1. The programme at a glance
Indication. A major depressive episode in an adult, mild to severe, single or recurrent, with or without concurrent medication.
Reference model. Beck and colleagues' cognitive therapy of depression, and behavioural activation in the form systematised by Martell, Dimidjian and Herman-Dunn. The protocol follows the order the data support: behavioural first, cognitive next, beliefs last.
Format. Sixteen individual sessions of 50 minutes: weekly until session 10, then fortnightly. Two booster sessions, at one month and at three months.
Target mechanism. Not to lift mood directly, but four changes: to restart behaviours that produce reinforcement, to interrupt rumination, to undo the interpretations that maintain withdrawal, and to make the patient able to recognise and stop the next episode.
| Session |
Focus |
Session output |
| 1 |
Assess, measure, set the frame |
Record installed, risk assessed |
| 2 |
The formulation, and the model |
Written formulation of the case |
| 3 |
Activation: observing |
One week of activities recorded |
| 4 |
Activation: scheduling |
Three activities planned |
| 5 |
The obstacles to activation |
Analysis of one failed schedule |
| 6 |
Rumination |
Signal and postponement installed |
| 7 |
Spotting thoughts |
Three-column record |
| 8 |
Putting thoughts to the test |
Five-column record |
| 9 |
Behavioural experiments |
One prediction tested |
| 10 |
Problem solving |
One problem worked in six steps |
| 11 |
Sleep, the body, substances |
Fixed rising time |
| 12 |
Relationships, and what has gone out |
Two contacts scheduled, one resumed |
| 13 |
Rules and beliefs |
One rule identified and loosened |
| 14 |
Self-criticism |
Written answer to the critical voice |
| 15 |
Consolidating |
Relapse sheet begun |
| 16 |
Review and relapse prevention |
Written plan in the patient's hand |
The tools in the last column are presented in the corresponding session: the daily record in session 1, the signal and postponement of rumination in session 6, the three- and five-column records in sessions 7 and 8.
What the patient takes away. Seven printable worksheets, listed in section 36 and downloadable from this page: my record, my days, my rumination, my thoughts, my experiments, my safety plan, my relapse plan.
What sets this programme apart from supportive work. Supportive work accompanies the episode. This programme does three things more: it restarts behaviours before the wish to act returns, it treats rumination as a behaviour rather than as a thought, and it explicitly prepares for the next episode — because there will probably be one.
2. Before you begin
The question that comes first: which depression is this?
Three decisions are taken before the first procedure, and they are not taken in one session.
Is this a major depressive episode? A fortnight of low spirits after a break-up is not one. Section 5 gives the elements, section 7 gives what must be ruled out.
Has there been a hypomanic or manic period? This is the most important question in the whole manual, and it is asked of every depressed patient, without exception. A bipolar disorder treated as unipolar depression exposes the patient to a switch, that is, a shift into a manic or hypomanic phase, which the antidepressant can promote; and the diagnosis is missed in a substantial proportion of cases because nobody asks. See section 7.
What is the suicide risk, now? Section 10 is devoted entirely to it, and it is not optional.
What this programme does not treat
It does not treat a grief that is running its course. The distinction is developed in section 5 and taken up again in section 7; it is delicate, and it is not settled by duration alone.
It does not treat bipolar depression. Behavioural and cognitive work has its place there, but within a different framework, with a mood stabiliser and close attention to the sleep-wake rhythm.
It does not, on its own, treat a severe depression with psychotic features, major retardation, or malnutrition. These are situations where psychiatric opinion comes first, and sometimes admission.
It does not replace medical assessment. A physical work-up is part of the assessment, and the question of medication belongs to a doctor.
How to use it
Read the whole manual before the first session, in particular sections 7, 10, 12 and 32: the differential, suicide risk, medication, and the conduct of activation. These are the four places where things go wrong.
Each session is described on the same frame: the aim, the run-through in steps, what you say, the common errors, and the criterion for moving on.
One warning particular to this disorder. Your patient will arrive saying that this is pointless, that he has not done the exercises, and that you are wasting your time. This is not a refusal of care: it is the symptom speaking. How you receive that sentence in session 3 often decides the rest of the treatment.
3. The clinical picture
What patients describe
They rarely describe sadness first. They describe an extinction: no more desire, no more taste for anything, no more impetus, and a fatigue that rest does not repair.
The phrases that recur: "I don't feel like anything", "I have to force myself for everything", "I don't recognise myself", "I'm a burden to the people around me", and "I can't see how this could ever get better". That last one is at once a symptom, a treatment target, and an element of suicide risk.
They also describe an inversion of the usual order: they are waiting to feel like it before doing it. That is exactly the opposite of what treatment asks, and the sentence is worth writing down verbatim in session 1 — it will be used in session 2, then in session 4.
The three groups of symptoms
Mood and anhedonia. Sadness, but above all loss of interest and pleasure. Anhedonia is often more informative than sadness: many patients do not say they are sad, they say they are empty, switched off, or they say nothing at all.
The somatic side. Sleep — initial insomnia, early waking, or hypersomnia —, appetite, weight, energy, retardation or agitation, reduced libido, pain.
The cognitive side. Concentration, indecision, slowed thinking, self-deprecation, guilt, pessimism, thoughts of death. Complaints of poor memory are frequent and most often stem from the depression rather than from dementia — an important point in older adults, in whom the two can coexist (section 7).
What maintains the episode
This is the useful part of the picture, because it is the part one treats.
Withdrawal. The patient does less, which leaves him fewer opportunities for reinforcement, so less desire, and he does even less. This loop is the protocol's main target.
Avoidance. Not laziness: a flight from what has become costly — people, work, tasks, and sometimes the bed itself becomes a refuge.
Rumination. Hours spent on "why is this happening to me", "what is wrong with me". It gives the feeling of looking for a solution and it worsens mood measurably.
Interpretations. The patient reads himself, others and the future negatively. He attributes his failures to an internal, global and stable cause — to himself, to everything he is, and for ever —, and he disqualifies anything that goes well.
Real consequences. Sick leave, conflict, debt, isolation — which in turn feed the episode. Section 24 deals with this point: some dark thoughts are accurate, and what is accurate is solved rather than restructured.
What depression is not
It is not sadness. Sadness is a normal emotion, directed at an object, that varies with circumstances. Depression is a lasting state that extinguishes responsiveness itself.
It is not a lack of willpower. This is the most important sentence in the feedback, and it can be demonstrated: will is precisely one of the functions the disorder impairs.
It is not a flaw of character, and it is not a strength of character to be admired either. Both discourses do harm.
It is not always a signal. Some episodes have an obvious biographical meaning, others arise in a life that is going well. Insisting on finding a cause in the latter produces guilt and nothing else.
Epidemiology, in four useful points
Depression is one of the commonest mental disorders, with a lifetime prevalence of the order of fifteen to twenty per cent, and a yearly prevalence around five per cent.
It is about twice as common in women, with a gap that appears in adolescence.
It is recurrent: after a first episode the risk of recurrence is high, and it rises with the number of previous episodes. This is the central argument for relapse prevention work. In this manual, "relapse" refers to any return of depression; strictly speaking, a relapse is a return of symptoms before recovery from the episode — a remission (section 9) sustained for several months — and a recurrence is a new episode after recovery.
And a substantial proportion of episodes are neither diagnosed nor treated, particularly in men, in older adults, and when the complaint presents in somatic form.
4. The model that guides this programme
Two complementary models, to be explained to the patient
This is the most profitable explanation in the programme, and it takes twenty minutes in session 2, with a sheet of paper.
The behavioural model. An event, a loss, exhaustion reduce activity. Reduced activity removes the sources of reinforcement — pleasure, accomplishment, contact. Lost reinforcement further reduces desire and activity (Lewinsohn, 1974). Depression is not only caused by what happened: it is maintained by what the person has stopped doing.
The practical consequence is counter-intuitive and must be said in so many words: action comes before the wish to act. One does not wait to feel like it in order to act; one acts, and the wish returns afterwards, partly and slowly.
The cognitive model. Within the episode, interpretation is distorted in three directions: about oneself, about the world, about the future — this is Beck's cognitive triad. These interpretations are not lies the patient tells himself: they are plausible readings, produced by a system that filters. During the episode, attention and memory mainly retain what confirms the negative. The interpretations are treated by putting them to the test, not by contradicting them.
What these two models imply
They give the programme's four principles, to be written down with the patient.
We start with behaviour. Because it is feasible when one can no longer think, because the behavioural component carries a large share of the effect, and because a patient who has obtained a concrete result becomes capable of cognitive work.
We act before feeling like it. Motivation is not the condition of starting, it is its consequence.
We treat rumination as a behaviour. Not as content to be corrected: as something one does, which has triggers, a duration, and which can be interrupted.
We put thoughts to the test rather than replacing them. Including when they turn out to be true.
What the model explains to the patient
Why he often feels worse in the morning. There is the circadian cycle, the body's twenty-four-hour clock, and the fact that nothing has yet been done that day.
Why rest does not repair. Because the fatigue of depression is not a sleep debt, and because prolonged rest removes still more reinforcement.
Why he thinks about the same thing endlessly without finding a solution. Because rumination is not problem solving: it is about causes and meanings, not about actions.
Why he judges himself so harshly. Because the filter is in place, and because the same filter prevents him from seeing that it is in place.
Why things are better when he is busy and worse when he is alone at home. That is the behavioural model, observed from the inside.
What the model implies you should not do
Do not wait for desire to return.
Do not aim at sadness first. It moves last. What moves first is the number of activities and the sense of accomplishment.
Do not reassure. "You'll see, it will get better" is probably true and completely inoperative.
Do not argue with negative thoughts head-on. You ask for evidence; you do not supply counter-arguments.
5. The DSM-5-TR criteria, reformulated
The criteria below are a reformulation in our own words, for reference. They do not replace the manual: refer to the DSM-5-TR (American Psychiatric Association, 2022) for the letter of the text and the application notes.
A major depressive episode involves the following.
At least five manifestations present during the same two-week period, representing a change from previous functioning, and among which must figure either a depressed mood or a loss of interest or pleasure.
The manifestations concerned: depressed mood for much of the day, nearly every day; markedly diminished interest or pleasure in almost all activities; significant change in weight or appetite; insomnia or hypersomnia; agitation or retardation observable by others; fatigue or loss of energy; feelings of worthlessness or excessive guilt; difficulty concentrating or indecisiveness; recurrent thoughts of death, suicidal ideation, a plan or an attempt.
Clinically significant distress or impairment of functioning.
No attribution to a substance or a medical condition.
No history of mania or hypomania, which would point to a bipolar disorder: this condition applies to the diagnosis of major depressive disorder, not to the episode, which can also occur in a bipolar disorder.
The specifiers that change management
The DSM-5-TR allows one to specify severity, the presence of psychotic features — which require psychiatric opinion —, of mixed, melancholic or atypical features, of anxious distress, peripartum onset, a seasonal pattern, and a single or recurrent episode.
Mixed features refer to a few hypomanic symptoms mixed in with the episode (see section 7); melancholic features, to profound anhedonia with morning worsening and early waking; atypical features, to a mood that remains reactive to events, with hypersomnia and increased appetite. Anxious distress is marked tension and worry during the episode.
Two of these matter particularly here. Anxious distress is associated with slower response and higher suicide risk. And the recurrent character governs the whole final part of the programme.
The most debated point, and what you need to know about it
The removal, in DSM-5, of the bereavement exclusion — the DSM-IV rule that set the diagnosis aside in the early period following a death — was and remains debated. The authors' argument is that grief does not protect against a depressive episode and that the exclusion deprived people of care they needed. The critics' argument is that it exposes normal sorrow to medicalisation.
In practice, do not decide on duration. Look at the form: in grief, the pain comes in waves, it is tied to the person who died, self-esteem is generally preserved, and pleasure remains possible between the waves. In a depressive episode, mood is continuous, anhedonia is global, and self-deprecation is aimed at the person himself (Zisook and Shear, 2009). See also the manual Treating prolonged grief in adults: a therapist's manual.
What these criteria do not say and what must be assessed
They do not measure rumination, which is nevertheless one of the best predictors of chronicity.
They say nothing about avoidance behaviours, which are the main target of treatment.
They say nothing about available support, nor about the real consequences already installed — debt, sick leave, isolation — which determine whether the programme is feasible at all.