HealthScorer

mental health

PHQ-9 explained: when a bad stretch becomes clinical depression

A bad week is not depression. Two weeks of low mood, sleep changes, and lost interest might be. The PHQ-9 is the nine-question tool clinicians use to draw that line. Here is what your score means and when it matters.

5/13/2026 7 min
A person sitting by a window with a coffee cup, looking out, soft morning light.
Photo on Unsplash

A bad week is not depression. A two-week stretch of low mood, broken sleep, and the kind of weight that makes brushing your teeth feel like work might be. The Patient Health Questionnaire (PHQ-9) is the nine-question tool clinicians around the world use to tell the difference.

This article walks through what each score band actually means, what the two-week clock is for, and when a number on a screen should turn into a phone call.

TL;DR

  • PHQ-9 has nine questions matched to the nine DSM-5 criteria for major depressive disorder, scored 0 to 27.
  • A score of 10 or higher is the threshold for further clinical evaluation, validated in 17,000+ patients by Levis 2019.
  • Bands: 0-4 minimal, 5-9 mild, 10-14 moderate, 15-19 moderately severe, 20-27 severe.
  • The 2-week window matters. PHQ-9 asks about the past 14 days, mirroring the DSM-5 minimum duration for a depressive episode.
  • Item 9 (thoughts of self-harm) is acted on regardless of total score. Any positive answer = same-day clinical contact.
  • For teens 13-17, the cutoff shifts up to 11 (Richardson 2010 in Pediatrics).
  • US crisis line: 988 (Suicide and Crisis Lifeline).

What the PHQ-9 is, and why it exists

Robert Spitzer and Kurt Kroenke published the PHQ-9 in 2001 in the Journal of General Internal Medicine. Their goal was modest and practical: a 9-question screen that a primary-care nurse could hand a patient in the waiting room, get back in under 3 minutes, and use to decide whether the appointment needed a 30-minute mental-health conversation or not.

The questions track the nine DSM-IV criteria for major depressive episode, one-to-one. Each is rated 0 (not at all) to 3 (nearly every day) for the past 2 weeks. The total runs 0 to 27.

Kroenke’s original 2001 study validated it against the structured clinical interview in 6,000 primary-care and obstetric patients. The threshold of 10 caught 88% of true major-depression cases and ruled out 88% of patients who did not have it. Twenty years later, Levis 2019 in the BMJ pulled together individual data from 58 studies and 17,357 patients and confirmed the same: at PHQ-9 ≥ 10, sensitivity is 0.88 and specificity is 0.85. The number has held.

The 2-week clock

The questionnaire opens with: “Over the last 2 weeks, how often have you been bothered by any of the following problems?”

That fortnight is not arbitrary. DSM-5 requires symptoms to be present for at least 2 consecutive weeks before a major depressive episode can be diagnosed. The PHQ-9 is built around that floor. A week of sadness after a breakup, a rough Monday after a sleepless weekend, the low couple of days that follow a stomach bug — none of these are what the questionnaire is built to measure.

Two weeks is also where the math of natural mood variation runs out. Most people swing through a few bad days a month. Sustained low mood, sustained sleep change, sustained loss of interest, lasting fourteen days in a row, is not the same thing.

What each band actually means

Kroenke 2001 proposed the now-standard scoring bands, and Manea 2012 in CMAJ meta-analytically tightened them:

ScoreBandWhat it usually means
0-4Minimal or noneWithin typical mood variation. No clinical action implied.
5-9MildSymptoms present but rarely require treatment alone. Watch the trend.
10-14ModerateThreshold for clinical interview. Treatment options discussed.
15-19Moderately severeTreatment usually indicated. Psychotherapy and/or medication.
20-27SevereActive treatment. Combination therapy typically recommended.

These are statistical bands, not diagnostic labels. A 14 does not mean “you have moderate depression.” It means “your symptoms cluster the way moderate depression clusters.” The interview, the history, the timeline — those decide what the score means for you.

NIMH summarises it the same way on its depression topic page: the questionnaire is a screen, not a diagnosis.

Item 9: the question that floats free of the total

The ninth question asks: “Thoughts that you would be better off dead, or of hurting yourself in some way.”

Any answer above zero — even “several days” — is treated as a safety signal independent of the rest. A score of 4 on the full PHQ-9 with a 1 on item 9 is acted on differently than a score of 4 with a 0. Kroenke 2001 designed it deliberately so that suicidality could not be diluted by other items.

If you answer above zero, two paths are standard practice:

  1. Same-day contact with your GP or your existing therapist.
  2. Or, if “today” feels too unstable, the 988 Suicide and Crisis Lifeline — call or text 988 in the US. The line is staffed 24 hours.

The point is not that thoughts of self-harm mean you are about to act. Most people who have those thoughts at some point in their life never do. The point is that the question is the one most worth not sitting on.

What “moderate” looks like in practice

A PHQ-9 of 12 (low end of moderate) maps to a person who, over the past 2 weeks:

  • Sleeps badly most nights (item 3 scored 2 or 3).
  • Feels tired most days (item 4 scored 2 or 3).
  • Has lost some interest in things that used to feel good (item 1 scored 2).
  • Notices their mood is low most days (item 2 scored 2).
  • Has some trouble concentrating at work or school (item 7 scored 1 or 2).
  • Items 5, 6, 8, 9 may be at 0 or 1.

That cluster is what the threshold of 10 is catching. It is not “everyone has bad days.” It is a 2-week stretch where most of the boxes that DSM-5 calls major-depressive episode are at least partially ticked.

When the PHQ-9 is misleading

Three situations push the score around without the underlying clinical picture matching:

Acute grief. In the first 2 to 6 weeks after a significant loss, PHQ-9 scores rise the same way they would for primary depression. DSM-5 removed the “bereavement exclusion” in 2013, so the score will not auto-correct for it. The clinical interview does.

Physical illness with overlapping symptoms. Hypothyroidism, anaemia, sleep apnoea, B12 deficiency — each can score the somatic items (sleep, energy, appetite) without low mood being primary. A high PHQ-9 in someone who has not had a basic blood panel in a year is a flag to do the blood panel first.

Single bad week inside a normal year. The 2-week window is a floor, not a ceiling. If you took PHQ-9 on the worst Tuesday of an otherwise stable year, you might score 12 today and 4 next month. Re-testing in 2 to 3 weeks separates a bad fortnight from a depressive episode.

The PHQ-9 number is data, not a verdict.

What to do with your number

  • 0-9. No action beyond noticing. If life context (a divorce, a job loss, a move) makes the score feel light, that context matters more than the score.
  • 10-14. Book a GP appointment within 1-2 weeks. Bring the score. Ask about a referral to psychotherapy and whether your context suggests medication.
  • 15-19. Book within the week. Be honest about item 9. Most GPs at this band recommend both therapy and an SSRI conversation.
  • 20-27. Same-week appointment. If item 9 is above zero, same-day contact (GP or 988).

If your score is 10+ and you also have anxiety symptoms, the matched GAD-7 is the parallel tool. Many people who screen positive on PHQ-9 also screen positive on GAD-7. Take both, bring both.

You can take the PHQ-9 itself here, in your browser, with nothing saved: PHQ-9 depression screener.

Worth remembering

A bad fortnight is not the same thing as depression, and a high PHQ-9 score is not a diagnosis. But the questionnaire is good at one specific job: telling the difference between mood you can ride out and mood that benefits from a conversation with someone trained.

Ten or higher, for two weeks or longer, is the line. Item 9, any value above zero, is its own line and does not wait for the total. The rest is what the appointment is for.

Frequently asked questions

What does a PHQ-9 score of 12 mean?
A score of 12 falls in the moderate band (10 to 14). Kroenke 2001 originally proposed treatment consideration starting at 10, and the Manea 2012 meta-analysis in CMAJ confirms 10 as the most balanced cutoff for major depression. A 12 is not a diagnosis. It is a strong signal that a conversation with a GP or counsellor is worth scheduling within the next week or two.
From what PHQ-9 score is it considered depression?
Ten. Levis 2019, a meta-analysis in the BMJ pooling individual data from 17,357 participants across 58 studies, confirmed that a PHQ-9 of 10 or higher gives the best balance of sensitivity (88%) and specificity (85%) for major depression. Below 10 the score is still informative but rarely warrants a diagnostic interview on its own.
Should I see a doctor with a PHQ-9 score of 10?
Yes, especially if the score has been at 10 or above for more than two weeks. Ten is the threshold the American Psychiatric Association and NIMH use to recommend a full clinical interview. Booking your GP is a reasonable first step. If you have any thoughts of self-harm, regardless of the total score, contact 988 (the US Suicide and Crisis Lifeline) the same day.
What does a PHQ-9 score of 20 mean?
Twenty falls in the severe band (20 to 27). Kroenke 2001 framed this band as requiring active treatment, typically a combination of pharmacotherapy and psychotherapy. A score this high paired with any positive answer on item 9 (thoughts of self-harm) is a same-day clinical contact, not a wait-and-see. NIMH lists 988 as the immediate US contact.
Can a PHQ-9 score of 5 still mean I should worry?
Usually no, but the trend matters. A 5 falls in the mild band (5 to 9). If your score climbed from 2 to 5 in a month, that is more meaningful than a stable 5 you have lived with for a year. Kroenke 2001 notes that mild scores predict functional impairment less reliably than scores at 10 or above, so the score alone is rarely a trigger for treatment.
What if I score high on PHQ-9 but feel mostly fine?
PHQ-9 measures the past two weeks. If you completed it on a particularly bad day or week, your average state may be different. Re-test in 7 to 14 days. Levis 2019 found PHQ-9 reliability is highest when administered twice across a 2-week window. A single elevated score that drops the second time around tells a different story than one that stays elevated.
PHQ-9 vs Beck Depression Inventory: which is better?
Different tools, comparable performance. The Manea 2012 meta-analysis in CMAJ found PHQ-9 sensitivity at 88% and specificity at 85% at the cut-off of 10. Beck's Depression Inventory (BDI-II) lands at similar numbers. PHQ-9 is shorter (9 items vs 21), free, and aligned with DSM-5 criteria. BDI is older, copyrighted, and slightly more sensitive to anxious-depression presentations. For most primary care contexts, PHQ-9 wins on practicality.
Is PHQ-9 reliable for teenagers?
Yes, with a small caveat. Richardson 2010 in Pediatrics validated PHQ-9 in 442 adolescents (ages 13 to 17) and found 89.5% sensitivity at a cutoff of 11 (slightly higher than the adult 10). For teens 13 to 17 the recommended cutoff is 11. The questions stay the same; the threshold for further evaluation moves up by one point because adolescent baseline mood scores run a touch higher.
How often should I retake the PHQ-9?
Every 2 to 4 weeks if you are tracking treatment response, weekly if you are in active therapy with significant symptom shift. Kroenke 2001 and the APA both recommend PHQ-9 as a longitudinal measure: the trend over months tells more than a single number. Daily retakes are noise; the questions ask about the past 2 weeks, so the score does not move meaningfully in 24 hours.
Can grief make PHQ-9 look like depression?
Yes, especially in the first 2 to 6 weeks after a major loss. DSM-5 removed the bereavement exclusion in 2013, so a PHQ-9 score elevated during acute grief now reads numerically the same as one elevated by primary depression. NIMH guidance recommends a clinician distinguish grief-driven low mood from major depressive disorder if symptoms persist beyond 2 months or include item-9 thoughts.
What does PHQ-9 item 9 about thoughts of self-harm mean?
Item 9 asks how often, in the past 2 weeks, you have had thoughts that you would be better off dead or of hurting yourself. Any answer above zero is a safety flag and is acted on independently of the total score. Kroenke 2001 designed it this way deliberately. If you answer above zero, contact 988 (US Suicide and Crisis Lifeline) or your GP the same day.
Does PHQ-9 diagnose anxiety too?
No. PHQ-9 measures depression. For anxiety, GAD-7 is the matched 7-item tool (also from Spitzer and Kroenke). Many people score above threshold on both. Levis 2019 notes that PHQ-9 and GAD-7 correlate at roughly 0.7, meaning they overlap but are not interchangeable. If you score 10+ on PHQ-9 and feel anxiety is the bigger problem, take GAD-7 too and bring both numbers to your GP.
What is the lowest PHQ-9 score that is still depression?
Five, in theory. Scores of 5 to 9 are the mild band, and a small fraction of people in that band meet DSM-5 criteria for minor depression. But the cut-off where the score reliably flags major depressive disorder is 10. Below 10 the score is suggestive, not diagnostic, and the clinical interview decides.
Does HealthScorer save my PHQ-9 answers?
No. PHQ-9 is calculated entirely in your browser. Nothing about your answers, your score, or your IP address leaves your device or hits any HealthScorer database. There is no account, no login, no analytics tied to your responses. The numbers you see are yours alone.

Sources

  1. The PHQ-9: validity of a brief depression severity measure — Kroenke K, Spitzer RL, Williams JB (Journal of General Internal Medicine, 2001) — Society of General Internal Medicine [peer-reviewed] PMID 11556941
  2. Accuracy of the PHQ-9 for screening to detect major depression: individual participant data meta-analysis — Levis B, Benedetti A, Thombs BD (BMJ, 2019) — BMJ Publishing Group [PubMed meta-analysis] PMID 30967483
  3. Optimal cut-off score for diagnosing depression with the Patient Health Questionnaire (PHQ-9): a meta-analysis — Manea L, Gilbody S, McMillan D (CMAJ, 2012) — Canadian Medical Association [PubMed meta-analysis] PMID 22184363
  4. Depression — National Institute of Mental Health [government health body]
  5. 988 Suicide and Crisis Lifeline — Substance Abuse and Mental Health Services Administration (SAMHSA) [government health body]
  6. Evaluation of the Patient Health Questionnaire-9 Item for Detecting Major Depression Among Adolescents — Richardson LP, McCauley E, Grossman DC, et al. (Pediatrics, 2010) — American Academy of Pediatrics [PubMed review] PMID 20603258