlled in these studies. For example, the CDC-Kaiser study doesn’t control for childhood poverty. It’s possible that poverty during childhood may be related to both a higher ACE score and increased risk of depression or anxiety in adulthood. Our regression analysis also didn’t control for other variables, which means that some of the relationship that we identified in our data may be better explained by other factors. While it's plausible that childhood trauma causes worse adult mental health, we have to be cautious about attributing ALL of that effect to a causal role - some of it could be correlational but not causal.
There are a number of factors other than childhood experiences that contribute to the severity of depression or anxiety - e.g. current life stressors, personality traits, coping mechanisms and skills, genetics, current treatments being used, the strength of social relationships, etc. Given this multitude of factors, it’s not obvious how much we should expect any one factor to contribute. Looked at in this way, the ACE score accounting for 10% of the variation in depression scores of adults at a point-in-time snapshot, may be quite substantial - yet, many people now speak of childhood trauma as being the primary source of adult depression or anxiety, which seems quite unlikely given our data.
Implications
Our data (though it is on a much smaller number of participants) support the claims of the larger studies that ACEs are linked to the risk of many serious problems in adulthood on a population level. This gives additional reasons (over and above the immediate harm of the experiences themselves) that, from a societal standpoint, reducing potentially traumatic events in childhood may be an important policy priority. If most (or all) of this link is causal, then reducing adverse childhood experiences could improve adult mental health.
In contrast, when focusing on treatment of adults with depression and/or anxiety, our study suggests that the role of childhood trauma may be overemphasized. We think this may be due in part to misinterpretation of what the population-level data means for individuals. Our study suggests that only 7-10% of the variability in people's individual anxiety or depression levels can be predicted from their ACE score. The other 90%-93% of variance in depression or anxiety is not explained by ACEs score, meaning that there are a number of other factors that shape how much anxiety or depression someone is experiencing (or that ACEs scores are simply not taking into account enough factors to accurately model the experiences of individuals).
This means that although the impact of childhood trauma at the population level is clear, at the individual level, there is a lot of variation in how it relates to adult mental health. While childhood trauma may be the key to some people's mental health challenges, thankfully, it is very far from a guarantee of adult mental health struggles.
This article is an abridged version of a longer piece, which you can read here. And if you’d like to learn more about anxiety and depression, you could try reading our data-based explainer essay about them, or try our free ‘What to Know About Suicide’ tool: