张向阳教授:《如何利用临床资料发表5分以上的文章》(下)

《如何利用临床资料发表5分以上的文章》（下）

------暨如何答复审稿者意见

在上两篇文章（点击阅读：《如何利用临床资料发表5分以上的文章》（上）和《如何利用临床资料发表5分以上的文章》（中））里，我们通过实例分析讲解了如何利用单纯的临床资料，发表高影响因子文章，尤其是如何答复审稿人的方法和技巧。

发表SCI文章，尤其是高分的文章，答复好审稿人的意见至关重要，常常需要与审稿人“斗智斗勇”。审稿人形形色色，有的比较友善，有的比较苛刻和挑剔，有一些基本原则（也是我个人的经验之谈）可以遵循：

（1）逐条答复：对审稿人提出的每一点意见/建议都专门列出来，逐条详细答复，不留死角。千万记得一点：确实审稿者的某个或一些意见很难答复，不能抱有侥幸心理不答复，这样就可能留给审稿者一种不诚实的印象。

（2）如何对待审稿人的苛刻、甚至错误的意见：常常遇到这种情况，审稿人好像不了解我们的研究内容，提出的意见不准确甚至是错误的，而一些审稿者很苛刻，让人感到是“鸡蛋里挑骨头”，甚至是攻击我们文章的结果或论点，这些都让我们感到审稿人很武断和随意，甚至还有敌意。确实科学界也有“歧视”，也有“对手和敌人”，也有潜规则。即便如此，我们答复审稿人意见时，一定要抱着非常虚心、平和的态度，冷静下来，试着去想想为什么审稿人会在这一点上找麻烦？作为作者，如果能够站在审稿者的位置，想出审稿人为什么会提出这个意见，那么就能做出很好的答复了

我们的目的是让审稿者和编辑接受我们的story，最终让我们的文章被接受和发表。如果我们只是采用负性情绪去答复审稿者意见，那最终审稿者和编辑会采用同样的方式对待我们的稿件，真是因小失大、得不偿失！为了让我们的论文尽快接受和发表，一定要想到我们老祖宗的智慧“小不忍则乱大谋”。

下面继续讲解如何答复第三位审稿人的意见，希望大家先看懂审稿者的问题（尤其是背后的含义），然后再仔细阅读我的答复。

Question 1：A possible weakness of this paper is the lack of assessment of lethality which may be important to classify suicide attempts as more or less serious.
Answer：This point is excellent. Indeed, the lack of assessment of lethality which may be important to classify suicide attempts as more or less serious is a weakness of this study, which has been added as one of the limitations in the last paragraph of the Discussion showing as “Fourth, the suicide attempt data was collected from medical case notes and interviews with the patients and their family members rather than measured by a structured instrument at the time of the suicide attempt. Moreover, we did not assess the severity of suicide attempts, which is important to examine the clinical correlates with lethality”.
（点评：这位审稿人提出来一个很好、很关键的问题：本文对精神分裂症行为没采用任何量表评定，只是简单问病人的几个相关问题。确实这是个明显的设计不足，乖乖地承认limitation 吧。）

Question 2：Definition of first episode: it is not clear what the definition for first episode is. Is it first symptom onset, or first hospitalization?
Answer：The definition of first episode has been provided in the Methods section on Page 6, showing as “In addition, first episode was defined as first symptom onset in this study.”
（点评：这位审稿人提出一个简单、但很重要的问题：如何定义首发精神分裂症。）

Question 3：The authors found that the attempters were more likely to smoke. However, the definition of smoking was not provided, such as a cut-off of 1 or more smoked cigarette every day. Also, please clarify if ex-smokers were grouped together with current smokers to form a large group of smokers. If so, the authors should also provide statistics in the result section for three groups: smokers, ex-smokers and non-smokers.
Answer：This point is very excellent. We  have provided the detailed definition of smoking in the Methods section on Page 7, showing as “In our current study, we took the following definitions for smoking behavior45: Current smokers—Adults who have smoked 100 cigarettes in their lifetime and currently smoke cigarettes every day (daily) or some days (nondaily). Former or ex-smokers—Adults who have smoked at least 100 cigarettes in their lifetime, but they currently do not smoke.  Never smokers—Adults who have never smoked a cigarette or who smoked fewer than 100 cigarettes in their entire lifetime. In our present study, former smokers were excluded, and only never smokers were included as non-smokers”.
（点评：因为我们本文发现企图组的吸烟率较高，说明吸烟可能是促进的危险因素，所以这位审稿人提出一个很重要的问题：如何定义吸烟？ 同时对那些戒烟组是如何处理的？ 纳入的统计分析了吗？）

Question 4：Previous study showed that schizophrenia patients who had suicide attempt had better executive functioning. However, RBANS- the cognitive tool used in this study does not have executive domain.
Answer：This point is very excellent. Indeed, previous study showed that schizophrenia patients who had suicide attempt had better executive functioning. Unfortunately, RBANS does not have executive domain and we did not use the other cognitive test to capture executive functioning. This point has been added to the last paragraph of the Discussion section showing as “Second, we only used the RBANS to investigate the following neurocognitive characteristics: memory, visuospatial/ constructional, language, attention, and delayed memory, but without additional neurocognitive features that have been explored. While the RBANS has been validated as a reliable screening test for cognitive impairment, it is not designed to be sensitive for cognitive impairment specific to schizophrenia, particularly in domains such as executive functioning which has been reported to be closely associated with suicide in schizophrenia 30,31. Thus, there may be important differences in cognitive functioning that were not captured with the RBANS, which is one of the main methodological limitations of our current study”.
（点评：这位审稿者提出来类似上面第一位审稿人的问题：认知检测量表RBANS没能检测到执行功能，所以我们只能把此点作为limitation写到讨论的倒数第二段。）

下面是答复稿全文， 其中蓝色字体是根据审稿人的问题/意见加入的新内容：

Prevalence, clinical correlates and cognitive function of suicide attempts in first-episode and drug naïve patients with schizophrenia

Abstract
Background: It is well-established that patients with chronic schizophrenia have a substantially higher rate of attempted and completed suicide than the general population. However, actual prevalence at first-episode psychosis is relatively unknown. Previous studies show that suicidal schizophrenia patients demonstrate higher cognitive function than non-suicidal patients, with mixing results. The aims of the study were to examine suicide attempt prevalence and its associations with demographic, clinical variables and cognitive function in Chinese first-episode, drug-naïve (FEDN) schizophrenia patients using a cross-sectional and case-control design.

Method: 357 FEDN inpatients meeting DSM-IV diagnosis of schizophrenia and 380 healthy controls were enrolled and completed a detailed in-house questionnaire. The suicide attempt data were collected from the medical record and interviews with the patients and their family members. The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) was administrated for cognition from the 28 patients with and 95 without a history of suicide attempt and 151 healthy controls. Also, patients were rated on the Positive and Negative Symptom Scale (PANSS). This study was conducted from June 2013 to December 2015.

Results: We found a suicide attempt rate of 12.0% in these first episode schizophrenia inpatients. The attempters were more likely to smoke (34.4% vs. 17.9%; x2=5.48, p=0.019), and had lower severity of negative symptoms (F=4.12, df=1,354, p=0.043) as compared to FEDN patients without a suicide attempt. All five RBANS subscales (all<0.001) except for the Visuospatial/Constructional (p>0.05) index showed significantly lower cognitive performance for FEDN patients than healthy controls. Among the FEDN patients, the suicide attempters performed better than non-attempters on attention (F=5.12, df =1,121, p=0.025) with an effect size of 0.49. The following variables were independently associated with suicide attempt under multivariate regression analysis: the PANSS negative symptom (Wald x2=7.90 df=1, p=0.005; adjusted OR=0.807, 95% CI: 0.696~0.936) and Attention (Wald x2=4.69, df=1, p=0.03; adjusted OR=0.957, 95% CI: 0.918~0.997).

Conclusions: FEDN patients with schizophrenia attempt suicide more often than the general population. The suicidal patients were more likely to smoke, had reduced severity of negative symptoms, and showed better attention.

Keywords: Schizophrenia; suicide; prevalence; risk factor; first-episode drug naïve; attention

Introduction

People with schizophrenia are more likely to attempt suicide 1,2 and have early mortality 3. The lifetime prevalence of suicide in schizophrenia ranges from 20 to 50% 4-7, 8,9. Schizophrenia is estimated to reduce life expectancy by 10~20 years, and suicide is the biggest single cause of premature death that contributes to this shortened life span 2,10. Risk factors for suicide common to both schizophrenia patients and the general population include: female gender, higher levels of education, single or divorced, living alone, lower socioeconomic status, symptoms of depression and anxiety, early adversities, history of previous suicide attempts and family history of suicide 11, 12. Especially, hopelessness has been widely recognized as an independent risk factor for suicide and suicidal-behaviors in both patients with schizophrenia and the general population 13, 14. Also, hopelessness has been documents as an important risk factor for suicidality in patients with physical diseases 15. Risk factors for suicide specific to schizophrenia include: younger age at onset, longer duration of untreated psychosis (DUP), increased positive and decreased negative symptoms, a greater level of insight 3,16, non-compliance with antipsychotic medication 17, higher premorbid IQ, or fewer cognitive deficits 18,19. Further, substance use has been considered as important predictor for suicide and suicide-related behaviors 20, 21. Noticeably, the presence of cannabis use/dependence has been reported to enhance suicide risk in schizophrenia patients, especially in adolescence. Cannabis use/ dependence may be critically involved in suicidal behavior among psychotic youths with those attempting or completed suicide who reported some additional risk factors for suicide such as mood disorders, stressful life events, interpersonal problems, poor social support, lonely lives, and feelings of hopelessness 20. In addition, a history of prior suicide attempt is strongly predictive of future completed suicide, and the majority of schizophrenia patients who commit suicide have at least one previous attempt 22. However, the ability to appropriately screen for and prevent suicidality in schizophrenia patients is difficult, because of the incomplete knowledge about its phenomenology in this special patient population.

Until recently, most studies reporting an elevated prevalence of suicide were conducted with chronic and medicated schizophrenia patients 23. In a recent study, we found that chronic and medicated patients with schizophrenia had 9.2% prevalence of suicide attempts in a Chinese Han population 24. Only a few studies have reported suicide in the first episode of psychosis (FEP) 9,25. This is a critical period in the disease, as the rate of suicide-related mortality is higher among subjects recently diagnosed with schizophrenia 9, 26, and suicide risk peaks in the first few years after onset 7. The suicide risk is two-fold higher at the onset of psychotic illness than later on in the illness 8. Thus, it is of great importance to investigate the relevant risk factors in this population, particularly since early intervention and therapy may improve the suicidality among FEP patients 9, 27.

Schizophrenia patients display cognitive impairments across a number of domains, including learning, memory, attention, working memory, information processing, social cognition, and executive function 28, 29. Previous studies show that suicidal schizophrenia patients demonstrate higher cognitive function than non-suicidal patients 18, 30. A recent study showed that patients with schizophrenia or schizoaffective disorder who had lifetime history of suicide attempt had better executive functioning and good insight 31. However, other studies reported that suicide intent in patients with schizophrenia and schizoaffective disorder was not correlated with cognition 22, 32,33 or with lower cognitive function 34. Thus, there is some controversy on this issue and further investigation which more closely identifies the cognitive function at the time of suicide attempt will provide resolution. Interestingly, neurocognitive dysfunctions are usually present even in the prodromal phases of schizophrenia patients. Recent studies suggested that neurocognitive impairments, especially in attention and working memory abilities and declarative memory abilities, may be considered a robust characteristic of high-risk (CHR) state of psychosis, especially in individuals who later develop psychosis 35. Thus, interventions targeting the enhancement of neurocognitive functioning are warranted even in the prodromal phases of the illness for prevention of suicidal behaviors in this population.

Only a few systematic studies have investigated the socio-demographic and clinical correlates of suicide attempts in Chinese schizophrenia patients 24. Those which evaluated the prevalence of suicide attempt and demographic and clinical risk factors in chronic patients with schizophrenia have mixed results 22, 24,36, 37. None of these, however, have evaluated first-episode psychosis in drug naïve patients, which, as discussed above, represent a particularly high risk population. On the other hand, there exist the complex inter-relationships between antipsychotic medication, clinical symptoms, cognitive deficits and suicidality in chronic patients with schizophrenia. For example, more negative symptoms were found to be associated with suicide 7; however, there were the opposite findings 38, with conflicting results 39-41. Also, literature has reported mixed results regarding the association between antipsychotic treatment and suicide-related behavior in schizophrenia 42. For example, some antipsychotics especially classified as second generation such as clozapine were reported to have anti-suicide effect 43; however, others found a higher suicide risk among patients who were on antipsychotics, possibly due to side effects, such as extrapyramidal symptoms and tardive dyskinesia 44. Thus, studying suicide in first-episode and unmediated patients can disentangle between an antisuicide effect due to antipsychotic medication and an antisuicide effect of the characteristic of the disease itself, and also minimize the potential impact of other confounders, such as illness duration and the psychiatric and medical comorbidities that are associated with chronic illness. In this study, we recruited a large sample of first-episode and drug naïve (FEDN) patients with schizophrenia in a Chinese Han population (n= 357) to investigate the prevalence of suicide attempts among FEDN Chinese schizophrenia patients and to determine the socio-demographic, clinical correlates and cognition of suicide attempts.

Method
Subjects 

Three hundreds and fifty-seven (male/female=173/184) FEDN inpatients were recruited in the Beijing Hui-Long-Guan hospital, a city owned psychiatric hospital from June 2013 to December 2015. All hospital admissions were screened for patients who met the following 6 criteria: (1) an acute episode that met DSM–IV criteria for schizophrenia, as assessed by two independent experienced psychiatrists using the Structured Clinical Interview for DSM-IV (SCID) at study intake; (2) these patients were followed for 3 months as inpatients after admission and included if only the second 3-month evaluation was consistent in diagnosing schizophrenia; (3) the duration of symptoms was not longer than 60 months at admission; (4) patients had no prior treatment with antipsychotic medication; (5) patients were Han Chinese and between 18 and 45 years of age; (6) the current psychotic symptoms were of moderate severity or greater as measured with the Clinical Global Impression – Severity scale (CGI-S) score ≥4.  Exclusion criteria for the first episode schizophrenia patients were: (1) with current major medical problems; (b) history of any brain diseases; (c) family history of neurologic disorder; (d) lifetime history of alcohol or substance dependence except tobacco, or having alcohol or substance abuse within the past 6 months before the study; (e) the patient refused to provide written informed consent and take part in cognitive assessment. In addition, first episode was defined as first symptom onset in this study. 

Three hundred and eighty normal controls (male/female=179/201) with age ranging from 18 to 45 years were recruited randomly from the local community in Beijing during the same period. All participants were recruited without any selection, and they were interviewed by trained investigators supervised by a research psychiatrist. Current mental status and personal or family history of any mental disorder was assessed. No controls had either a personal or family history of a psychiatric disorder.

We obtained a complete medical history, physical examination and laboratory tests from all subjects. They were in good physical health, and any subjects with medical illnesses or drug and alcohol abuse/dependence except tobacco were excluded. The Institutional Review Board for the Beijing Hui-Long-Guan hospital approved the research protocol, and all subjects provided informed consent.

Socio-demographic characteristics

Research staff administered a detailed questionnaire that asked for the general information, socio-demographic characteristics, smoking and alcohol drinking behavior, and history of suicide attempts. In our current study, we took the following definitions for smoking behavior45: Current smokers—Adults who have smoked 100 cigarettes in their lifetime and currently smoke cigarettes every day (daily) or some days (nondaily). Former or ex-smokers—Adults who have smoked at least 100 cigarettes in their lifetime, but they currently do not smoke.  Never smokers—Adults who have never smoked a cigarette or who smoked fewer than 100 cigarettes in their entire lifetime. In our present study, former smokers were excluded, and only never smokers were included as non-smokers. 

According to the report of the World Health Organization (WHO), the outcome-based term “fatal suicidal behavior” or “completed suicide” has been proposed for suicidal acts that results in death, and similarly, “non-fatal suicidal behavior” for suicidal actions that do not result in death. Such actions are also often called “attempted suicide” 46. Hence, in this study, we defined a suicide attempt as an intentionally self destructive act performed with at least some intent to die 47, but not resulting in death. During the research interview all subjects were asked about previous suicide attempts with the following details: the number of attempts, the exact date for each suicide attempt, and the methods. The screening question was “In your entire lifetime did you ever attempt suicide?” Responses were coded as yes or no. History of suicide attempts were confirmed by a review of medical records supplemented by a clinical diagnostic interview of patients and, whenever possible, their relatives by a qualified psychiatrist. Additional visits to their family members, relatives or even friends and coworkers were requested for subjects with missing or ambiguous suicide-related data, such as the number of suicide attempts, the date for suicide attempt and the exact methods.  Since our study design was retrospective and the information for suicide attempt were collected from medical case notes and interviews with the patients and their family members, a possible recall bias could not be ruled out completely. To reduce the recall bias, we specifically collected these data related to suicidal behaviors.

Clinical assessment

Two experienced psychiatrists assessed the psychopathology and symptom severity of the patients using the Positive and Negative Syndrome Scale (PANSS) 48 and CGI 49. To ensure consistent and reliable ratings, the two psychiatrists simultaneously attended a training session in the use of the PANSS and CGI prior to the start of the study. After training, they maintained an intra-class correlation coefficient of greater than 0.8 on both the PANSS and the CGI total scores at repeated assessments during the course of this study.

In addition, the Chinese translation of the standardized Fagerstrom Test for Nicotine Dependence (FTND) was employed to measure the degree of nicotine dependence 50.
Cognitive performance 

We utilized the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) 51 for cognitive performance testing. The RBANS is comprised of 12 subtests that are used to calculate 5 age-adjusted index scores and a total score. The average score of the general population is 100 with a standard deviation of 15. Test indices include: Immediate Memory (comprised of List Learning and Story Memory tasks), Visuospatial/Constructional (comprised of Figure Copy and Line Orientation tasks), Language (comprised of Picture Naming and Semantic Fluency tasks), Attention (comprised of Digit Span and Coding tasks), and Delayed Memory (comprised of List Recall, Story Recall, Figure Recall, and List Recognition tasks). Our group previously translated RBANS into Chinese and established its clinical validity and test-retest reliability among patients with schizophrenia 52.

Statistical analysis

Group differences were compared using one way analysis of variance (ANOVA) for continuous variables and chi squared (x2) test for categorical variables. The prevalence of suicide attempt was analyzed by x2 test. The odds ratio (OR) for suicide attempt risk in schizophrenia as compared to healthy controls was derived from logistic regression analysis after controlling or socio-demographic factors. We compared RBANS scores among the three groups (suicide attempters vs. non-attempters vs. controls) using ANOVA. Fisher’s least significant difference (LSD) test was used to perform post-hoc pair-wise between-group comparisons. Bonferroni correction was used to adjust for multiple testing. When significance was found in ANOVA, the effects of the relevant variables were added to the logistic regression analysis model as covariates. Effect sizes were calculated for the two-way comparisons. In addition, a binary logistic regression analysis was conducted to assess which factors were most strongly associated with suicide attempts. All statistical analysis used SPSS version 15.0 with two tailed p values of 0.05.

Results    
Prevalence of suicide attempts in FEDN patients    

Table 1 shows that the patients did not differ from healthy controls with respect to demographic characteristics except for lower body mass index (BMI) in schizophrenia (p<0.001). The frequency of suicide attempt was 12.0% (43/357) for patients and 0.79% (3/380) for controls, with an OR of 17.21 (95% CI: 5.29–56.00; X2=39.8, p<0.001). This difference remained significant after using logistic regression to control for the socio-demographic confounders, including gender, age, education, smoking and BMI (X2=21.31,p<0.001, adjusted odds ratio=16.39; 95% CI: 4.98–53.26). Within the schizophrenia group, there was no significant difference in the frequency of suicide attempt between male and female patients (11.6% male vs. 12.5% female) (X2=0.07, p=0.79).

As we mentioned in the Introduction section, our previous study showed that the prevalence of suicide attempts was 9.2% (48/520) in the chronic and medicated patients with schizophrenia 24. Although the prevalence of suicide attempts appears higher in FEDN than chronic medicated patients (12.0% vs. 9.2%), there was no significant difference between these datasets (x2=1.80,df=1, p=0.18), possibly due to limited sample size.

Clinical characteristics between the patients with and without history of suicide attempt

Table 2 shows that smoking was more common in attempters (34.4%) than non-attempters (17.9%; x2=5.48, p=0.019). This difference remained significant after using logistic regression to adjust for clinical characteristics including gender, age, education, BMI and age of onset (x2=5.79, p=0.016; adjusted OR=3.33, 95% confidence interval, 1.25~8.86). Compared to non-attempters, attempters showed a significantly lower PANSS negative symptom subscore (F=4.12, df=1,354, p=0.043; Bonferroni corrected p>0.05). In stepwise multiple logistic regression analysis for associations with suicide attempts, we added the demographic and clinical data including sex, age, education, smoking and BMI, age of onset, and the PANSS positive, negative and general psychopathology subscale scores together into the model and found that the PANSS negative symptom (Wald x2=6.32 df=1, p=0.012; adjusted OR=0.836, 95% CI: 0.742~0.943) and smoking (Wald x2=5.31, df=1, p=0.021; adjusted OR=0.887; 0.797~0.986) remained significant.

Cognitive functioning in suicide attempters and non-attempters in FEDN patients

The RBANS total and index scores for the 28 patients with and 95 without a history of suicide attempt and 151 healthy controls are shown in Table 3.  The subjects from these three subgroups were matched for demographics including sex, age, education, smoking and BMI (all p>0.05). When comparing schizophrenia patients to healthy controls, one-way ANOVA showed significantly lower cognitive scores on the RBANS total and nearly all of its 5 subscale scores (all p<0.001; Bonferroni corrected p<0.01) except for Visuospatial/Constructional index (p>0.05), with effect sizes ranging from 0.74 to 1.13. 

In addition, we made clinical and demographic comparisons between patients who had the cognitive evaluation vs. the one who did not, to check for a selection bias that could have increased the protective role of poor cognition. However, there was no any significant difference in demographic variables including sex, age, education, smoking and BMI, age of onset, the PANSS total and its three subscale scores as well as the rate of suicide attempt  (all p>0.05). Further, both the suicide attempters and non-attempters showed significant lower cognitive scores on the RBANS total and 4 domain scores (all p<0.001; Bonferroni corrected p<0.01) except for Visuospatial/Constructional index (both p>0.05) (Table 3). When comparing suicide attempters to non-attempters within the schizophrenia group, attempters performed better on the Attention domain (F=5.12, df =1,121, p=0.025; Bonferroni corrected p>0.05), with an effect size of 0.49. However, there were no significant differences in the RBANS total and any other domain scores between the attempters and non-attempters (all p>0.05). When adding these data including sex, age, education, smoking and BMI, age of onset, and the PANSS positive, negative and general psychopathology subscale scores together into a single stepwise multiple logistic regression analysis of all risk factors evaluated, the PANSS negative symptom (Wald x2=7.90 df=1, p=0.005; adjusted OR=0.807, 95% CI: 0.696~0.936) and Attention (Wald x2=4.69, df=1, p=0.03; adjusted OR=0.957, 95% CI: 0.918~0.997) remained significant for suicide attempt.

Discussion  

In our present study, the frequency of suicide attempts in schizophrenia was 12%, which was exactly the same as the recent report of the Chinese community-dwelling schizophrenia in Beijing 37, close to 10.6% attempted suicide rate in first episode psychosis in Hong Kong 53, and close to a previous report with the rate of suicide attempts of 7.5% in Chinese schizophrenia outpatients of a rural community 22. A recent epidemiological survey showed that the overall prevalence of suicidal ideation, plans and attempts was 2.8, 1.6 and 1.3% in a rural sample, and 1.8, 1.3 and 0.9% in an urban sample, respectively in the Beijing Municipality 54. Thus, our results showed that the prevalence of lifetime suicide attempt in the FEDN patients with schizophrenia was appropriately 11 times greater than in the general population in Beijing (12.0% vs.1.1%), suggesting a marked increased prevalence of suicide attempts in both FEDN and chronic medicated patients (9.2%) with schizophrenia in the Chinese population. However, most reports on suicide attempts in FEDN or chronic medicated schizophrenia patients among Chinese show lower prevalence as compared to the 20%-50% prevalence estimated in Western countries 6,7.

Interestingly, suicide-related mortality is higher among subjects recently diagnosed with schizophrenia (≤5years from diagnosis) 26. This risk appears to peak in the first year of illness and then steadily decline over the following years 9. Recently, a Norwegian study showed that in the early phases of first episode of psychosis, 38.8% of patients reported suicidal ideation and 25.9% attempted suicide before any treatment 33. In our present study, although we found a higher prevalence of suicide attempts in FEDN than chronic medicated patients (12.0% vs. 9.2%), there was no significant difference between them, suggesting that there is no significant change in suicide attempt rate over the progression of the illness course, and most attempts occur early in the disease.

In our present study, we found that smoking and reduced negative symptoms were associated with suicide attempt in FEDN patients. This corroborates our previous study where nicotine dependence severity was associated with suicide attempts in chronic, medicated schizophrenia patients 24, suggesting that this is a risk factor in both categories. Several epidemiological studies have reported an association between smoking and suicidal behaviors, but the results are inconsistent 55. However, a recent meta-analysis has demonstrated that smoking is associated with an increased risk of suicidal behaviors in general 56. In chronic patients with schizophrenia, several studies also showed increased suicidal risk among smoking schizophrenia patients 47. Taken together, these studies suggest that smoking may be a contributing factor for suicide in schizophrenia.

Further, our study found that lower negative symptom correlated with an increased risk of suicide attempt, and that positive symptoms had no association. Many studies found that the negative symptoms of schizophrenia may increase suicidal experience in psychotic individuals; however, the results have been inconsistent 39-41. For example, earlier studies found that individuals with chronic schizophrenia and other psychotic disorders at risk for suicide exhibited fewer negative symptoms 38. Further, a recent study showed that more negative symptoms trended towards significantly lower suicide risk in recent onset psychosis 7. On the other hand, increased negative symptoms significantly correlate with severity and intensity of recent suicidal ideation even after adjustment for depression scores 57. Our current study showed that the negative symptom was significantly lower in suicide attempters, suggesting that negative symptoms may be a protective factor for suicide attempt risk in FEDN patients with schizophrenia. It is hard to provide a reasonable explanation for the lower negative symptoms in suicide attempters in first-episode schizophrenia patients found in our current study. We speculate that higher negative symptoms are associated with the progressive loss of social drive, the diminished capacity to experience affect, and the indifference toward future 57, which are markedly disabling, but may decrease or even eliminate the painful hopelessness, self-awareness and stigma that are associated with suicide-related behaviors. Finally, there was no effect for positive symptoms in our current study, which is consistent with previous findings for positive symptoms 7, 58,59. A few studies have reported a significant association, but this does not to appear to be a strong relationship 18, 60. Taken together, whether positive or negative symptoms show a significant association with suicide-related behaviors is still mixed.  The possible explanations related to this mixed finding may be due to the patients in different stage of disease progression (acute vs chronic or active phase vs remission), different age at onset or illness duration,  exposure to antipsychotic treatment (naive vs medicated), or different ethnic origin or genetic background of the patients studied. Therefore, the relationship of clinical symptoms and suicide in patients with schizophrenia deserves a further investigation in a large sample size of FEDN patients using a longitudinal design.  

Interestingly, we found that patients who had a history of suicidal attempt outperformed non-suicide attempters in the RBANS Attention domain, which was consistent with the findings of earlier studies 18, 30-31. For example, a previous study found an association between suicidality and better performance on attention, together with psychomotor speed, verbal working memory and executive function in schizophrenia patients 30. A study by Nangle et al 18 similarly observed that schizophrenia patients who had previously attempted suicide performed better on attention and verbal fluency than non-attempters. Interestingly, the previous studies also found an association between suicidality and better performance on tests of executive function 18, 30. Two recent studies in this area have affirmed the role of executive functioning in increased suicidality 31, 61. Although we did not measure executive functioning directly in our present study, attention is an important component of executive function and thus can be seen as a proxy for it. Executive function relates to the capacity to shift attention from one stimulus to another, initiate or cease engaging in a given behavior, to evaluate risk, and to develop plans of action and to carry them out 18. Impairment in this supervisory attentional system can result in difficulties with goal formulation and an inability to plan effectively 18, 51. Thus, those schizophrenia patients with the relatively higher attention and executive functioning may have greater ability to formulate plans and initiate behavior directed towards suicide attempt 18.  However, some other studies failed to find significant relationship between cognitive performance and suicidal behavior in schizophrenia 19, 32,33. This discrepancy may arise from differences in techniques of measuring cognitive performance, differences in sampling of patients in different stages of disease progression (acute vs. chronic or active phase vs. remission), different illness courses (first-episode vs multiple episodes), exposure to different type, dosage and length of antipsychotics, different definitions of suicidal categories, or socio-cultural factors which have close relationship with suicidality. Taken together, these results show the complexity and difficulty in determining the relationship between suicidality and cognitive functioning in schizophrenia, which warrants further investigation in a large sample from the different ethnicities.

There are several limitations of this current study. First, this is the cross-sectional case-control nature of the study design and cannot show direct causality between suicide attempts and the risk factors in patients with schizophrenia. Thus, the main findings of our study should be regarded as the exploratory/preliminary nature. Second, we only used the RBANS to investigate the following neurocognitive characteristics: memory, visuospatial/ constructional, language, attention, and delayed memory, but without additional neurocognitive features that have been explored. While the RBANS has been validated as a reliable screening test for cognitive impairment, it is not designed to be sensitive for cognitive impairment specific to schizophrenia, particularly in domains such as executive functioning which has been reported to be closely associated with suicide in schizophrenia 30,31. Thus, there may be important differences in cognitive functioning that were not captured with the RBANS, which is one of the main methodological limitations of our current study. Third, premorbid IQ was not measured among the subjects in the study, which would have been prudent since premorbid IQ definitely has an impact on performance on cognitive tasks. Thus, the impact of premorbid IQ on cognitive performance should be adjusted in the future studies. Fourth, the suicide attempt data was collected from medical case notes and interviews with the patients and their family members rather than measured by a structured instrument at the time of the suicide attempt. Moreover, we did not assess the severity of suicide attempts, which is important to examine the clinical correlates with lethality.

Clinical Points:
•    The suicide attempt data were collected from 357 Chinese first-episode, drug-naïve schizophrenia and 380 healthy controls. Some of them were measured with cognition.
•    The patients showed higher suicide attempt rate of 12.0% and excessive cognitive impairments than controls.
•    The attempters were more likely to smoke, had lower severity of negative symptoms and better attention than non-attempters in patients.

References
1. Pompili M., Amador X. F., Girardi P., et al. Suicide risk in schizophrenia: learning from the past to change the future. Annals of General Psychiatry 2007; 6: 10
2. Balhara YP, Verma R. Schizophrenia and suicide. East Asian Arch Psychiatry. 2012; 22(3):126-133.
3. Hor K, Taylor M. Suicide and schizophrenia: a systematic review of rates and risk factors. J Psychopharmacol 2010; 24(4 Suppl):81-90.
4. Altamura A. C., Bassetti R., Bignotti S., Pioli R., Mundo E. Clinical variables related to suicide attempts in schizophrenic patients: a retrospective study. Schizophrenia Research 2003; 60(1): 47–55.
5.  Suokas J. T., Perälä J., Suominen K., Saarni S., Lönnqvist J., Suvisaari J. M. Epidemiology of suicide attempts among persons with psychotic disorder in the general population. Schizophrenia Research 2010; 124(1–3):22–28.
6. Aleman A, Denys D. Mental health: a road map for suicide research and prevention. Nature 2014; 509: 421–423
7. Castelein S, Liemburg EJ, de Lange JS, et al. Suicide in recent onset psychosis revisited: significant reduction of suicide rate over the last two decades – a replication study of a Dutch incidence cohort. PLoS One 2015; 10(6): e0129263.
8. Nordentoft M, Madsen T, Fedyszyn I. Suicidal behavior and mortality in first-episode psychosis. J Nerv Ment Dis 2015; 203(5): 387-392.
9. Ventriglio A, Gentile A, Bonfitto I, Stella E, Mari M, Steardo L, Bellomo A. Suicide in the early stage of schizophrenia. Front Psychiatry. 2016; 7:116.
10. Lopez-Morinigo JD, Ayesa-Arriola R, Torres-Romano B, et al. Risk assessment and suicide by patients with schizophrenia in secondary mental healthcare: a case-control study. BMJ Open 2016; 6(9):e011929.
11. Fuller-Thomson E, Hollister B. Schizophrenia and Suicide Attempts: Findings from a Representative Community-Based Canadian Sample. Schizophr Res Treatment. 2016; 2016: 3165243.
12. Hawton K., Sutton L., Haw C., Sinclair J., Deeks J. J. Schizophrenia and suicide: a systematic review of risk factors. The British Journal of Psychiatry 2005;187(1):9–20.
13. Aloba O, Esan O, Alimi T.  Adaptation of the Beck Hopelessness Scale as a suicide risk screening tool among Nigerian patients with schizophrenia. Int J Psychiatry Clin Pract. 2017;  doi: 10.1080/13651501.2017.1356928. [Epub ahead of print]
14. Cassidy RM, Yang F, Kapczinski F, Passos IC. Risk Factors for Suicidality in Patients With Schizophrenia: A Systematic Review, Meta-analysis, and Meta-regression of 96 Studies. Schizophr Bull. 2017. doi: 10.1093/schbul/sbx131. [Epub ahead of print]
15. Eskelinen M, Korhonen R, Selander T, Ollonen P. Suicidal ideation versus hopelessness/helplessness in healthy individuals and in patients with benign breast disease and breast cancer: a prospective case-control study in Finland. Anticancer Res. 2015; 35:3543–3551
16.  Barret EA, Mork E, Færden A, et al. The development of insight and its relationship with suicidality over one year follow-up in patients with first episode psychosis. Schizophr Res 2015; 162:97–102
17. Tiihonen J, Lönnqvist J, Wahlbeck K, Klaukka T, Tanskanen A, Haukka J. Antidepressants and the risk of suicide, attempted suicide, and overall mortality in a nationwide cohort. Arch Gen Psychiatry 2006; 63(12):1358-1367.
18. Nangle JM, Clarke S, Morris DW,  et al. Neurocognition and suicidal behaviour in an Irish population with major psychotic disorders. Schizophr Res 2006; 85: 196–200.
19. Zoghbi AW, Al Jurdi RK, Deshmukh PR, et al. Cognitive function and suicide risk in Han Chinese inpatients with schizophrenia. Psychiatry Res 2014; 220 (1-2):188-192.
20. Serafini G, Pompili M, Innamorati M, Rihmer Z, Sher L, Girardi P. Can cannabis increase the suicide risk in psychosis? A critical review. Curr Pharm Des. 2012;18(32):5165-87.
21. Ostergaard MLD, Nordentoft M, Hjorthøj C. Associations between substance use disorders and suicide or suicide attempts in people with mental illness: a Danish nation-wide, prospective, register-based study of patients diagnosed with schizophrenia, bipolar disorder, unipolar depression or personality disorder. Addiction 2017; 112(7):1250-1259.
22. Ran MS, Xiang MZ, Mao WJ, et al. Characteristics of suicide attempters and non-attempters with schizophrenia in a rural community. Suicide Life Threat Behav 2005; 35(6):694-701.
23. Reutfors J, Clapham E, Bahmanyar S, et al. Suicide risk and antipsychotic side effects in schizophrenia: nested case-control study. Hum Psychopharmacol 2016; 31(4):341-345.
24. Zhang XY, Al Jurdi RK, Zoghbi AW, et al. Prevalence, demographic and clinical correlates of suicide attempts in Chinese medicated chronic inpatients with schizophrenia. J Psychiatr Res 2013; 47(10):1370-1375.
25. Pompili M, Serafini G, Innamorati M, et al Suicide risk in first episode psychosis: a selective review of the current literature. Schizophr Res. 2011;129(1):1-11.
26.  Fleischhacker WW, Kane JM, Geier J, et al. Completed and attempted suicides among 18,154 subjects with schizophrenia included in a large simple trial. J Clin Psychiatry 2014; 75(3): e184–190
27. Pompili M, Serafini G, Innamorati M, et al. Unmet treatment needs in schizophrenia patients: is asenapine a potential therapeutic option? Expert Rev Neurother 2011; 11(7):989-1006.
28. Bora E. Differences in cognitive impairment between schizophrenia and bipolar disorder: Considering the role of heterogeneity. Psychiatry Clin Neurosci 2016 ;70(10): 424-433.
29. Schulz SC, Murray A. Assessing cognitive impairment in patients with schizophrenia.  J Clin Psychiatry 2016; 77 (Suppl 2): 3-7.
30. Kim CH, Jayathilake K, Meltzer HY. Hopelessness, neurocognitive function, and insight in schizophrenia: relationship to suicidal behavior. Schizophr Res 2003; 60: 71–80
31. Verma D, Srivastava MK, Singh SK, Bhatia T, Deshpande SN. Lifetime suicide intent, executive function and insight in schizophrenia and schizoaffective disorders. Schizophr Res 2016; 178(1-3):12-16.
32.  Potkin SG, Anand R, Alphs L, Fleming K. Neurocognitive performance does not correlate with suicidality in schizophrenic and schizoaffective patients at risk for suicide. Schizophr Res 2003; 59: 59–66
33. Barrett EA, Sundet K, Simonsen C, et al. Neurocognitive functioning and suicidality in schizophrenia spectrum disorders. Compr Psychiatry. 2011; 52(2):156-163.
34. Kosidou K, Dalman C, Fredlund P, et al. School performance and the risk of suicide attempts in young adults: a longitudinal population-based study. Psychol. Med. 2014; 44: 1235-1243
35. Seidman LJ, Shapiro DI, Stone WS, et al.  Association of Neurocognition With Transition to Psychosis: Baseline Functioning in the Second Phase of the North American Prodrome Longitudinal Study. JAMA Psychiatry 2016; 73(12):1239-1248.
36. Xiang YT, Weng YZ, Leung CM, Tang WK, Ungvari GS. Socio-demographic and clinical correlates of lifetime suicide attempts and their impact on quality of life in Chinese schizophrenia patients. J Psychiatr Res 2008; 42(6):495-502.
37. Yan F, Xiang YT, Hou YZ, et al. Suicide attempt and suicidal ideation and their associations with demographic and clinical correlates and quality of life in Chinese schizophrenia patients. Soc Psychiatry Psychiatr Epidemiol 2013; 48(3):447-454.
38. McGirr A, Tousignant M, Routhier D, et al. Risk factors for completed suicide in schizophrenia and other chronic psychotic disorders: a case-control study. Schizophr Res 2006; 84(1):132-143.
39. Kelleher I, Corcoran P, Keeley H, et al. Psychotic symptoms and population risk for suicide attempt: a prospective cohort study. JAMA Psychiatry 2013; 70(9):940–948
40. De Vylder JE, Hilimire MR. Suicide risk, stress sensitivity, and self-esteem among young adults reporting auditory hallucinations. Health Soc Work 2015; 40(3):175–181
41. Koyanagi A, Stickley A, Haro JM. Subclinical psychosis and suicidal behavior in England: findings from the 2007 adult psychiatric morbidity survey. Schizophr Res 2015; 168(1–2): 62-67
42. Leucht S, Cipriani A, Spineli L, et al.  Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet 2013; 382(9896):951-962.
43. Ernst CL, Goldberg JF.  Antisuicide properties of psychotropic drugs: a critical review. Harv Rev Psychiatry 2004; 12(1): 14–41.
44. Reutfors J, Bahmanyar S, Jonsson EG, et al. Medication and suicide risk in schizophrenia: a nested case-control study. Schizophr Res 2013; 150 (2–3): 416–420
45. US Centers for Disease Control and Prevention. Health behaviors of adults: United States, 2005-2007. Vital and health statistics 2010; 245 (10): 80
46. Krug EG, Dahlberg LL, Mercy JA, Zwi AB, Lozano R. World report on violence and health, vol. 1. Geneva: World Health Organization; 2002; 185.
47. Kao YC, Liu YP, Cheng TH, Chou MK. Cigarette smoking in outpatients with chronic schizophrenia in Taiwan: relationships to socio-demographic and clinical characteristics. Psychiatry Res 2011; 190(2-3):193-199.
48. Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull 1987; 13:261–276
49. Guy W. ECDEU Assessment Manual for Psychopharmacology, Revised. Bethesda, MD, US Department of Health, Education, and Welfare, 1976
50. Fagerstrom KO. Measuring degree of physical dependence to tobacco smoking with reference to individualization of treatment. Addictive Behaviors 1978; 3: 235-241.
51. Randolph C, Tierney MC, Mohr E, Chase TN. The Repeatable Battery for the Assessment of Neuropsychological Status (RBANS): preliminary clinical validity. J Clin Exp Neuropsychol 1998; 20(3):310-319.
52. Zhang BH, Tan YL, Zhang WF, et al. Repeatable battery for the assessment of neuropsychological status (RBANS) as a screening test in Chinese: reliability and validity. Chinese Mental Health Journal 2009; 28: 865-869
53. Chang WC, Chen ESM, Hui CLM, Chan SKW, Lee EHM, Chen EYH.. Prevalence and risk factors for suicidal behavior in young people presenting with first-episode psychosis in Hong Kong: a 3-year follow-up study. Soc Psychiatry Psychiatr Epidemiol 2015; 50: 219–226.
54. Ma X, Xiang YT, Cai ZJ, et al. Lifetime prevalence of suicidal ideation, suicide plans and attempts in rural and urban regions of Beijing, China. Aust N Z J Psychiatry 2009; 43(2):158-166.
55. Li D, Yang X, Ge Z, et al. Cigarette smoking and risk of completed suicide: a meta-analysis of prospective cohort studies. J Psychiatr Res 2012; 46(10):1257-1266.
56. Poorolajal J, Darvishi N. Smoking and Suicide: A Meta-Analysis. PLoS One 2016; 11(7):e0156348.
57. Gill KE, Quintero JM, Poe SL, et al. Assessing suicidal ideation in individuals at clinical high risk for psychosis. Schizophr Res 2015; 165(2–3):152–156
58. Pratt D, Gooding P, Johnson J, Taylor P, Tarrier N. Suicide schemas in non-affective psychosis: An empirical investigation. Behav Res Ther 2010; 48: 1211–1220.
59. Fedyszyn IE, Robinson J, Harris MG, Paxton SJ, Francey S. Predictors of suicide-related behaviors during treatment following a first episode of psychosis: The contribution of baseline, past, and recent factors. Schizophr Res 2012; 140: 17–24.
60. Mitter N, Subramaniam M, Abdin E, Poon LY, Verma S. Predictors of suicide in Asian patients with first episode psychosis. Schizophr Res 2013; 151: 274–278.
61. Delaney C, McGrane J, Cummings E, et al. Preserved cognitive function is associated with suicidal ideation and single suicide attempts in schizophrenia. Schizophr Res 2012; 140(1-3):232-236.

Table 1

Table 2

Table 3