In Adolescent Depression, Immune Markers Move in Both Directions
- In 280 adolescents (128 healthy controls, 78 with major depressive disorder without psychotic features, 74 with psychotic features), IL-6 and IL-10 were lower and IL-17A was higher in both depression groups than in controls (q < 0.05).
- Th17/regulatory polarization showed the largest group effect (partial η2 = 0.158) and was higher in both patient groups. TGF-β1 differed across the groups, mainly because levels were higher in the group without psychotic features than in controls.
- No cytokine, immune axis or correlation link distinguished the two depression groups. The correlation structure differed from controls at eight links in the psychotic group and at one link in the non-psychotic group.
- Adding immune features to kynurenine markers raised the AUC for healthy-versus-depressed discrimination from 0.806 to 0.868 (ΔAUC 0.064), but not for discrimination between the subtypes. The authors call this classification exploratory. The immune axes were not significantly associated with the kynurenine markers (permutation P = 0.577).
The phrase "depression is inflammation" reaches the consulting room in simplified form, often attached to the idea of a blood test. A study of adolescents from a Beijing hospital gives a less tidy picture. Some cytokines were higher than in healthy peers, and others were lower.
Who was measured and how
The authors, based at Beijing Anding Hospital, enrolled 280 adolescents: 128 healthy controls, 78 with major depressive disorder (MDD) without psychotic features and 74 with MDD with psychotic features. Serum levels of 11 cytokines were measured by ELISA. The team also computed five T-helper-related immune-axis scores that were specified in advance. The same cohort had earlier shown changes in tryptophan-kynurenine markers, but inflammatory markers had not been measured.
Lower, higher and the same
IL-6 and IL-10 were lower and IL-17A was higher in both MDD groups than in controls (q < 0.05). The largest effect was for Th17/regulatory polarization (partial η2 = 0.158), higher in both patient groups. Between the two MDD groups, no cytokine, axis or correlation link differed. The immune axes were not significantly associated with the kynurenine markers (permutation P = 0.577). Adding immune features to those markers raised the AUC for healthy-versus-MDD discrimination from 0.806 to 0.868, but not for telling the subtypes apart.
A caution for the inflammation formula
A client or a parent who has read that depression is an inflammatory illness may expect every marker to point up. In this sample IL-6 was lower and IL-17A was higher. That is a change in balance, and the formula "depression is inflammation" is too coarse to survive it. I would not describe psychotic depression to a family as an immunologically separate illness on this evidence, and I would not offer a cytokine panel as a diagnostic step. When the inflammation idea comes up in session, acknowledge it, say that some immune markers shift in some adolescents with depression, and keep the treatment plan where it is.
In this adolescent sample, depression came with lower IL-6 and IL-10 and higher IL-17A, which is a change in balance and not a uniform rise in inflammation.
The authors call the elastic-net classification exploratory. Their conclusion that the two depression groups were not distinguished applies to the peripheral immune markers measured in this study.