# Zero symptoms in common? It depends what you count.

One person cannot sleep and loses weight. Another sleeps too much and gains weight. Both may have depression. Yet opposite manifestations can meet the same criteria: sleep and weight. That distinction changes what “zero” means.

Marius Comper · Revised 5 September 2026 · [DSM](https://www.ncbi.nlm.nih.gov/books/NBK571021/box/ch4.b4/?report=objectonly)

## The same criterion can contain opposite experiences

For a major depressive episode, the symptom rule requires at least five of nine criteria within the same two-week period, including depressed mood or loss of interest or pleasure. Two sets of five drawn from nine must share at least one element.

5 + 5 − 9 = 1At least one shared criterion. Zero is impossible at this level.

Below are two hypothetical profiles. The first meets five symptom criteria, the second six. The listed manifestations differ, but three criteria are shared.

Same criteria, different manifestations

Criterion | Profile A | Profile B | 
| --- | --- | --- |
1. Mood | Depressed mood | — | 
2. Interest / pleasure | — | Loss of interest or pleasure | 
3. Weight / appetite | Weight loss | Weight gain | 
4. Sleep | Insomnia | Hypersomnia | 
5. Psychomotor activity | Agitation | Slowing | 
6. Energy | Fatigue | — | 
7. Self-worth / guilt | — | Excessive guilt | 
8. Concentration / decisions | — | Difficulty concentrating | 
Criteria 3, 4 and 5 are shared. A dash means a manifestation is absent in this example. Criterion 9, recurrent thoughts of death or suicide, is absent in both profiles and omitted from the table.

Counting insomnia and hypersomnia separately can yield “zero identical manifestations” in these lists. Grouping them under the sleep criterion yields overlap. Neither count establishes that two people have nothing clinically in common.

The lists illustrate only the symptom rule. Diagnosis also requires significant distress or impaired functioning and assessment of other explanations. For major depressive disorder, a history of mania or hypomania also matters. The table does not establish a diagnosis or select a treatment. [DSM](https://www.ncbi.nlm.nih.gov/books/NBK571021/box/ch4.b4/?report=objectonly)

## How many ways to meet the same rule?

Diagnostic rules that allow multiple combinations are called polythetic. The calculation treats each criterion as present or absent. It ignores severity and does not tell us how often a profile occurs in clinical practice.

Major depressive episode227

At least 5 of 9, with at least one of the first two criteria.

256 − 29 = 227

Of all 256 selections containing at least five criteria, remove the 29 containing neither core criterion.

PTSD636,120

Four groups: at least 1 of 5, 1 of 2, 2 of 7 and 2 of 6.

31 × 3 × 120 × 57

Count the allowed selections within each group, then multiply. Trauma exposure, duration over one month and the other diagnostic requirements still apply.

ADHD, through age 16116,220

At least 6 of 9 in at least one of the two groups.

512² − 382²

From age 17, the threshold falls to five and the same calculation gives 196,608 combinations. Onset, duration, presence in multiple settings and functional impairment are outside this count.

These are combinations allowed by symptom rules, not hundreds of thousands of established diseases or observed profiles. The PTSD count uses DSM-5 criteria for adults and children older than six, retained in DSM-5-TR. [PTSD](https://www.ptsd.va.gov/professional/treat/essentials/dsm5_ptsd.asp) · [ADHD](https://www.cdc.gov/adhd/hcp/clinical-care/index.html)

In PTSD, two selections can be entirely disjoint even at the level of the 20 symptoms: for example, B1/C1/D2/D4/E1/E3 and B2/C2/D1/D6/E2/E6. Each meets the group thresholds. Both must still satisfy the shared clinical requirements of the diagnosis.

## Observed variation is a different question

Variation also appears in data. Fried and Nesse examined 3,703 outpatients with depression in STAR*D and identified 1,030 symptom profiles. The most common occurred in just 1.8% of participants. [Fried & Nesse, 2015](https://pubmed.ncbi.nlm.nih.gov/25451401/)

The 1,030 profiles do not contradict the 227 combinations: the study separated manifestations grouped within DSM criteria and converted responses into presence or absence. The result depends on the instrument and this coding. It is an observation in a treatment-seeking sample, not a distribution for the whole population.

For the reader, the point is simple: a diagnostic name and a total score leave important information out. Which symptoms a person has, how severe they are and how they affect daily life matter. The number of combinations alone does not decide a diagnosis’s usefulness.

## From differences between people to a hypothesis about causes

Denny Borsboom’s network theory proposes that symptoms can sustain one another. A possible loop links insomnia, fatigue, difficulty concentrating and worry that disrupts sleep again. The arrows express a causal hypothesis. Counting criteria does not prove it. [Borsboom, 2017](https://pmc.ncbi.nlm.nih.gov/articles/PMC5269502/)

Insomnia → Fatigue → Poor concentration → Worry → Insomnia

Illustrative loop. The relationships and their strength may differ between people.

One possible consequence of strong feedback is hysteresis: after an external input disappears, a system can remain in a state different from its starting state. History matters. The model below demonstrates this mathematical property; it does not reproduce a person’s course of depression.

### A system that retains its state

Compare weak and strong feedback. Apply the input, then remove it. With strong feedback, activation remains high.

Feedback 

Experiment result: with strong feedback, activation rises from about 0.02 to 1.00 with the input and remains at 0.98 after its removal. With weak feedback it returns to 0.02. Values are abstract units, not clinical scores.

Equation and limits of the example

x′ = 1 / [1 + exp(−(c × x + s − 4))]

Start at x = 0.02. Feedback c is 2 or 8; input s is 0 or 3. Each action computes 100 steps from the previous state. Changing feedback and resetting start a new experiment. Coefficients are chosen for demonstration, not estimated from patient data. One number here compresses the whole system.

In research, a network estimated from correlations does not automatically establish that one symptom causes another. Group-level relationships need not describe a particular person. Robinaugh and colleagues call for robust empirical phenomena and formal theories that explain them. [Robinaugh et al., 2020](https://pubmed.ncbi.nlm.nih.gov/31875792/)

## What overlap can and cannot tell us

Multiple diagnoses can include similar problems. Sleep appears in both depression and PTSD criteria. This can complicate category boundaries, but does not establish that multiple diagnoses are merely a catalogue error. A shared symptom alone establishes neither diagnosis.

In a network hypothesis, a symptom connecting two groups might act as a bridge. Recommending an intervention requires evidence that changing it produces a clinical benefit. A drawn connection or a central node is not enough.

Three distinct results remain. Rules permit different profiles. Studies can measure variation in observed profiles. Models propose explanations for how problems arise and persist. A diagnosis can guide assessment without telling a person’s whole story.

## Sources and method

Original calculations on binary symptom rules; no prevalence estimates. This article discusses DSM and does not generalize its results to ICD. Sources and bibliography checked on 5 September 2026.

[DSM](https://www.ncbi.nlm.nih.gov/books/NBK571021/box/ch4.b4/?report=objectonly)

[PTSD](https://www.ptsd.va.gov/professional/treat/essentials/dsm5_ptsd.asp)

[ADHD](https://www.cdc.gov/adhd/hcp/clinical-care/index.html)

[Fried & Nesse, 2015](https://pubmed.ncbi.nlm.nih.gov/25451401/)

[Borsboom, 2017](https://pmc.ncbi.nlm.nih.gov/articles/PMC5269502/)

[Robinaugh et al., 2020](https://pubmed.ncbi.nlm.nih.gov/31875792/)

https://mariuscomper.uk/zero-simptome-comune/en/