Duration of untreated bipolar disorder: a multicenter study

Little is known about the demographic and clinical differences between short and long duration of untreated bipolar disorder (DUB) in Chinese patients. This study examined the demographic and clinical features of short (≤2 years) and long DUB (>2 years) in China. A consecutively recruited sample of 555 patients with bipolar disorder (BD) was examined in 7 psychiatric hospitals and general hospital psychiatric units across China. Patients’ demographic and clinical characteristics were collected using a standardized protocol and data collection procedure. The mean DUB was 3.2 ± 6.0 years; long DUB accounted for 31.0% of the sample. Multivariate analyses revealed that longer duration of illness, diagnosis of BD type II, and earlier misdiagnosis of BD for major depressive disorder or schizophrenia were independently associated with long DUB. The mean DUB in Chinese BD patients was shorter than the reported figures from Western countries. The long-term impact of DUB on the outcome of BD is warranted.

Assessment procedure. All patients with a diagnosis of BD receiving treatment in the participating hospitals/units were invited to take part in the study and if agreed, they were referred by their treating psychiatrists to the investigators. Patients satisfying the above entry criteria were invited to participate in the study. Patients' basic demographic and clinical data were collected with a form designed for the study in the course of a clinical interview supplemented by a review of the medical records. DUB was defined as the interval between the onset of the first mood episode and the first treatment with a mood stabilizer 4,5 . Mood stabilizers prescribed included lithium, valproate, carbamazepine and lamotrigine. There is no widely accepted gold standard threshold for defining short/long DUB. Following earlier studies 3 and the recommendation of the Chinese Society of Psychiatry Expert Committee, in this study the 2-year cutoff point differentiated between short (≤ 2 years) and long DUB (> 2 years).
Statistical analysis. The data were analyzed using SPSS Version 21.0 for Windows. Comparison of the socio-demographic and clinical characteristics between the short and long DUB groups was made using independent sample t-test, Mann-Whitney U test, and chi-square test, as appropriate. Multiple logistic regression analysis with the "Enter" method was used to determine the demographic and clinical variables that were independently associated with long DUB. Long DUB was the dependent variable, while variables that significantly differed between the two groups in the above univariate analyses were entered as independent variables. The one-sample Kolmogorov-Smirnov test was used to check the normality of the distribution of the continuous variables. The level of significance set at 0.05 (two-tailed).

Results
Altogether 555 patients were invited to participate in the study; 35 (6.3%) refused thus 520 (93.7%) fulfilled the study entry criteria and formed the study sample. The mean DUB was 3.2 ± 6.0 years. Long DUB accounted for 31.0% of the sample. Table 1 presents the basic demographic and clinical characteristics for the whole sample and separately for patients with short and long DUB. Patients in the long DUB group were older, had a longer duration of illness, more family history of mood disorders, more likely to have BD-II and an earlier misdiagnosis of BD for major depressive disorder or schizophrenia.

Discussion
To the best of our knowledge, this was the first multicenter study in China that compared the demographic and clinical features between BD patients with short and long DUB. The main finding is that Chinese BD patients had a mean DUB of 3.2 years; 31.0% of the sample had long DUB. Long DUB was significantly associated with longer duration of illness, diagnosis of BD type II, and earlier misdiagnosis of BD for major depressive disorder or schizophrenia. The rate of participation in the study was high, reflecting a good relationship between patients and their treating psychiatrists. Respect for the medical profession is still rather high in China -although it is being eroded -and patients are unlikely to refuse a request from their doctors.
The mean DUB in this study was shorter than figures of 6.7-20 years reported from Western countries [3][4][5]7,17 , but similar to the finding of 3.9 years reported from China 18 . The different definitions of DUB may partly account for the inconsistency of the results. For example, studies used the interval between the first symptoms and diagnosis, but others referred to the time between the first episode and any treatment or prescription of mood stabilizers in defining DUB 4 . Furthermore, there were more patients with BD-I than with BD-II in this sample (76.7% vs 23.3%), which would shorten DUB as BP-I patients are a great deal easier to identify 3,19 . In addition, the discrepancy in DUB between studies may be due to the differences in sampling methods, the distribution of bipolar subtypes and the type diagnostic assessment. All participating psychiatric units or hospitals in this study are located in major cities thus patients could easily access mental health services. In addition, psychiatrists in major centres are usually better trained to ascertain BD resulting in relatively short DUB.
In multivariate analyses DUB was independently associated with several clinical variables. Patients with longer DUB had a longer duration of illness, which is consistent with earlier findings 3 . BD-II has been a major contributing factor to long DUB 4,19,20 , which is also confirmed in this study. Hypomania is often perceived as normal Scientific RepoRts | 7:44811 | DOI: 10.1038/srep44811 condition by patients and their families, therefore it is not spontaneously reported 21,22 . Furthermore, depressive episodes usually last longer than hypomania in BD-II, thus patients seek help for depression more often than for hypomania 23 . Clinicians sometimes overlook the history of hypomania if patients present with a depressive episode. All these factors contribute to the failure to diagnose BD-II leading to an extended DUB.
Multiple episodes of major depression frequently occur prior to the first episode of hypomania or mania 24,25 . In addition, depressive phases occur more frequently than hypomanic or manic phases 26 . Therefore BD is frequently misdiagnosed for major depressive disorder 21,25,27 , which, in turn, leads to longer DUB as found in this study. Similarly, BD could manifest initially with psychotic symptoms prior to the first signs of mania that would suggest the diagnosis of schizophrenia, which is another reason for longer DUB 28 . This is consistent with the finding of this study. Early onset of BD (≤ 21 years) was reported to be a risk factor for long DUB 4,29-32 , but this was not confirmed in this study.
The major strength of this study is its large, multi-center sample. However, the results should be interpreted with caution due to several methodological limitations. First, similar to most studies, some of the data were collected retrospectively, such as onset age and that of the first episode, thus potential recall bias could not be avoided. Due to logistical reasons, some important variables, such as genetic and neuroimaging data, could not be collected. Second, due to the cross-sectional study design, the causality between DUB and other variables could not be examined. Third, second-generation antipsychotics (SGAs) show certain mood-stabilizing properties. Considering that the purpose of the SGA prescription were difficult to identify in cross-sectional surveys, only lithium, sodium valproate, carbamazepine and lamotrigine were classified as mood stabilizers. Fourth, there is no widely accepted consensus

Variables
The whole sample (n = 520) DUB ≤ 2 years (n = 359) DUB > 2 years (n = 161) P-value  Table 2. Demographic and clinical characteristics independently associated with long DUB by multiple logistic regression analysis. Study sites were controlled for. Total model: P < 0.001, R 2 = 0.427.
regarding the definition of short/long DUB. Finally, following earlier studies 4,5 prescription of mood stabilizers identified DUB although mood stabilizers do not necessarily indicate the diagnosis of BD.
In conclusion, the mean DUB in Chinese BD patients was shorter than the figures in the literature. The differences between long and short DUB with respect to Chinese BD patients' demographic and clinical characteristics were consistent with those found in Western clinical settings. Prospective studies examining the long-term clinical outcome of DUB are warranted.