The present study examined the association between depressive disorders, depression characteristics and antidepressant medication with inflammation in a large cohort of depressed persons and controls. After taking a large set of possibly confounding factors into account, it was found that men with current depressive disorders had higher levels of CRP, and marginally higher levels of IL-6, but not of TNF-α. No overall associations were found in women. Increased inflammation was in particular found in depressed men with an older age of depression onset (CRP, TNF-α). Users of SNRI (men only), TCA and TeCA had increased levels of CRP and IL-6, whereas men using SSRI had significantly lower levels of IL-6.
Our results confirm previous findings of immune dysregulation in depressed persons.3, 4
In men, effect sizes for current depression were small to moderate for both CRP and IL-6. These effect sizes are comparable to those reported by the meta-analysis of Howren,4
in which stronger effects were found for studies using clinical interviews compared with studies using self-report questionnaires. Compared with symptom questionnaires, a depressive disorder diagnosis is less confounded by somatic health problems. Together with the fact that our results were elaborately adjusted for possibly confounding factors, our findings suggest a true relationship between depression and inflammation in men.
No clear associations between depression and inflammation were found in women. This is in line with results from another large and relatively young cohort study in which history of depression was associated with CRP levels in men, but not women.5
Hormonal changes throughout female life due to phase of menstrual cycle, use of hormonal contraceptives, menopause and use of estrogens greatly impact on inflammation levels,6, 7, 8
which could confound a clear association between depression and inflammation. However, in our study, associations between depression and inflammation appeared absent in both pre- and postmenopausal women and adjusting for several hormonal factors did not change our findings. Another explanation might be that in women, psychosocial factors have a larger role in depression and therefore override the effects of biological factors. For instance, insufficient social support and stressful life events have been found to pose a greater risk for depression among women compared with men.26, 27
Results of increased inflammation were particularly present in men with late-onset depression. In contrast, characteristics that are more often associated with an early age of onset, such as higher severity and longer duration, were not consistently associated with increased inflammation. Interesting to note is that women in our study also had an earlier age of depression onset compared with men. A distinct etiology in late-onset versus early-onset depression was also found by Kendler et al.28
This study showed that depression with an early age of onset was associated with a family history of depression, whereas late-onset depression was associated with a family history of vascular disease. Also, subclinical vascular dysregulations, such as atherosclerosis, have been found to relate to late-onset depression.11, 12
These findings are in line with the vascular depression hypothesis, which suggests that vascular damage in the brain predisposes to late-onset depression.29
As immune dysregulation is critically involved in vascular disease,13
this vascular damage could be the result of increased inflammation.
Instead of inflammatory or vascular, the true etiology of depression in this subgroup of men with late-onset depression might have a metabolic nature. Several studies have confirmed an association between CRP, IL-6 and TNF-α with the metabolic syndrome and visceral fat depots release cytokines.30
The metabolic syndrome and in particular visceral fat have been bidirectionally associated with depression in late life,31, 32
specifically in men. Men possess higher amounts of visceral fat compared with women and are therefore more likely to experience related inflammation and depression. Involvement of metabolic processes is further supported by our finding that the association between depression and inflammation in particular decreased after adjustment for BMI.
Several biological mechanisms could further explain the relationship between depression and inflammation. Depression has been associated with dysregulation of important stress systems of the human body, that is, the hypothalamus–pituitary–adrenal-axis33
and the autonomic nervous system.34
Although the hypothalamus–pituitary–adrenal-axis in normal situations should temper inflammatory reactions, prolonged hyperactivity of the hypothalamus–pituitary–adrenal-axis could result in blunted anti-inflammatory responses to glucocorticoids resulting in increased inflammation.35, 36
In addition, both decreased parasympathetic37
as well as increased sympathetic nervous system activity38
have been associated with increased inflammation. Furthermore, pro-inflammatory cytokines might inhibit hippocampal neurogenesis,39
which could lead to a reduced hippocampal volume,40
which is also seen in depression.41
Also, several inflammatory markers have been shown to promote indoleamine-2,3-dioxygenase activation,42
which catalyzes tryptophan, the precursor of serotonin, to kynurenine, thereby indirectly reducing the availability of serotonin.43
Lastly, specific genes might underlie both increased inflammation and depression, as several inflammation-related genes have been associated with susceptibility to major depression.44
Our study also shows that inflammation levels differ across persons using different types of antidepressant medication. The highest inflammation levels were found in men using SNRI, TCA or TeCA, while IL-6 levels in men using SSRI were lower compared with medication-free depressed men. Interesting to note is that the first three classes of medication have a combined serotonergic/noradrenergic effect, whereas SSRI act purely serotonergic. Earlier studies found decreases in inflammatory marker levels after antidepressant treatment, mainly SSRI (see ref. 14
for an overview), whereas two recent large studies found that use of antidepressants, mainly TCA, was associated with elevated inflammation levels.15
Possibly, noradrenergic effects are driving increased inflammation mechanisms. Noradrenaline is a part of the human stress response and has been suggested to potentiate cytokine production.45
Use of SNRI, TCA and TeCA has also been observed to disturb functioning of autonomic nervous system,46
and the metabolic syndrome.48
Although it is possible that persons using SNRI, TCA or TeCA are in other ways different from SSRI users and medication-free depressed persons, we constructed our groups in such a way that depression severity levels were very comparable. In addition, we adjusted our analyses for a large set of covariates and have therefore taken possible differences in lifestyle or disease factors into account. On the other hand, evidence suggests that increased inflammatory activity before treatment predicts non-response.49, 50
Possibly, persons with elevated inflammation did not respond to SSRI and were therefore prescribed SNRI, TCA or TeCA.
Associations found in this study were not always consistent across all inflammatory markers. Correlations between inflammatory markers were only modest and were highest between CRP and IL-6. Regulation of the immune system is rather complex and involves many different inflammatory mechanisms. Most consistent findings were found for CRP, which is a very general marker of inflammation. IL-6 and TNF-α, on the other hand, only tap part of the immune system. This seems to suggest that inflammation is indeed involved in depression, but it is still unclear which part of the immune system is most critically involved. Nonetheless, expected associations with covariates (for example, age, smoking, alcohol use, BMI, physical activity and somatic disease (medication)) were found for all inflammatory markers.
What do the findings of our study implicate with regard to treatment of depression? Considering the heterogeneity of depression and the fact that current treatment of depression is only effective in about a third to a half of patients,1, 2
this indicates that new treatments are needed for specific subgroups. Our finding of increased inflammation in men with late-onset depression, together with previous findings of high inflammation indicating non-response to antidepressants,49, 50
might suggest that this specific subgroup could benefit from alternative treatments, with anti-inflammatory medication being a likely candidate. Preliminary evidence from studies among patients treated with anti-inflammatory agents for other indications suggests that these agents may have beneficial effects on mood.51
One study found positive effects on mood in medically healthy, major depressed patients.52
On the other hand we found, like others,14
that SSRI might have a beneficial effect on inflammation, suggesting that SSRI could be effective in depressed patients with immune dysregulation through this anti-inflammatory effect. Furthermore, behavioral interventions, such as exercise, have been shown to normalize immune and metabolic dysregulation,53
as well as to improve depressive symptoms to some degree,54
and might therefore be an indicated treatment for a immune/metabolic depression subgroup. At this moment, these considerations for treatment implications are still very speculative. Follow-up (longitudinal) research should confirm and further delineate an inflammatory (or metabolic) depression subtype, taking into account age, sex, depression characteristics and course. Experimental studies are needed to examine and compare the effects of different types of currently available antidepressants, anti-inflammatory medication and behavioral (exercise) interventions on both immune/metabolic parameters and depression.
Our study has some important strengths such as a large sample size, assessment of multiple inflammatory markers, clinical diagnoses of depression, adequate adjustment for potential confounders, the ability to examine the role of depression characteristics and antidepressant medication. However, some limitations need to be acknowledged. As our data are cross-sectional, we cannot make any inferences about the direction of the association. Longitudinal studies are needed to investigate whether immune dysregulation is a precursor or the result of depression (treatment), or whether this relationship is bidirectional. The few available prospective studies have shown contradicting results.9, 55, 56
Further, like most other studies, we assessed circulating levels of inflammatory markers, which show a high degree of intra-individual variation, which could explain the rather modest overall associations between depression and inflammation in our study.
In conclusion, our study suggests that immune dysregulation has a role in a subgroup of depressed persons, in particular in men with a late-onset depression. Treatment trials should further examine the differential effects of different types of antidepressants on inflammation. Whether a specific treatment strategy (SSRI, anti-inflammatory drugs, exercise) is indicated for a subgroup of late-onset depressed patients with immune dysregulation needs to be further investigated, using longitudinal and experimental study designs.