Is depression biochemical?

By Martin E.P. Seligman, PhD
APA President

There is popular belief that unipolar depression is a biochemically caused illness. I am quite open to this possibility, but what bothers me most about this belief is the actual evidence for biochemical, as opposed to cognitive, causation of depression. Let?s look at this with somewhat more rigor than is usual in presidential columns.

Consider two propositions:

B. Unipolar depression can be caused by some specific biochemical state (B) (the best researched being serotonin deficits).

C. Unipolar depression can be caused by some specific cognitive state (C) (the best researched being pessimistic explanatory style).

Here is a brief, and nonexhaustive list of six kinds of evidence I think are relevant to establishing a strong causal link between a biochemical or cognitive state and a psychological problem. I will also give my own (conservative) evaluation of the actual state of the evidence using a Michelin Guide 0?3 stars approach (but please don?t ask me to send you the complete list of citations until after my term of office is over).

*** = Lots of good (experimental and longitudinal) evidence

** = Some good (experimental or longitudinal) evidence

* = Some not so good evidence

0 = No evidence

First, global treatment:

B. Some biochemical treatment which involves changing B relieves depression.***

C. Some cognitive treatment which involves changing C relieves depression.***

Second, specific treatment ingredient:

B. The specific improvement of B during treatment relieves depression.*1/2

C. The specific improvement of C during treatment relieves depression.**1/2

Third, experimental manipulation:

B. Increasing or decreasing B modifies depression.*

C. Increasing or decreasing C modifies depression.** 1/2

Fourth, longitudinal measurement (prospective involving causal modeling):

B. Changes in B go on to change depression.*

C. Changes in C go on to change depression.**

Fifth, laboratory models of depression in animals and subclinical human populations manipulating B and C change the depressive analogue and the subclinical symptoms of depression in the appropriate way:

B. *

C.**

Sixth heritability of the specific mechanism, since unipolar depression is around 0.25 heritable based on twin studies and adoptive studies:

B. Some specific biochemical mechanism proposed to cause depression is heritable. 0

C. Some specific cognitive mechanism proposed to cause depression is heritable.**

Overall then, out of 18 possible stars, biochemical causation gets 7.5 stars, while cognitive causation gets 14 stars. So the actual evidence is weak that any known biochemical state can cause unipolar depression and moderately strong that at least one known cognitive state, pessimistic explanatory style, can cause unipolar depression.

But CNN, Newsweek, most managed-care organizations, most psychiatrists and consequently the American public, think otherwise. Is there a double standard about evidence here? Why?