State anxiety tracks the pain, trait anxiety does not reach significance: 43 spinal pain patients in Moscow

Chronic spinal pain reaches our offices already braided with mood and anxiety, and the standard move is to name the comorbidity and stop there. A small study from Sechenov University does two less common things: it puts a psychiatrist in the room with every single patient, and it keeps state and trait anxiety apart instead of collapsing them into one "anxiety" variable. Both choices change what the data say.

A psychiatrist in the room

Forty-three patients with chronic non-specific cervicalgia, lumbalgia or lumboischialgia were assessed. Radiculopathy, spinal stenosis, acute spinal trauma, malignancy within five years and marked organic cognitive impairment were excluded; the age window was 18 to 74. Mean pain was 6.74 ± 1.18 on the numeric rating scale, mean Central Sensitization Inventory 36.77 ± 11.46, and SF-12 summary scores were 35.4 ± 8.82 physical and 41.56 ± 9.55 mental. The sample was pre-selected for elevated anxiety or depression, so it is an enriched clinical group and carries no prevalence claim.

Then the psychiatrist saw everyone, and every patient came away with a diagnosis. Twenty-five of the 43 carried a depressive-spectrum diagnosis: recurrent depressive disorder in 14 and depressive episode in 11. Generalised anxiety disorder accounted for 11 more, with mixed anxiety and depressive disorder in 4 and single cases of panic disorder, unspecified anxiety disorder and agoraphobia. Set that against the mean Beck score of 12.88, which the paper reports without assigning it a severity band. The gap between the questionnaire and the interview is the most portable thing in this paper.

The state and trait split is the second. Mean state anxiety was 30.63 ± 9.16 and mean trait anxiety 45.93 ± 10.56, which complicates the story rather than confirming it: the group sat low on state anxiety and high on trait anxiety, and it was the low-scoring measure that moved with the pain. State anxiety moved with pain intensity (r = 0.406, p = 0.007) and with reduced physical activity (r = -0.355, p = 0.020); trait anxiety reached significance on neither, at r = 0.142 (p = 0.363) and r = -0.237 (p = 0.127). The second of those coefficients is not a trivial estimate; it simply did not clear the threshold in a sample this size. The regression models repeat the pattern: depression and state anxiety predicted both pain and activity, while trait anxiety survived only as a predictor of the mental component of quality of life (β = -0.364, p = 0.007). Trait anxiety is not inert here, though. All three affective measures correlated with central sensitisation scores, trait anxiety included (r = 0.337, p = 0.027).

What to do with the distinction

Practically, this separates two jobs. State anxiety is the arousal in the room today, and it is what within-session regulation, graded exposure and activity pacing are built for. Trait anxiety is a longer piece of work, and in this sample it showed up not in how loud the pain was but in how poor life felt overall. If a patient's pain report and activity level are moving, look at current arousal and mood before reaching for a temperament formulation.

Second, do not let a low questionnaire score close the question. A mean Beck of 12.88 and a psychiatrist-confirmed depressive episode or recurrent depression in 25 of 43 patients cannot both be the whole story, and the interview is the one that produced treatable diagnoses. Screeners belong at the front of an assessment, not at the exit.

Third, physical activity is the lever worth watching. Depression and state anxiety both predicted lower IPAQ scores while trait anxiety did not reach significance, and deconditioning is one of the few elements in this loop a therapist can move directly. The design is cross-sectional, so "predicted" means "was associated with in a regression model" and nothing here establishes which way the arrow runs.