Depressive disorder is a well-known chronic, recurrent and disabling mental disease with high direct and indirect costs to society in both western and eastern cultures.1-3 Although a large number of novel antidepressants have been introduced over the past few decades, at least 40% of depressed patients show only partial or no response to initial or even multiple antidepressants medication.4,5 Thase et al6 had characterized 5 stages of treatment-resistant depression (TRD) in 1995. In this model, the second-stage criterion of TRD, which shows little effect to adequate dose and time treatment of 2 or more antidepressants with different mechanisms of action, is the most widely accepted in clinical research. In clinical work, TRD patients often show the following features: late onset, accompanying anxiety, psychotic symptoms or cognitive impairment, recurrent attacks or previous lack of effect, together with other psychosis, personality disorder, mental retardation or physical diseases, persistent stress factors, and lack of effective social support system. But there is still no good means to objectively identify TRD by any instrumentation.
The cerebrum will suffer an overload of indifferent stimulus if its inhibiting function is weakened, and this will cause many kinds of psychiatric symptoms. Sensory gating (SG) P50 is an electrophysiological index which reflects the brain inhibiting function by testing the brain auditory evoked potentials. It has two testing mode, the conditioning and testing stimulus paradigm and the stimulus train paradigm.
In the past, most studies on P50 were aimed at schizophrenia. The concordant findings of these studies were that the ratios of testing stimulus (S2) and conditioning stimulus (S1) (S2/S1) on the P50 of schizophrenia patients were higher than those of health adults which revealed that these patients had defects on SG.7-9 There are still few studies on P50 of mood disorders, especially on the P50 of TRD patients. Olincy et al10 found that the P50 deficits also existed in patients of bipolar disorder who had a history of psychotic symptoms. Recently Jiang et al11 reported that patients with first episode depression also had deficits of P50 suppression, they could not filter irrelevant information effectively. It seems that TRD patients may also have P50 deficits. Are these deficits more severe in TRD patients compared to non treatment-resistant depression (NTRD) patients and healthy controls (HC)? What is the relationship between the P50 deficits and the clinical symptoms in TRD? Can we use P50 characteristics to identify TRD earlier or to predict a curative effect? This study tries to answer these questions.
All 100 patients were enrolled from the Shanghai Mental Health Centre including outpatients and inpatients. Inclusive criteria were the following: diagnosed depressive episode according to the International Classification of Diseases-10 (ICD-10) by 2 or more experienced psychiatrists; had at least a score of 17 on 17-item Hamilton Rating Scale for Depression (HAMD-17); aged from 17 to 72; had at least an education level of junior high school; had adequate audio and visual level for the study. Exclusive criteria were the following: had severe physical diseases or severe suicide attempt; gravida or women lactating; had a recent treatment with electro-convulsive treatment (ECT); history of poor compliance with treatment; alcohol or drug abuser; had previous definite manic or hypomanic episode. Fifty patients were first depressive episode patients who were free of psychotropic drugs at least two weeks before they were enrolled. After the measurement of their baseline P50, they were given a 6-week treatment of a single antidepressant such as fluoxetine, paroxetine, or sertraline. Thirty-nine of them (25 men and 14 women, mean aged (33±13) years, mean education time was (10±2) years, mean course of disease was (15±13) weeks, mean scores of baseline HAMD-17 was 22.80±4.35) showed good reaction to treatment. The scores of their HAMD-17 decreased more than 50% without any manic or hypomanic episode, and they were enrolled to the NTRD group, while the other 11 patients were excluded because there was a lack of curative effect to one antidepressant but still did not meet the inclusive criteria of TRD. Other 50 patients (32 men and 18 women, mean aged (45±14) years, mean education time was (10±2) years, mean course of disease was (15±12) years, mean score of baseline HAMD-17 was 23.42±4.29) showed little improvement (the scores of HAMD-17 decreased less than 50%) to enough dose (such as at least 150 mg/d of imipramine) and time (6 weeks) treatments of 2 or more antidepressants with different mechanism of action and were free of psychotropic drugs at least one week before they were enrolled to the TRD group. After the measurements of their baseline P50, they were given a single treatment of 75 mg or 150 mg venlafaxine (Efexor, USA). Thirty-six of them had a repeated measurement of P50 and HAMD-17 (mean score was 22.04±4.10) after an 8-week treatment duration, others withdrew the informed consent or failed to follow-up. All the patients were told the details of the study and signed the informed consent by themselves or their legal guardians.
There were 51 HC, 30 men and 21 women, mean aged (45±12) years (range 23–63). They had received an education of (10±2) years who were volunteers recruited at the same time from the Shanghai Mental Health Centre, Shanghai Jiao Tong University School of Medicine and the community nearby. They did not have any psychosis, mental retardation, physical diseases, history of alcohol or drug abuse or history of major psychosis in first degree relatives.
No significant difference in sex and education level was found among the three groups. Significant differences in age and course of disease were found between TRD and NTRD groups while no significant difference in scores of baseline HAMD-17 was found. Compared to the scores of baseline HAMD-17, no significant change was found after venlafaxine treatment in TRD group.
P50 test was performed in a soundproof room using auditory conditioning stimulus (S1)-testing stimulus (S2) paradigm. Recordings were obtained between 10:00 a.m. and 12:00 p.m. while subjects were in a supine position and staring at a fixed target to maintain alertness. P50 features were obtained using Bravo (Nicolet Biomedical, USA). The recording electrode was attached at central zone (Cz) according to the 10-20 international system for electrode placement, with the mastoid leads as reference and prefrons zone (Fpz) as ground. All electric resistances were less than 5 kΩ. A 0.10-ms duration square wave of 85 db pairing burst sound (S1 and S2) which was triggered by outlaid signal generator was delivered to the subjects through earphone. A conditioning- testing paradigm presented 32 pairs of auditory stimuli with an intrapair interval of 0.5 seconds and interstimulus interval of 10 seconds. The input signals were amplified by amplifier and were analyzed in a window of 200 ms. The P50 which was elicited by S1 was called conditioned stimulus wave (S1-P50), and the P50 which was elicited by S2 was called testing stimulus wave (S2-P50).
Several indexes of P50, including the latencies and amplitudes of S1-P50 and S2-P50, were identified. Three expressions reflecting the deficit of SG (including S1–S2, S2/S1 (%) and 100 × (1–S2/S1)) were also measured.
All data were expressed mean ± standard deviation (SD) and statistical analysis was performed using SPSS 17.0. The difference among three groups of P50 measures was evaluated using multivariate analysis of variance (MANOVA) and Kruskal-Wallis test. The correlation between P50 and the scores of HAMD-17 was performed using the Pearson correlation analysis. The longitudinal change of P50 before and after the antidepressant administration was tested using a paired-sample t test. P <0.05 was considered statistically significant.
Abnormalities of the baseline P50
The baseline data of P50 from the patient group were used to analyze the difference of P50 and three P50 expressions among the three groups. Because significant differences in age and course of disease were found between the TRD and NTRD groups, we found no correlation between the three P50 parameters by Pearson correlation analysis: S2-P50 amplitude, S2/S1 (%) and 100 × (1–S2/S1) (r=0.135, 0.100 and –0.094, P >0.05). We also found that age had no significant effect (F=0.276, P >0.05) on S2-P50 amplitude (F=0.136, P >0.05), S2/S1 (%) (F=0.001, P >0.05) and 100 × (1–S2/S1) (F=0.000, P >0.05) by MANOVA. No significant differences were shown between S1-P50 latency and amplitude; but significant differences were shown between S2-P50 latency and amplitude, also between S1–S2, S2/S1 (%) and 100 × (1–S2/S1). Increased S2-P50 amplitude was shown in the TRD and NTRD groups (P <0.01 and P <0.05) compared with the HC group, but no difference was shown on S2-P50 amplitude between the TRD and NTRD groups (P >0.05). A bigger value for S2/S1 (%) and a smaller value for 100 (1–S2/S1) were shown in the TRD and NTRD groups (P <0.01) compared with the HC group. While compared with the NTRD group, the TRD group also showed a bigger value for S2/S1 (%) and a smaller value for 100 × (1–S2/S1) (P <0.05, Table 1).
|view in a new window |
Table 1. P50 expressions among the three groups
Follow-up of P50 changes
After 8-week medication, the scores of HAMD-17 in 36 TRD patients were not significantly changed (P >0.05), while the P50 and three P50 expressions also showed no change compared with the baseline P50 measurements (P >0.05, Table 2).
|view in a new window |
Table 2. P50 before and after the medications in TRD patients
Correlations between three P50 expressions and the scores of HAMD-17
Correlations between the three P50 expressions and the scores of HAMD-17 were evaluated using the data before and after medication from 36 TRD patients. Before medication, the Pearson correlation analysis revealed a positive correlation between S2/S1 and the scores of HAMD-17 (r=0.802, P <0.01). Both S1–S2 and 100 × (1–S2/S1) also showed a negative correlation with the scores of HAMD-17 (r= –0.802, P <0.01). But after 8-week medication, a Pearson correlation analysis revealed no correlation between the three P50 expressions and the scores of HAMD-17 (r= –0.225, 0.271 and –0.271, P >0.05, Table 3).
|view in a new window |
Table 3. Correlation between different P50 expressions and the scores of HAMD-17 before and after the medications in TRD patients
The auditory system is an important pathway for people to accept information. The cerebrum of a healthy adult orderly imports and elaborates the information from the elementary centrum to the senior centrum. Normal cerebrum has a select and filter process for information of stimulus which is called SG. Cerebrum abstinently elaborates the information for the preliminary attention process and admits the significant information into the more senior centrum by means of SG. SG had been compared to a spigot which precisely regulated the amount of sensory stimulations into the cerebrum and avoids sensory overload. A defect of SG will induce the overload of indifferent stimulus and all kinds of psychiatric symptoms correlated with attention.
P50 is an evoked potential of middle latency, which is a normal phase wave emerging at 30–90 ms after auditory stimulus. P50 using auditory conditioning stimulus (S1)–testing stimulus (S2) paradigm is essentially a double pulse suppression which validating the SG phenomenon. The ratio of S2/S1 is called P50 suppression. P50 suppression is known as a reliable indicator to measure SG.12
Our study showed that there were significant differences among the three groups in the latency and amplitude of S2-P50 after the influence of age and course of disease were ignored. Increased S2-P50 amplitude was shown in the TRD and NTRD groups compared with the HC group, but no difference on S2-P50 amplitude between TRD and NTRD groups. When the normal cerebrum accepts paired stimulations with certain intervals and identical qualities, its reaction to the second stimulations (S2) is suppressed. Because the second stimulations are indifferent stimulations which have not brought new information to the centrum, they are suppressed so that they will not enter the cenbrum excessively. For example, if you lived near a street, you would always be disturbed by the noises at first. But several days later, you would not be bothered with that any more if your SG function was good. Increased S2-P50 amplitude in the TRD and NTRD groups suggested that both TRD and NTRD patients had defects in the ability of early information processing at the attention stage, the focus on outer stimulus were weakened and a many indifferent stimulations (S2) were registered in the cerebrum causing the sensory overload This can interfere with the normal cognitive process and produce psychiatric symptoms. This result was consistent with other studies on schizophrenia patients.7-9 Experiments showed that some kind of disposed neuromechanism and abnormal dopaminergic and cholinergic systems which resided in hippocampus, thalamencephal and pallium played important roles in changes of SG function.13
In clinical work, we found that some patients showed an insufficient reaction to something which was important to them. We supposed that they should have deficits on amplitude of S1-P50 as well as S2-P50. Based on this assumption, we tried to find some expressions of S1 and S2 to reflect SG function more sensitively. Smith et al14 investigated the expressions of P50 on SG and found that the ratio of S2/S1 was most valuable to discriminate patients and health adults. Our study showed that there was no significant difference among the three groups in latency and amplitude of S1-P50 which suggested that TRD and NTRD patients might not have deficits in reactions to important stimulations. This result was not consistent with other studies on schizophrenia patients.7-9 There were significant differences in S1–S2, S2/S1 and 100 × (1–S2/S1) among the three groups, and also differences in S2/S1 and 100 × (1–S2/S1) between each two of the three groups. It was clear that both the TRD and NTRD groups had significant deficits in SG function and the deficits were more severe in TRD patients. So we thought the combination of the three expressions could more intuitively show the degree of damage to the cerebrum P50 function.
There are now molecular genetics studies on P50 deficit of SG. Chen et al15 investigated the changes of SG function before and after treatments in first episode schizophrenia patients and considered that the P50 variations were possible early changes in schizophrenia patients which had the characteristics of trait markers. Another study had proven that there was a genetic linkage between the deficits of P50 suppression and the alpha-7 nicotinic receptors (α7-NR), which are the subtypes of cholinergic receptors.16 Schulze et al17 investigated bipolar disorder patients who had a history of psychotic symptoms and their healthy relatives, and found that the SG P50 deficit was possibly an inner-phenotype of psychotic bipolar disorder which might reflect the influence of gene susceptibility to the psychotic symptoms. However, we found the TRD patients always had psychotic symptoms, and whether the SG P50 deficit was an inner-phenotype of TRD or an inner-phenotype of psychotic symptoms is still unknown. Some scholars thought that the deficits of P50 suppression were state markers correlated with the symptoms of mood disorders, and were trait markers correlated with schizophrenia.18 If they are trait markers they will not change with the clinical symptoms, and if they are state markers they will change with the clinical symptoms.19
This study showed that TRD patients had no significant changes in their P50 measurements and different P50 expression after treatment, which suggested that the deficits of P50 suppression were possibly trait markers of TRD. So the correlations between three P50 expressions and the scores of HAMD-17 from 36 TRD patients were evaluated using the data before and after medication. Before medication, there was a positive correlation between S2/S1 and the scores of HAMD-17, and both S1–S2 and 100 × (1–S2/S1) also showed a negative correlation with the scores of HAMD-17. After medication, there was no correlation between the three P50 expressions and the scores of HAMD-17. From these results we concluded that the severity of SG P50 deficits in TRD patients might correlate with the severity of their clinical symptoms. Currently, venlafaxine has become an acceptable first-line treatment choice,20 especially in more severe depressive patients. So we chose it to treat TRD patients in this study. But this drug did not achieve
the anticipated effect, and there was no significant change in the scores of HAMD-17 after treatment. It is uncertain whether P50 deficits are trait markers or state markers of TRD and whether they could be relieve with the improvement of clinical symptoms, a point which should be further investigated.21
1. Cassano P, Fava M. Depression and public health: an overview. J Psychosom Res 2002; 53: 849-857.
2. Lu J, Ruan Y, Huang Y, Yao J, Dang W, Gao C. Major depression in Kunming: prevalence, correlates and co-morbidity in a south-western city of China. J Affect Disord 2008; 111: 221-226.
3. Hwu HG, Chang IH, Yeh EK, Chang CJ, Yeh LL. Major depressive disorder in Taiwan defined by the Chinese diagnostic Interview Schedule. J Nerv Ment Dis 1996; 184: 497-502.
4. Fava M, Davidson KG. Definition and epidemiology of treatment-resistant depression. Psychiatr Clin North Am 1996; 19: 179-200.
5. Golden RN, Nemeroff CB, McSorley P, Pitts CD, Dubé EM. Efficacy and tolerability of controlled-release and immediate-release paroxetine in the treatment of depression. J Clin Psychiatry 2002; 63: 577-584.
6. Thase ME, Rush AJ. When at first you don't succeed: sequential strategies for antidepressant nonresponders. J Clin Psychiatry 1997; 58: 23-29.
7. Light GA, Braff DL. Do self-reports of perceptual anomalies reflect gating deficits in schizophrenia patients. Biol Psychiatry 2000; 47: 463-467.
8. Light GA, Geyer MA, Clementz BA, Cadenhead KS, Braff DL. Normal P50 suppression in schizophrenia patients treated with atypical antipsychotic medications. Am J Psychiatry 2000; 157: 767-771.
9. Olincy A, Ross RG, Harris JG, Young DA, McAndrews MA, Cawthra E, et al. The P50 auditory event-evoked potential in adult attention-deficit disorder: comparison with schizophrenia. Biol Psychiatry 2000; 47: 969-977.
10. Olincy A, Martin L. Diminished suppression of the P50 auditory evoked potential in bipolar disorder subjects with a history of psychosis. Am J Psychiatry 2005; 162: 43-49.
11. Jiang KD, Su H, Chen XS, Lou FY, Liang JH, Chen XW. Sensory gating deficits assessed by the P50 event-related potential in subjects with first-episode depression. Chin J Psychiatry (Chin) 2006; 39: 197-200.
12. Chen XS, Wang HX, Zhang MD. Study advancements on sensory gating theory and animal model of schizophrenia. Chin J Psychiatry (Chin) 2006; 39: 247-249.
13. Liu QR, Huang BC, Sun JL. Practical clinical electrophysiology. Beijing: Chinese Scientific & Technology Publishing; 2006: 95-108.
14. Smith DA, Boutros NN, Schwarzkopf SB. Reliability of P50 auditory event-related potential indices of sensory gating. Psychophysiology 1994; 31: 495-502.
15. Chen XS, Zhang MD, Lou FY, Wang HX, Wang JJ, Guo Q, et al. Follow-up study on sensory gating in first-episode schizophrenia. Chin J Psychiatry (Chin) 2006; 39: 193-196.
16. Freedman R, Coon H, Myles-Worsley M, Orr-Urtreger A, Olincy A, Davis A, et al. Linkage of a neurophysiological deficit in schizophrenia to a chromosome 15 locus. Proc Natl Acad Sci U S A 1997; 94: 587-592.
17. Schulze KK, Hall MH, McDonald C, Marshall N, Walshe M, Murray RM, et al. P50 auditory evoked potential suppression in bipolar disorder patients with psychotic features and their unaffected relatives. Biol Psychiatry 2007; 62: 121-128.
18. Light GA, Malaspina D, Geyer MA, Luber BM, Coleman EA, Sackeim HA, et al. Amphetamine disrupts P50 suppression in normal subjects. Biol Psychiatry 1999; 46: 990-996.
19. Chen XS, Lu YZ, Wang JJ, Wang HX, Zhang MD, Lou FY, et al. Relationship between event-related potential P300 and first episode schizophrenia. Chin Med J 2007; 120: 339-341.
20. Jain R. Single-action versus dual-action antidepressants. Prim Care Companion J Clin Psychiatry 2004; 6: 7-11.
21. Wang Y, Fang YR, Chen XS. Study advancements on sensory gating P50 of mental disorders. J Neurol Neurorehabilitation 2007; 4: 199-200.