Childhood Trauma and Addiction: Did I Cause This?

01 The question underneath the question

When families search for the link between childhood trauma and addiction, they are usually not doing research. They are checking a verdict they have already half-passed on themselves: this happened because of something I did, or something I failed to give.

That question deserves an actual answer rather than reassurance. So here is what the evidence says, including the parts that are uncomfortable in both directions.

02 Yes — the association is real

Childhood adversity is genuinely associated with later substance use disorder, and the relationship is dose-dependent: the more adversity, the higher the risk. This is one of the more consistently replicated findings in the field, and nothing below is meant to wave it away.

In one large prospective study using objectively documented maltreatment — medical records and registers rather than adults recalling their childhoods — the crude risk of substance use disorder was more than six times higher in the exposed group.

If the story stopped there, the popular version would be right. It does not stop there.

03 The number that actually answers the question

The problem with the raw association is that families share more than a home. They share genes, and they share circumstances — poverty, illness, instability — that raise both the odds of adversity and the odds of addiction independently. A correlation between the two cannot tell you which is doing the work.

Researchers get around this with designs that hold family background constant — comparing siblings raised in the same family where one was exposed and one was not, comparing twins, following adoptees. A 2023 review pooled 34 such studies covering 54,646 people. The result is the most direct answer available:

Roughly half the apparent effect was never caused by the adversity at all. What remains is real — the authors describe it as consistent with a small causal contribution, and it held steady across different methods and different psychiatric outcomes. But it is small, and the researchers say plainly that much of the risk comes from wider genetic and environmental factors rather than from the maltreatment itself.

The same pattern appears in the sibling study mentioned above. Comparing exposed people against their own unexposed siblings — same parents, same house — the elevated risk dropped from more than six-fold to roughly three-fold. Still real. Much smaller than it first looked.

And in the background sits a figure rarely quoted to families: alcohol use disorder is approximately 49% heritable, while the shared environment — the home itself, the thing a parent actually builds — accounts for about 10% of the variation in who develops it.

04 "Feeling unloved" is not what these studies measured

This distinction matters more than any other on this page, and it is almost always skipped.

The strong causal findings above come from studies of severe maltreatment — abuse and neglect serious enough to appear in hospital records and official registers. That is not the same thing as a child feeling unloved, overlooked, or not close to a parent.

When researchers measure something closer to that softer experience — attachment security, the felt sense of a reliable bond — the association with later substance use is much weaker. Pooling 665 effect sizes across 34 samples and 56,721 people, the correlation between earlier attachment and later substance use was r = −.11. That is a small relationship. It points in the expected direction, and the same review found the path running more strongly from attachment to later substance use than the reverse — but it is nowhere near the scale the popular account implies.

So if the question is "was my home cold enough to cause this?", the honest answer is that the evidence linking that specific experience to addiction is real but weak — and considerably weaker than the evidence about severe maltreatment, which is what gets quoted.

05 What a "small effect" actually means for one family

Effect sizes describe populations, not people, and this is where the number becomes genuinely useful.

A small average effect means the exposure is a risk factor and not a mechanism. Researchers studying maltreatment make the same point directly: not everyone who experiences it goes on to struggle with substance use, and many people adapt without that outcome. Because the effect is small rather than deterministic, childhood adversity is neither necessary nor sufficient for addiction to develop — it is one contributor among several. Any account treating addiction as essentially always downstream of childhood trauma is claiming something stronger than the evidence supports.

Which cuts both ways, and both are worth sitting with. It means a difficult childhood does not sentence anyone. It also means a loving one is not a guarantee — and families who did everything they could think of are not looking at evidence of a failure they cannot find.

06 The part that is usually backwards

Families often describe isolation as the thing that started it — they withdrew, and then they used. Sometimes that is true. But the direction of travel is not what most people assume.

In a study using genetic variants to test causal direction, the clearest finding ran the other way: starting to smoke increased later loneliness, with strong statistical support. The reverse path — loneliness causing smoking — had only weak support.

An important limit, because it is specific to the substance: that same study found no clear causal link between loneliness and alcohol in either direction — not loneliness leading to drinking, and not drinking leading to loneliness. Findings about one substance do not transfer to another, and for alcohol specifically the loneliness story is not established.

What this changes practically: the withdrawal a family sees now is substantially a product of the addiction rather than only a cause of it. The person pulling away is not necessarily revealing what was missing all along.

07 So — did you cause it?

On the evidence: almost certainly not in the way you are afraid of.

The strongest causal findings concern severe, documented maltreatment, and even there the causal share shrinks substantially once genetics and family circumstances are accounted for. The gentler thing most parents are actually worried about — not enough warmth, not enough time, the wrong words in a hard year — sits in the part of the literature where effects are small. Meanwhile roughly half the liability for alcohol use disorder is heritable, and the shared home environment accounts for around a tenth of it.

None of that means the relationship was irrelevant, and this page is not an argument that nothing you did mattered. Connection is protective. It is simply not the switch, and treating it as one has a cost: it hands a parent a guilt the evidence does not support, at precisely the moment their attention is most needed elsewhere.

The more useful question is not what caused this. It is what helps now — and there the research is far more actionable, because it concerns things still in front of you rather than things behind.