PHQ-9 and GAD-7: why these two questionnaires became the clinical gold standard

Why this pair became the standard

In 2001, Kurt Kroenke, Robert Spitzer, and Janet Williams published the PHQ-9 — a brief 9-item depression scale. Five years later, the same authors plus Bernd Löwe released the GAD-7 — a 7-item anxiety companion. Over two decades, this pair has almost entirely displaced longer scales in primary care. The USPSTF, NICE, HEDIS — every major clinical guideline now recommends them as first-line screening.

Three reasons converge. Brevity: 16 items total, 3–5 minutes to complete. Public domain: free to use, no licensing, translated into 80+ languages. Direct DSM alignment: the 9 PHQ-9 items are literally the 9 DSM-IV and DSM-5 criteria for major depressive disorder. GAD-7 was designed by the same team as a companion instrument, with the same scale (0–3) and same cutoff (≥10). What emerges is a single workflow, not two separate tools.

What PHQ-9 measures

PHQ-9 is 9 items, each corresponding to one DSM-IV symptom of major depressive disorder. The assessment period is the past 2 weeks. The scale runs from 0 ("never") to 3 ("nearly every day"). Total score range: 0–27. The original validation by Kroenke et al. (2001) used a sample of 6,000 primary care patients. At a cutoff of ≥10, the instrument showed sensitivity of 88% and specificity of 88% against a blinded clinical interview — a rare convergence of two high parameters at once.

Eighteen years later, an individual participant data meta-analysis by Levis, Benedetti & Thombs (2019) in the BMJ replicated this result almost exactly: 58 studies, 17,357 participants, 2,312 cases of major depression. Against a semi-structured interview (the reference closest to clinical practice), sensitivity was 88%, specificity 85% at a cutoff of ≥10. The 2001 figures held up against the largest verification ever organized.

• Item 1: Little interest or pleasure in doing things — anhedonia
• Item 2: Feeling down, depressed, or hopeless — depressed mood
• Item 3: Trouble with sleep — sleep disturbance
• Item 4: Tiredness or little energy — fatigue
• Item 5: Poor appetite or overeating — appetite change
• Item 6: Feeling bad about oneself, guilt — worthlessness
• Item 7: Trouble concentrating — cognitive impairment
• Item 8: Moving slowly or being restless — psychomotor change
• Item 9: Thoughts of being better off dead — suicidal ideation, immediate clinical evaluation signal

"A PHQ-9 score of 10 or above had a sensitivity of 88% and a specificity of 88% for major depression. These characteristics plus its brevity make the PHQ-9 a useful clinical and research tool."— Kroenke, Spitzer & Williams, 2001, Journal of General Internal Medicine

What GAD-7 measures

GAD-7 was created 5 years after PHQ-9 by the same research group. An initial pool of 13 candidate items was reduced to 7 through factor analysis, item-total correlations, and mapping to DSM-IV criteria for GAD. The scale, assessment period, and cutoffs are identical to PHQ-9. Validation on a sample of 2,740 primary care patients (Spitzer et al., 2006) yielded sensitivity of 89% and specificity of 82% at a cutoff of ≥10.

An important property of GAD-7 that is often missed: it is validated not only for generalized anxiety disorder. Kroenke et al. (2007) in the Annals of Internal Medicine showed GAD-7 works as a screener for panic disorder, social phobia, and PTSD with AUC of 0.80–0.91. One brief questionnaire covers four anxiety syndromes. GAD-7 is not a narrow GAD instrument but a general anxiety screener. The first two GAD-7 items form GAD-2 — an ultra-brief two-item screen.

Thresholds and interpretation: same logic, two scales

The cutoffs of both instruments follow one principle and are easy to remember. 0–4 — minimal or none. 5–9 — mild. 10–14 — moderate. 15+ — severe (PHQ-9 adds a 20+ boundary for "very severe"). The clinical cutoff of ≥10 is the primary threshold for both instruments. Below — monitoring; above — intervention is indicated.

Item 9 of PHQ-9 deserves separate attention. It is a direct question about thoughts of death and self-harm. Any score above zero on this item requires immediate clinical risk assessment — regardless of the total score. This is a rule, not a recommendation: like item 10 in the EPDS, item 9 works as a standalone safety signal, independent of the total. GAD-7 has no equivalent item — if suicide risk is suspected in an anxious patient, add PHQ-9 or a specialized suicidality screening instrument.

PHQ-9 + GAD-7 over time: the MBC standard

The brevity of this pair is not just convenience — it is what makes weekly monitoring possible without overloading the client. Clients asked to complete the BDI-II or CORE-OM every week tend to resist. Three to five minutes of PHQ-9 + GAD-7 pass unnoticed. This is what turns the pair from a screening instrument into a measurement one: a series of readings shows the therapy trajectory in real time.

Löwe et al. (2004) established the minimal clinically important difference (MCID) for PHQ-9 at 5 points. For GAD-7 the commonly cited figure is approximately 4 points. This is a working benchmark for interpreting change: a decrease of 5+ PHQ-9 points from baseline is a clinically noticeable improvement, not measurement noise. MBC methodology involves plotting scores from session to session and discussing the graph with the client — which itself has therapeutic effect, independent of the therapy content.

Limitations and when to add other instruments

The pair has boundaries. PHQ-9 is focused on the 9 DSM criteria for major depression and does not cover the spectrum of depressive phenomena better captured by BDI-II (self-deprecation, sense of punishment, body image change). GAD-7 covers cognitive anxiety and worry but is less sensitive to somatic anxiety with pronounced physiological arousal — here BAI or HADS-A is more precise. For patients with predominantly somatic complaints or in medical settings, HADS is often preferable, specifically designed with somatic items excluded.

• BDI-II — for more detailed depression assessment or a wider severity range (0–63 vs PHQ-9's 0–27)
• BAI — when primary complaints are somatic: palpitations, dizziness, panic
• HADS — in medical settings where physical symptoms may confound PHQ-9 and GAD-7
• DASS-21 — when a third construct (stress) is needed alongside depression and anxiety
• EPDS — in the perinatal period, where PHQ-9 somatic items produce false positives

A gold standard is not "the best instrument for every case" but "the instrument to use by default." PHQ-9 + GAD-7 cover 80% of screening tasks in primary care and psychotherapy practice. Specialized instruments exist for the remaining 20%. The real strength of the pair emerges not from a single reading but from the series — when the client sees their own graph, and the clinician has objective data for clinical decisions. This is what turns 16 brief items into the foundation of measurement-based care.