PHQ-9 and GAD-7: What Mental Health Screeners Can (and Can't) Tell You

If you've ever been handed a short questionnaire at a doctor's office that begins "Over the last two weeks, how often have you been bothered by any of the following problems?" — you've almost certainly met the PHQ-9 or the GAD-7. These two brief screeners are the workhorses of mental health measurement worldwide, used in clinics, research studies, and a growing number of apps. They're genuinely useful tools. They're also widely misunderstood — treated as verdicts when they're really more like weather reports. Here's what they measure, how the scoring actually works, and — most importantly — what a score can and cannot tell you.

What the PHQ-9 measures

The PHQ-9 (Patient Health Questionnaire-9) was developed in the late 1990s by Robert Spitzer, Kurt Kroenke, Janet Williams, and colleagues, with its validation published in 2001. It asks how often, over the past two weeks, you've been bothered by nine experiences: low mood, loss of interest or pleasure, sleep problems, low energy, appetite changes, feeling bad about yourself, trouble concentrating, moving or speaking noticeably slowly or restlessly, and thoughts that you'd be better off dead or of hurting yourself.

Each item is rated from 0 ("not at all") to 3 ("nearly every day"), giving a total from 0 to 27. The nine items deliberately mirror the nine symptom criteria clinicians use when assessing depression — which is why the tool became so widely adopted.

One item deserves special mention: item 9 asks about thoughts of death or self-harm. Any answer above zero on that item is worth discussing with a professional promptly, regardless of what the total score says.

What the GAD-7 measures

The GAD-7 (Generalized Anxiety Disorder-7) came from the same research team, with Bernd Löwe, in 2006. Its seven items ask how often, over the past two weeks, you've been bothered by: feeling nervous or on edge, not being able to stop or control worrying, worrying too much about different things, trouble relaxing, restlessness, irritability, and feeling afraid that something awful might happen.

Items use the same 0–3 scale, for a total of 0 to 21. Although it was designed around generalized anxiety, scores also tend to run high in panic and social anxiety — so in practice it works as a broad anxiety signal rather than a narrow one.

How the scoring bands work

Both questionnaires come with conventional bands, established in the original validation research. For the PHQ-9:

• 0–4 — minimal depression symptoms
• 5–9 — mild
• 10–14 — moderate
• 15–19 — moderately severe
• 20–27 — severe

For the GAD-7:

• 0–4 — minimal anxiety symptoms
• 5–9 — mild
• 10–14 — moderate
• 15–21 — severe

A score of 10 or above on either measure is the threshold most often cited as warranting further evaluation by a professional. But notice the careful wording those bands deserve: they describe the level of symptoms you reported over the past two weeks. Not your character, not your prognosis, not a diagnosis. Which brings us to the point that can't be made strongly enough.

Screening is not diagnosis — full stop

A screener is a smoke detector, not a fire report. It's designed to be sensitive — to catch as many people as possible who might benefit from a closer look — and that sensitivity comes at a cost: it flags plenty of situations that a fuller assessment would explain differently.

A questionnaire cannot see context. It doesn't know you're three weeks into grieving someone you loved. It doesn't know you have a newborn and haven't slept more than four hours in a month, which alone will light up items about sleep, energy, and concentration. It doesn't know about your thyroid condition, your medication change, your night-shift schedule, or the layoff rumors at work. A trained professional weighs all of that — your history, your circumstances, your physical health, how long things have lasted and how much they interfere with your life — before reaching a diagnosis. No total score can do that arithmetic.

So hold both directions of the caution. A high score does not mean you "have" depression or an anxiety disorder — it means your recent symptom load is high and a conversation with a professional is a good idea. And a low score does not invalidate your struggle — plenty of real difficulties don't happen to match the nine or seven items on these particular forms.

Why tracking over time is the real value

Here's the part most people miss: a single score is a snapshot, and snapshots are noisy. One rough week, one bad night's sleep before you filled it in, one stressful afternoon — any of these can nudge a score a few points. The signal isn't in the number; it's in the trend.

Taken every couple of weeks — matching the "last two weeks" window the questions actually ask about, so there's no point taking them daily — these screeners become something much more useful: a trend line. Is the line drifting down over two months? Something you're doing is helping. Did it jump eight points? Worth asking what changed. Has it sat above 15 for six weeks despite your best efforts? That's meaningful information to bring to a professional, far more persuasive than "I've been feeling kind of bad lately."

Trend lines get even more useful when you pair them with notes about what you were doing — starting therapy, changing jobs, exercising, going through a breakup. The score tells you how much; it never tells you why. For the why, you need the CBT side of the ledger: noticing what your mind is actually doing, like the patterns in our guide to common cognitive distortions, and what you're doing behaviorally week to week.

How to talk to a professional about your scores

If you've been tracking and want to bring the results to a doctor or therapist, a little structure goes a long way:

• Bring dates and scores, not just impressions — "I scored 14, 16, and 15 over the past six weeks" is concrete and useful.
• Describe the change, if there is one: "This was a 6 in March and it's been over 12 since May."
• Say what you've already tried — exercise, self-help exercises, cutting back on drinking — and what happened.
• Ask open questions: "What do you make of these numbers?" rather than "Do I have depression?"

Clinicians use these exact instruments every day; you'll be speaking their language. Some apps make this easier by letting you export your history — Untangle: CBT Companion, for example, includes free PDF export and clinician sharing — but a handwritten list of dates and scores works fine too.

Privacy: think before you type

One more thing, because screener answers are sensitive health information — especially item 9 on the PHQ-9. Before using any digital screener, it's worth knowing: Where are your answers stored — on your device, or on a server, and in what country? Are responses shared with or sold to any third parties? Can you export your history? Can you delete it entirely? Reputable tools answer these questions plainly in their privacy policy rather than burying them. If you can't find the answers at all, that is itself an answer.

The bottom line

The PHQ-9 and GAD-7 are quick, free, well-validated ways to put a rough number on two hard-to-measure things — and to watch that number move over time. Used that way, they're excellent companions to self-help work and to professional care. Just never let a number tell you who you are. It can't. It was never designed to.

This article is for education, not medical advice. If you're in crisis, contact your local emergency number or the 988 Suicide & Crisis Lifeline (US).

Referencer

1. The PHQ-9: Validity of a Brief Depression Severity Measure (Kroenke, Spitzer & Williams, 2001) — Journal of General Internal Medicine (PubMed)
2. A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7 (Spitzer, Kroenke, Williams & Löwe, 2006) — Archives of Internal Medicine (PubMed)
3. Patient Health Questionnaire (PHQ) Screeners — Pfizer / phqscreeners.com
4. Depression — National Institute of Mental Health