Multiple treatments are effective for major depressive disorder (MDD), but the outcomes of each treatment vary broadly among individuals. Accurate prediction of outcomes is needed to help select a treatment that is likely to work for a given person. We aim to examine the performance of machine learning methods in delivering replicable predictions of treatment outcomes.

Of 7732 non-duplicate records identified through literature search, we retained 59 eligible reports and extracted data on sample, treatment, predictors, machine learning method, and treatment outcome prediction. A minimum sample size of 100 and an adequate validation method were used to identify adequate-quality studies. The effects of study features on prediction accuracy were tested with mixed-effects models. Fifty-four of the studies provided accuracy estimates or other estimates that allowed calculation of balanced accuracy of predicting outcomes of treatment.

Eight adequate-quality studies reported a mean accuracy of 0.63 [95% confidence interval (CI) 0.56–0.71], which was significantly lower than a mean accuracy of 0.75 (95% CI 0.72–0.78) in the other 46 studies. Among the adequate-quality studies, accuracies were higher when predicting treatment resistance (0.69) and lower when predicting remission (0.60) or response (0.56). The choice of machine learning method, feature selection, and the ratio of features to individuals were not associated with reported accuracy.

The negative relationship between study quality and prediction accuracy, combined with a lack of independent replication, invites caution when evaluating the potential of machine learning applications for personalizing the treatment of depression.

The modality of assessment used at a University Clinical Practice in Brazil is interventive psychodiagnosis in which the active participation of children and families is considered. Orientation is given following the input provided by children and their parents.

Evaluating the use of an electronic form to be fulfilled during the observation of a child’s play in psychological session.

A child at the age of 5yrs 4m was brought for psychological assessment with the complaint of aggressiveness and irritability. His parents answered the Child Behavior Checklist (CBCL -1 1/12 5 yrs) and the Psychology interns had to observe the child’s play and fulfill an electronic form in which the choice of toys and plays, motricity, creativity, symbolic abilities, frustration tolerance, adequation with reality were verified.

The results of CBCL indicated that the child was within the clinical range regarding anxiety and depression along with somatic complaints. The indicators observed in the electronic form such as rigidity in the modality of play, the lack of adequate ability of impersonating in role-playing, the difficulty of using creativity during play unless he was guided by peers or the Psychology interns and the constant anguish of separating himself from his parents were crucial for parents’ orientation. The psychological treatment lasted five months and benefited from the information obtained through the form once the symptoms of irritability and aggressiveness were reduced.

This modality of assessment can be instructional for parents and may also reduce financial and time costs once provides specific indicators to observe during play.

Postpartum depression (PPD) is the commonest postpartum psychiatric condition, with prevalence rates around 20%1. PPD is associated with a range of adverse outcomes for both the mother and infant2. Therefore, identifying modifiable risk factors for perinatal depression is an important public health issue3.

To explore the role of dysfunctional attitudes towards motherhood in postpartum depressive symptoms and disorder.

247 women were evaluated in the third (12.08±4.25 weeks) and sixth months (31.52± 7.16 weeks) postpartum with the Attitudes Towards Motherhood Scale4, the Postpartum Depression Screening Scale5 and the Diagnostic Interview for Psychological Distress-Postpartum6. Correlation analysis was performed followed by linear/logistic regression analysis when the coefficients proved significant (p<.05), using SPSS.

Dysfunctional beliefs towards motherhood concerning judgement by others and maternal responsibility positively correlated with depressive symptoms at the third (.528; .406) and the sixth months (.506; .492) postpartum. Those dysfunctional beliefs were predictors of depressive symptoms at the third (ß=.440; ß=.151) and sixth months (ß=.322; ß.241) explaining 29.4% and 30.2% of its variance, respectively. Having dysfunctional beliefs at the third month significantly increase the likelihood of being diagnosed with Major Depression (DSM5) both in the third (Wald=9.992, OR=1.169; Wald=16.729, OR=1.231) and sixth months (Wald=5.638, OR=1.203; Wald=7.638, OR=1.301) (all p<.01).

Cognitive distortions should be included in the assessment of risk factors for PPD. Early identification of women presenting motherhood-specific cognitive biases may be crucial for implementing preventive interventions favoring a more positive and healthier motherhood experience.

No significant relationships.

Motherhood-related beliefs are characterized by themes of failure and maternal role idealization. Recent studies found that postpartum depression/PPD is both predicted and a predictor by/for dysfunctional beliefs/DB. Additionally, it is possible that when contextual factors (eg. lack of social support) are present, women may anticipate the parenting experience as being of isolation, which in turn can lead to more dysfunctional attitudes.

To explore psychosocial risk factors for motherhood-DB.

233 women were evaluated in the second trimester (17.05±4.82 weeks) of pregnancy and in the third month (12.08±4.25 weeks) postpartum sociodemographically and psychosocially (years of education, previous children and social support) and the Portuguese validated self-report questionnaires to assess: perinatal depression; perinatal anxiety; perfectionism; negative affect; self-compassion; and repetitive negative thinking (all in T0). The Attitudes Towards Motherhood Scale was administered in the postpartum. When Pearson/Spearmen correlation coefficients proved significant (p<.05), linear/logistic (hierarchic) regression analysis were performed.

Motherhood-DB correlated significantly with all the variables, except for years of education, Other-oriented-Perfectionism and Common-Humanity. Motherhood-DB were significantly higher in women without previous children (p<.05). The final regression model was statistically significant (p<.001) explaining 15% of the Motherhood-DB variance, with Socially-Prescribed-Perfectionism and social support being the only statistically significant predictors. Hierarchic regression showed that even after controlling for social support, SSP significantly incremented the variance in 9%.

Our results highlight the need for preventive approaches to help women understand the origins of their dysfunctional beliefs (perfectionism, the myths of perfect motherhood) and for the promotion of positive cognitions.

No significant relationships.

The main purpose of the present study was to evaluate whether complications related to surgery for exostoses are associated with a decrease in patients’ quality of life.

This was a retrospective study for which the following information was collected: sex, age, pre- and post-operative symptoms, pre- and post-operative audiological evaluation results, surgical approach, instruments used, complications, and Glasgow Benefit Inventory score.

The study included 67 patients (94 ears). The three main complaints reported were wax retention, otitis externa and hearing loss. Surgical complications occurred in 14.9 per cent of patients. Patients experienced a significant benefit from surgery, especially in relation to somatic state, with a global Glasgow Benefit Inventory score of + 44.3. No significant difference was found between the global Glasgow Benefit Inventory changes and surgery-related complications (p = 0.093).

After surgery for exostoses, the vast majority of patients showed improvement. Complications related to surgery in general do not seem to influence patients’ satisfaction with surgery.

Insight and treatment adherence in serious mental illness, namely psychotic disorders, are well recognized as strong predictors of prognosis; several psychometric instruments have been developed for their evaluation.

To analyze the relation between self and hetero-evaluated insight and treatment attitudes in a clinical sample of psychiatric patients, besides assessing it's clinical correlates and relevance.

60 patients with serious mental illness (schizophrenia and bipolar disorder with psychotic features) were assessed using the Portuguese versions of the ‘Insight Scale' (IS), ‘Medication Adherence Rating Scale’ (MARS) and ‘Insight and Treatment Attitudes Questionnaire’ (ITAQ). General psychopathology and functioning scales were also applied, such as the BPRS, BDI-II and FAST. Relevant sociodemographic and clinical variables were also obtained. Statistical analysis was conducted using SPSS version 19.

Self-reported insight using the IS was not correlated with the insight subscale of the ITAQ, a hetero-evaluation instrument. Similarly, we found no correlation regarding attitudes to treatment when comparing self-report (MARS) and hetero- evaluation (treatment attitudes subscale of the ITAQ). Nonetheless, patients with a history of medication non-compliance and worse clinical outcomes had significantly lower (p< .05) scores in hetero-evaluation measures of insight and treatment attitudes (ITAQ), yet exhibited no differences in the self-evaluation measures of those dimensions.

Insight and treatment attitudes assessments can be valuable in clinical practice, contributing to decisions in both in- and outpatient settings involving treatment planning and level of monitoring. Clinician-rated instruments are probably more reliable, with clearer prognostic relevance.

To investigate in non-depressed perinatal women, whichBeck Depression Inventory-II (BDI-II) and Postpartum Postpartum DepressionScreening Scale (PDSS) scores and proportions of symptomatic items significantlydiffer through the perinatal period.

273 women were assessed with the Diagnostic Interviewfor Genetic Studies, BDI-II and the PDSS at the third trimester of pregnancy/T0(mean age =29.42±7.159 years; mean weeks gestation=32.6±3.61),three/T1, six/T2 and twelve/T3 months postpartum.

Significantlyhigher scores in T0 than in T1/T2/T3 and in T1 than in T2/T3 were found for: Total PDSS, Sleeping-Eating Disturbances, Emotional Liability, Total BDI-II andits dimension Somatic-Anxiety. Only T0 and T1/T2 didnot significantly differ in Anxiety/Insecurity, Mental Confusion, and Loss of Self. In Guilty/Shame only T1 was significantly higher than T3. Suicidalthoughts were significantly higher in T0 than T1 and lower in T1 than T2/T3.

Theproportions of symptomatic items systematically and significantly decreased fromT0 to T3 in changes of sleep, loss of interest in sex, feeling overwhelmed. Significantlyhigher symptomatic proportions in T0 than in all post-partum moments were foundfor: loss of energy, irritability, fatigue, changes in appetite, crying, indecisiveness and worthlessness (these last three also higher in T1 than inT3). Agitation was significantly higherin T1 than in T0. All the cognitive-affective symptoms from BDI-II and PDSSdid not differ.

From pregnancy to the twelfth month postpartum the depressivesymptomatology, particularly somatic-affective symptoms, significantlydecreased. However even for the women who are not clinically depressed, theperinatal period requires considerable psychological adjustment.

Factors associated with postpartum/PPT anxiety have been insufficiently investigated.

To identify correlates of PPT anxiety.

201 3-months postpartum women completed the Portuguese versions of Postpartum Depression Screening Scale (PDSS), Beck Depression Inventory/BDI-II, Profile of Mood States, Difficult Infant Temperament Questionnaire/DITQ, Multidimensional Perfectionism Scale and questions on sociodemographic variables, Lifetime history of insomnia/LTHD, Lifetime history of depressive symptomatology/LTHDS, Current insomnia, Health perception/HP, Stress perception/SP, Perceived social support/PSS, Quality of life/QOL, Health problem/complication postpartum, Sensibility to hormonal fluctuations, Type of delivery and Type of feeding. Postpartum anxiety was measured with the Anxiety/Insecurity (AI) subscale of the PDSS.

AI significantly correlated with LTHDS (rS=.32), LHI (rS=.18), Current insomnia (rS=.32), BDI-II (r=.76), SP (rS=.38), PSS (rS=.25), Perceived QOL (rS=.37), Health perception (rS=.29), Health problem in the postpartum (rS=.26), Negative Affect/NA (r=.66), Positive Affect/PA (r=.58), Conditional Acceptance/CA (r=.29) and DIT (r=.38) (all p< .01). Mean comparisons revealed that women with vs. without LHDS, with vs. without LTHI, good sleepers vs. with insomnia syndrome, high vs. low SP, low vs. high PSS, bad/very bad vs. good/very good QOL, bad/very bad vs. good/very good HP, high (< M+1DP) vs. low (>M-1DP) DIT, CA, NA and low vs. high PA had significantly higher mean scores in AI (all p< .01). Linear regression model composed of all correlated variables explained 53.7% of AI variance and showed that NA, PA and DIT are significant (p< .05) predictors of AI.

Our findings are in accordance with previous research and contribute to the progress on this topic.

Postpartum depressive symptomatology is highly prevalent and has negative impact in the entire family.

To identify correlates of postpartum depressive symptomatology.

201 3-months postpartum women completed the Portuguese versions of Postpartum Depression Screening Scale/PDSS, Profile of Mood States, Difficult Infant Temperament Questionnaire/DITQ, Multidimensional Perfectionism Scale and questions on sociodemographic variables (marital status, work status, educational level, parity), Lifetime history of insomnia/LTHD, Lifetime history of depressive symptomatology /LTHDS, Current insomnia, Sleep needs, Health perception/HP, Stress perception/SP, Perceived social support/PSS, Quality of life/QOL, Health problem/complication postpartum, Sensibility to hormonal fluctuations, Type of delivery and Type of feeding.

Total PDSS significantly correlated with LTHDS (rs=.35), LHI (rs=.22), Current insomnia (rs=.37), SP (rs=39), PSS (rs=.25), Perceived QOL (rs=.31), Health perception (rs=.28), Health problem or complication in the postpartum (rs=.16), Negative Affect/NA (r=.59), Positive Affect/PA (r=-.67), Conditional Acceptance/CA(r=.29) and DIT (r=.40) (all p< .01). Mean comparisons revealed that women with vs. without LHDS, with vs. without LTHI, good sleepers vs. with insomnia syndrome, high vs. low SP, bad/very bad vs. good/very good QOL, bad/very bad vs. good very good HP, high (< M+1DP) vs. low (>M-1DP) DIT, CA, NA and low vs. high PA had significantly higher mean scores in AI (all p< .01). Linear regression model composed of all correlated variables explained 53.9% of depressive symptomatology variance and showed that NA, PA, DIT and HP are significant (p< .05) predictors of AI.

These results are in line with previous findings and contribute to the progress on this topic.

The Children Eating Attitudes Test (ChEAT; Maloney et al. 1988) is a well-established 26-item scale designed to measure a wide range of problematic eating attitudes and behaviours among children and adolescents.

To analyse ChEAT reliability and validity in a Portuguese adolescent girls sample.

565 high-school girls (mean age 15.76 ± 1.571; mean BMI 20.42 ± 2.745) answered the Portuguese versions of ChEAT and of the Contour Drawing Figure Rating Scale (CDFRS; Thompson & Gray, 1995). to study the temporal stability 124 girls answered the ChEAT again after approximately six weeks.

Cronbach's alpha was of .76. the test-retest Pearson correlation was of 0.61. A four factors structure (explained variance=44.06%) was selected: Factor (F) 1 Fear of Getting Fat, F2 Restrictive and Purging Behaviours, F3 Food Preoccupation, F4 Social Pressure to Eat. the body satisfaction as assessed through CDFRS was negatively correlated with the total ChEAT (-.35), F1 (-.47) and F2 (-.23) (all p > .001); and positively correlated with F4 (.26, p < .001). Significant mean differences (all p < .01) were found between the three CDFRS groups (Group -1 Want to be thinner; Group 0 Satisfied; Group 1 Want to be fatter) in all eating behaviour dimension scores, except for F3; total ChEAT, F1 and F2 mean scores between groups significantly decreased through the body satisfaction groups -1, 0 and 1 and significantly increased for F4.

The Portuguese ChEAT psychometric characteristics are good. Factorial structure is in accordance with the original. It could be very useful to clinical and epidemiological purposes.

Suicidality has a multifactorial determination and is clinically under-diagnosed. Self-reported poor health, negative affect and sleep difficulties are associated with psychological distress, including suicidality.

To analyze the association between lifetime SRH, sleep difficulties, negative affect (NA) and suicidal ideation in pregnancy (T0) and post-partum (T1) and if they are predictors of suicidal ideation.

397 pregnant women completed the Portuguese version of POMS, PDSS and a set of items evaluating SRH, and sleep difficulties in the last trimester of pregnancy an three months of post-partum.

In pregnancy, suicidal ideation was significantly associated with lifetime SRH (p < .05), NA (p < .01) and difficulties of initiating sleep (DIS) (p < .01). In post-partum, suicidal ideation was also associated with DIS (p < .01), NA (p < .01), but not with SRH. In pregnancy, the predictors of the probability of having or not suicidal ideation are both SRH and DIS. In postpartum the suicidal ideation predictor was only DIS. However when NA was introduced in the regression model, NA was the only predictor of suicide ideation in both pregnacy and postpartum.

Difficulties of initiating sleep should be considered an indicator of suicidal ideation during the perinatal period. However NA might explain this association.

Postpartum depression (PPD) can occur through all the perinatal period and it is a public health problem. Positive and negative affect (at pregnancy and previous postpartum moments) are protective and risk factors for PPD. The Profile of Mood States (POMS) factor structure at pregnancy and three months postpartum has already been explored.

to explore the POMS factor structure at six and 12 months postpartum.

336 women (mean age=30,3; SD = 4,09; range=19-42 years) and 276 women (mean age=30,5; SD=3,99; range=19-41 years), respectively, filled in the Profile of Mood States (POMS), at six and 12 months postpartum.

A principal components analysis revealed, at six months postpartum, three components that explained 54,12% of the total variance: F1 Depression-Hostility (21 items; 39,5% explained variance/EV); F2 Anxiety-Anger/fatigue (14 items; 10,1% EV) and F3 Vigor-Activity (12 items; 4,8% EV); at 12 months postpartum, revealed also three components, that explained 46,16% of the variance: F1. Anxiety-anger/fatigue (13 items; explained 30,4% of the variance); F2. Depression-Hostility (6 items; 11,2% VE); F3. Vigor-Activity (14 items; 2,9% EV). At both postpartum moments, F1 and F2 can be summed to calculate Negative Affect (Cronbach Alpha, α=0,968 at six months and α=0,948 at 12 months). F3 corresponds to Positive Affect (α=0,863; α=0,875, respectively, at six months and 12 months).

The POMS factor structure at six and 12 months is robust, meaningful and can now be used to explore different aspects of the postpartum experience, at these postpartum moments.

It has been shown that perfectionism constitutes a risk factor both for insomnia and postpartum depression.

To analyse the relationship between perfectionism, depressive symptoms and insomnia in the postpartum.

201 3-months postpartum women (M=12.27 ± .91 weeks postpartum) completed the Portuguese versions of Postpartum Depression Screening Scale (PDSS) and Multidimensional Perfectionism Scale (MPS). Three MPS dimensions derived by factorial analysis with data from this sample were used: Self-Oriented Perfectionism/SOP, Social Prescribed Perfectionism-Conditional Acceptance/SPP-CA and SPP-Others Highs Standards/OHS. Women also answered 5 questions about sleep, considering the previous month, based on which three insomnia groups were formed: Good Sleepers/GS (women without insomnia symptoms or daytime impairment); Insomnia Symptoms Group (women with at least one insomnia symptom but no daytime impairment); Insomnia Syndrome Group/ISG (women with at least one insomnia symptom and sleep related daytime impairment).

Insomnia and PDSS were correlated (r = .39, pr < .01). SOP was not correlated with insomnia neither with PDSS total score. SPP-CA was significantly correlated with insomnia and with PDSS (rr > .25, pr < .01). SPP-OHS was only significantly correlated with insomnia (r = .14, pr < .05). SPP-CA mean scores were significantly higher in ISG than in GS (23.45 ± 3.032 vs. 18.98 ± 8.81, p = 002). Considering insomnia as a quantitative variable, linear regression showed that SPP-CA (β = .20, p = .005) and PDSS (β = .17, p = .021) were both significant predictors of insomnia, explaining 8.7% of variance (pr < .001). The mediation analysis revealed that SPP-CA partially mediated the relationship between PDSS and insomnia (IC 95% .008-.005).

SPP-C is a relevant correlate of postpartum insomnia.

Knowledge about the mechanisms underlying the relationship between perfectionism and eating behaviors in overweight/obese women is very scarce.

To investigate the relationship between perfectionism, eating behaviors and affect in overweight women.

The Portuguese validated versions of the Eating Disorders Examination Questionnaire/EDEQ, the Multidimensional Perfectionism Questionnaire and the Profile of Mood States were administered to an outpatient sample of 276 women (Mean age = 43.85 ± 11.89; Mean BMI = 32.82 ± 5.43) attending a weight loss treatment in a public hospital.

Correlations between Social Prescribed Perfectionism/SPP and EDEQ total (T) and its dimensional scores (Weigh and Shape Concern and Dissatisfaction/WSCD, Eating Concern/EC, Restraint) were all moderate (r > .30) and significant (p > .001). Self-Oriented Perfectionism/SOP was also significantly correlated with EDEQ-T, WSCI and Restraint (r = .20). Positive affect/PA was negatively correlated and Negative affect/NA was positively correlated with all EDEQ dimensions (r > .25; p > .001). SPP, but not SOP, was significantly correlated with PA (r = −.27) and NA (r = .34). Participants with high (>M+SD) vs. low (< MSD) SOP and SPP had significantly higher means in EDEQ-T, WSCI, EC and Restraint (all p < .001). Linear regression showed that SPP, PA and NA were predictors of EDEQ-T and WSCI; SPP and NA were predictors of EC; NA was the unique predictor of Restraint (all p < .001). Controlling for NA/PA, SPP still being a significant predictor of all EDEQ, WSCD and EC, accounting for significant increments of variance (4.4%, 2.9% and 4.3%, respectively; p < .001).

As in studies with other type of samples, SPP is related to disordered eating in overweight women.

In comparison with postnatal depressive symptoms, few studies considered antenatal depressive symptoms risk factors/correlates.

In a multidimensional work our aim was to explore the associations/predictive role of lifetime and current pregnancy variables for antenatal depressive symptoms.

Two hundred and thirty six women (M = 30.5; SD =4.02; variation = 19-41), at the third pregnancy trimester were interviewed with the DIGS/OPCRIT (to make lifetime/current depression diagnoses) and answered a booklet about sociodemographic variables, lifetime/current insomnia, current stress perception, perception of being an anxious person, gynecological variables, past health perception and pregnancy health problems, among others. They filled in the Profile of Mood States to assess anxious symptoms, Negative and Positive affect and the Postpartum Depression Screening Scale (PDSS) to assess antenatal depressive symptoms at pregnancy. Women with a depression diagnosis (DSM-IV/ICD-10, n =5) and on sleep medication at pregnancy were excluded from the analyses.

We found significant associations between antenatal depressive symptoms (PDSS total score), current job situation, perception of being an anxious person, lifetime history of depression (DSM-IV/ICD-10), past health perception, health problems at current pregnancy, stress perception at pregnancy, lifetime and current insomnia, anxious symptoms at pregnancy, Negative and Positive affect. In an hierarquical multiple regression past health perception, negative affect, positive affect and current insomnia were significant predictors of antenatal depressive symptoms.

Past health perception, negative affect (anxious and depressive symptoms) and insomnia at current pregnancy seem to be risk factors/correlates for the outcome. Positive affect seems to work as a protective correlate.

To investigate the symptom dimensions of psychosis using factor analyses/FA of lifetime symptoms of a sample of Portuguese psychotic patients.

FA of the OPCRIT items (56 signs and symptoms) of 684 patients

- Schizophrenia/Sz (73.8%), Other non-organic psychotic disorders/ONPD (6.0), Schizoaffective disorders/SzA (1.4%), Bipolar disorders/BP (18.1%) and Severe depression with psychosis/SDP (.8%) – assessed with the Diagnostic Interview for Genetic Studies.

Delusions of poverty, guilty delusions, nihilistic delusions, primary delusional perception, and somnolence, problems with appetite and weight and grandiosity were excluded from the final solution, as each were present in less than 10% of patients and did not load at any factor.

Following the Kaiser and the Cattel's Scree Plot criteria, a four factors structure was selected, which explained variance (EV) was of 60.25%. Based on items content, the meaningful four factors were denominated as follows: F1 Depression (EV 21.77%; a=.97); F2 Mania (5.72%; a=.97); F3 Delusions and hallucinations (2.97%; a=.92); F4 Disorganization and Negative symptoms (2.07%; a=.90). Mann-Whitney U tests revealed that the symptom dimensions that distinguished better between Dx categories were Delusions and hallucinations (only SDP vs. BP and Sc vs. SzA did not significantly differ) and Disorganization and Negative symptoms (only SDP vs. BP and SDP vs. ONPD did not significantly differ), with Sz patients presenting the highest scores.

This factorial structure is in accordance with other reports. Given the Schizophrenia (Sz) and Bipolar disorder (BP) heterogeneity and overlap, the present study contribute to phenotypic refinement and formulation of alternative psychosis phenotypes.

In Portugal there are no national epidemiologic registers on the prevalence/incidence of suicidal ideation and attempts. Although there is no evidence that suicide is higher in University students than in other people of the same age, it is a cause of premature death.

To analyze the prevalence and characteristics of suicidality in university students

549 University students (80.1% females) filled in a booklet of questionnaires that included the Suicidal Behaviour Section of the Diagnostic Interview of Genetic Studies (Azevedo et al, 1993; Nurnberger et al., 1994), which was slight modified.

0.4% reported lifetime poor psychological health and 1.3% reported lifetime hospitalizations due to emotional/behavioural problems. Lifetime thoughts of death, lifetime active suicidal ideation, and lifetime suicidal attempts were described by 20.4%, 11.9% and by 2.3% of the students, respectively. Regarding the more severe lifetime suicide attempt (N=15) the method used was predominantly self-poisoning (medication) (9/13), 40% (6/15) were hospitalized after this attempt, 53.3% (8/15) reported death wishes when the act was performed, and 60% (9/15) considered that death could be the consequence of the act. During last month suicidal ideation occurred in 1.9% of the students. During the last two weeks suicidal ideation without planning was found in 2.2% and 0.2% refereed they are going to attempt suicide if they will have an opportunity.

In university students recent, frequent and severe suicide ideation might occur. Results highlighted the need for suicidality prevention/early intervention and that Health University Care Services should screen systematically suicidal ideation.

The Eating Disorder Examination Questionnaire (EDEQ; Fairburn & Beglin, 1994) is a widely used measure composed of 28 items derived from the EDE Interview. The questionnaire has been considered an adequate measure of eating related psychopathology, including with overweight samples. EDEQ also assesses frequencies of eating disordered behaviors in terms of the number of episodes in the past four weeks (6 items not contributing to scores).

To investigate reliability and factorial structure of the EDEQ in a large sample of Portuguese overweight women.

The EDEQ was administered to an outpatient sample of 276 women (Mean age = 43.85 ± 11.89; Mean BMI = 32.82 Kg/m2 ± 5.43) attending a weight loss treatment consultation in a public medical center.

The EDEQ Cronbach a was .88. All items contributed to the internal consistency. A three factors structure (variance explained = 50.37%) was selected: Factor (F) 1 “Weight and shape concern/dissatisfaction” (a = .91); F2 “Eating concern” (a = .71) and F3 “Restraint” (a = .63). The mean total score was 2.36 (± 1.07). The regular occurrence of overeating episodes, selfinduced vomiting, laxative misuse and excessive exercise was reported by 10%, 1.5%, 4.1%, and 8.9% of participants, respectively.

The EDEQ Portuguese version showed adequate psychometric properties in a sample of overweight women. Its factorial structure, mean scores and specific eating disorder behaviors frequencies are in accordance with other studies with obese participants. The EDEQ Portuguese version could be very useful to clinical and epidemiological purposes in a broad range of eating disturbances.

1. (1) to identify the proportion of families that have at least one first degree relative (FDR) with major unipolar depression (MUD), across bipolar (BP) and schizophrenic (SC) multiplex families;

2. (2) to assess if there are significant demographic/clinic differences between these two groups.

The sample included 120 families with at least one first or second degree proband relatives diagnosed with BP or SC disorders. We selected BP and SC probands’ families that had at least one FDR with MUD, according to DSM-III-R. All families have been evaluated with Diagnostic Interview for Genetic Studies and selected according to the diagnostic definitions of DSM-III-R, based on the Operational Criteria Checklist. T-tests and chi-square tests used to analyze demographic and clinical differences between the two groups.

Out of the 60 BP families, 38.3% had at least one FDR with MUD; depressed relatives were predominantly female (68.7%) - sisters (35.5%), mothers (29%), daughters (3.2%). The average age of onset and length of illness was 35 and 15 years, respectively. Out of 60 SC families, 31.6% had at least one FDR with MUD; the majority of depressed relatives were females (81.8%) - sisters (42.9%), mothers (38%) daughters (4.8%). The average age of onset and length of illness was 30 and 19 years, respectively. There was no significant statistical differences between BP and SC families.

These results may support the thesis of a common genetic vulnerability to MUD, and the continuum viewpoint of affective and psychotic disorders.

Perfectionism has been associated to Perseverative Negative Thinking [PNT]. Both are transdiagnostic processes. PNT (in the form of worry and rumination) is a prime candidate when investigating negative components of perfectionism (Macedo et al., 2013).

To investigate if PNT mediates the relationship between perfectionism and negative affect [NA].

344 university students (68.4% girls) were evaluated using Portuguese validated versions of the Perseverative Thinking Questionnaire (two subscales:’Repetitive Thought’ [RT] and’Cognitive Interference and Unproductiveness’ [CIU]), Multidimensional Perfectionism Scale, Profile of Mood States and Perceived Stress Scale, with an additional item to evaluate perceived social support [PSS]. Only variables significantly correlated with the outcomes [NA] were entered in the hierarchic multiple regression models. Mediation analyses using Preacher and Hayes bootstrapping methodology were performed.

Perceived Stress [PS] ((=.245), PSS ((=-.257), Self-Oriented Perfectionism [SOP] ((=-.126), Self-Prescribed Perfectionism [SPP] ((=.122; p=.011), Concern over Mistakes [CM] ((=.102), Doubts about Actions [DA] ((=.115) and CIU ((=.110) were significant predictors (all p<.05) of NA. Controlling for PS and PSS, the perfectionism variables still predict depression, accounting for a significant increment of 16.7% (p<.01). The CIU, introduced in the model after the Perfectionism variables, significantly increments the NA variance in 1.6% (p<.01). CIU was a total mediator in the relationship between SOP and NA (IC95% .025-.229) and a partial mediator in the relationship between SPP (.176-.456), CM (.142-.855), DA (.143-.863) and NA.

PNT potentiates the relationship between negative perfectionism and NA. Positive perfectionism is associated with NA, only in the presence of PNT.