Serum cortisol level and depression severity in a sample of Brazilian elders

Address for correspondence: Luís Fernando Silva Castro de Araújo. Programa Terceira Idade (Proter), Instituto e Departamento de Psiquiatria, Faculdade de Medicina da Universidade de São Paulo. Rua Dr. Ovídio Pires de Campos, 785, 3o andar, Ceapesq, sala 14 – 05403-010 – São Paulo, SP, Brazil. Telephone: +55 (11) 3069-6973; Fax: +55 (11) 3069-8118. E-mail: laraujo@student.unimelb.edu.au Serum cortisol level and depression severity in a sample of Brazilian elders


Dear Editor,
The recent precision medicine movement has pushed research towards the identification of biomarkers that can be used for early diagnosis.We designed a pilot study that explores the correlation between blood cortisol level (BCL) with depression, depression severity and clinical comorbidities in a sample of aged Brazilian subjects.We hypothesized that BCL will significantly correlate to depression, its severity and with clinical comorbidities in our sample.
Participants were selected from an epidemiological study of older residents of the city of Sao Paulo 1 , which screened positive for depression (depression scale-D10-score ≥ 7) 2 and a pool of outpatients who received treatment for depression.Inclusion criteria were 60 years and older, and DSM-IV-TR 3 criteria for major depressive disorder based on a diagnostic interview by geriatric psychiatrists.Controls were 10 adults who were at least 60 years old without depression.Exclusion criteria included dementia, other organic mental disorder, and DSM-IV-TR-criteria-based diagnoses of any psychiatric disorder other than depression.Diagnosis and exclusion criteria were assessed with the CAMDEX interview 4 .
We followed 11 depressed subjects.Seven patients (63.6%) began depression after age 60 (late onset depression).Both groups had more female subjects (70% of controls and 54.5% of patients).The groups were similar in terms of marital status, mini-mental score, age and education.
In the initial appointment the subjects were assessed with: Mini Mental State Examination (MMSE) 5 ; the CAMCOG version validated for the Brazilian population 6 ; Montgomery-Asberg Depression Rating Scale (MADRS) 7 , Cumulative Illness Rating Scale (CIRS) 8 , Bayer Activities of Daily Living Scale (B-ADL) adapted for the Brazilian population 9 , and the Hamilton Rating Scale for Depression (HAM-D) 10 .To ensure that no subjects with incipient dementia would be included in the group we applied MMSE, CAMCOG and the B-ADL.We were unable to standardize the blood sample collection time, but all cases had it collected between 6-10 am (controls 6:29-9:42 am, mean = 8:57:55 am, mdn = 8:57 am; depressed subjects 6:19-9:52 am, mean = 9:23:44 am, mdn = 9:10 am).
These findings should be taken with caution, as we did not standardize the collection time for BCL.Nevertheless, they suggest that hypercortisolemic depressed elders comprise a subgroup within depressed subjects.Their clinical course may progress with more morbidity/comorbidities and functional deficits, as shown by the statistically significant relation to all four scales applied.Elevated BCL predicts depression (p = 0.037, df = 1, B = 0.34, SE = 0.162, Table 1), which suggests that BCL might be involved in the development of depression in aged patients.The odds of 1.402 means that for each raise of 1 unit (µg/dl) of cortisol level there is 40% increase in risk of depression (95% C.I. 1.020 -1.926).

Table 1 .
Logistic regression results for the prediction of diagnostic status from the BCL BCL: blood cortisol levels; B: the intercept; S.E.: standard error; Wald: Wald chi-square test; df: degrees of freedom; Sig.: significance; Exp(B): odds ratio; C.I.: confidence interval.