Sex difference in the progression of manic symptoms during acute hospitalization: A prospective pilot study

Abstract Objectives Acute mania is a serious medical condition that impacts men and women equally. Longtime presentation of manic symptoms is sex‐dependent; however, little is known about acute symptoms of mania. The objective of this study is to track and compare acute manic symptoms for sex differences during inpatient hospitalization. Methods All patients with bipolar mania admitted to a large university hospital between January and October 2017 were invited to participate in this longitudinal naturalistic follow‐up study. Manic (YMRS), depressive (MADRS), and psychotic (PAS) symptoms were tracked daily from admission to discharge. Results The total YMRS scores decreased significantly overtime (p < .0001) in both male (n = 34) and female (n = 23) patients (p = .7). However, male patients scored significantly higher in sexual interest (p = .01), disruptive and aggressive behavior (p = .01), and appearance (p < .001) while females had better insight into their illness (p = .01). Males and females received similar doses of lithium (p = .1), but males received significantly higher doses of valproic acid (VPA) in comparison with females (p = .003). However, plasma lithium and VPA concentrations at discharge were not significantly different between sexes. Conclusion Our results show sex differences in the progression of certain domains of manic symptoms in a cohort of 23 female and 34 male patients admitted to a large academic center in Turkey. Males, in this sample, exhibited more sexual interest, disruptive and aggressive behaviors, better grooming, and less insight compared to females. While these results are concordant with our preclinical findings and with anecdotal clinical observations, replication in larger samples is needed.

Additionally, BPI women are more likely to have lifetime history of attempted (Nivoli et al., 2011) and completed suicide (Clements et al., 2013). In contrast to these lifetime differences between male and female bipolar patients, sex differences in acute manic manifestations have received little attention. In a cross-sectional study, Young, Biggs, Ziegler, and Meyer (1978) assessed acute mania severity upon psychiatric admission using Young Mania Rating Scale (YMRS) in 149 consecutive patients and found that males scored significantly higher on the sexual interest item than female patients (Young et al., 2007). This study suggests subtle sex differences in the symptomatology of acute mania. However, only YMRS scores at admission were compared between subjects without further exploration of sex differences amid the course of manic symptoms during hospitalization or whether males and females received similar treatment.
Sex differences, in general, stem from a complex interplay between neurobiological and psychosocial factors. To explore the neurobiological aspect, we used our newly established lateral hypothalamic-kindled (LHK) rat mania model to examine sex differences in manic-like behaviors during mania induction (Abulseoud et al., 2014) and we found that male rats exhibited more sexual behavior, significantly increased locomotor activity during the light phase, and reduced rest interval, while females displayed significantly higher ethanol consumption and more frequent rearing behaviors (Abulseoud et al., 2015). To test whether these results translate in a clinical model, we sought to examine sex difference in the progression of manic symptoms during acute hospitalization in an academic medical center in Turkey. We chose such a facility to collaborate with because of their expertise in diagnosis, symptom quantification and treatment of acute mania, and the large volume of patients, and equally important, because of the unrestricted hospital length of stay. We hypothesized, based upon past research on acute mania (Hendrick et al., 2000;Kumar & Ram, 2001;Viguera et al., 2000;Young et al., 2007) and our work (Abulseoud et al., 2015), that the presentation of acute manic symptoms is sex-dependent.
Specifically, males will exhibit more aggressive and sexual manifestations compared to females. Our results document sex differences in several individual manic manifestations. University hospital is a large district hospital that serves the southern Marmara region with a population of about 4 million. The adult inpatient psychiatric unit has 110 beds (37 for females). All patients admitted to the unit between January and October 2017 and screened for history of bipolar diagnosis and current manic episode symptoms were invited to participate in this prospective naturalistic study. We excluded patients with schizoaffective disorder, intellectual disability, or illiteracy. Pregnant or lactating women were also excluded from the study. Study procedures were explained in detail for potential participants. Those who agreed to participate signed an informed consent and underwent detailed clinical interview by two Turkish board certified psychiatrists (GS and OS) assisted with two senior psychiatry residents. All cases were discussed prior to enrollment to confirm bipolar manic diagnosis. Physical examination, baseline laboratories including serum electrolytes, complete blood count, lipid panel, thyroid, liver, and kidney function tests were performed in all patients as part of routine work-up. Plasma valproate (VPA) levels were measured by color spectrophotometry (Belal, El-Kafrawy, Mahrous, Abdel-Khalek, & Abo-Gharam, 2016), and lithium levels are measured by flame photometric method (Levy & Katz, 1970). Blood samples were collected 12 hr after the last dose.

| ME THODS
Manic, depressive, and psychotic symptoms were tracked daily throughout the hospital length of stay (LOS) by the same team. To ensure consistency between different raters, the study team members had daily discussions about the ratings for individual study participants. We used Young Mania Rating Scale (YMRS; Young et al., 1978), Montgomery-Aberg Depression Rating Scale (MADRS; Montgomery & Asberg, 1979), and the Scale for the Assessment of Positive Symptoms (SAPS; Andreasen, 1984). YMRS is an 11-item clinician rating scale, each item is given a severity rating. There are four items that are graded on a 0 to 8 scale (irritability, speech, thought content, and disruptive/aggressive behavior), while the remaining seven items are graded on a 0 to 4 scale. MADRS is a 10-item scale: 1. Apparent sadness 2. Reported sadness 3. Inner tension 4. Reduced sleep 5. Reduced appetite 6. Concentration difficulties 7. Lassitude 8. Inability to feel 9. Pessimistic thoughts 10. Suicidal thoughts. Each item is scored 0-6. SAPS consists of 34 items divided into four positive symptom subscales: hallucinations, delusions, bizarre behavior, and positive formal thought disorder. Each subscale includes a global rating scale. In addition, detailed records for daily treatment and vital signs were also collected. Medication doses were recorded on daily basis throughout the hospital LOS.

| Statistical analysis
All data were presented as means (±SEM). Data were examined for normal distribution. Demographic and baseline laboratory data were examined for sex difference using student t test or Fisher exact test/ chi-square statistics as appropriate. Medications were grouped into lithium, valproic acid (VPA), antipsychotic, benzodiazepine, and other categories upon admission and at time of discharge. Antipsychotic medications were converted into chlorpromazine equivalent dose according to the classical mean dose (Leucht et al., 2015) and the minimum effective dose methods (Leucht et al., 2014). Benzodiazepines were converted into lorazepam dose equivalent according to published tables (Brett & Murnion, 2015). Separate two-factor ANOVAs were used to examine sex and hospitalization effects on different medication classes.
Rating scale scores were adjusted to hospital LOS into five time points: admission, 25% LOS, mid-LOS, 75% LOS, and discharge scores. Mean total scores (YMRS, MADRS, and SAPS) and individual items (YMRS and MADRS) were compared between male and female patients using separate repeated measures ANOVAs.
Analysis was done using GraphPad Prism V7 software. Results are considered significant when p < .05.

| Demographic information
A total of 72 patients (females: n = 30) with bipolar manic episode were admitted to the inpatient unit during the study interval from January to October 2017. A total of 58 patients (females: n = 24) consented for the study. Patients reported mild (MADRS < 19) to moderate (MADRS < 30) depressive symptoms. However, none met DSM-5 criteria for mixed episode as defined by having at least three out of six concurrent depressive symptoms during the majority of manic episode days (American Psychiatric Association, 2013). The patients who refused to participate in the study (total n = 14 [males: n = 8]) had difficulty understanding the study process and provide informed consent due to severe agitation or delirious mania. One female patient was discharged to another facility after 3 days and was excluded from analysis. The final cohort consisted of 34 males and 23 females. As shown in Table 1, the mean (±SEM) age was 37.6 (2.3) and 40.9 (2.4) years for males and females, respectively. Most of the females were married (65%) and homemakers (57%), while about half of male patients were married (50%), employed, or students (53%). More males compared to females reported legal problems (M vs. F: 40.6% vs. 17.4%, p = ns) and a past history of alcohol use (M vs. F: 51% vs. 13%, p = .004). The majority of males (68%) and females (83%) were classified as medium socioeconomic status based on reported monthly income (Low socioeconomic level = up to 999 Turkish Lera [TL], medium = 1,000-1,999 TL, and high = 2,000 TL or over).

| Bipolar disorder history
As detailed in Table 2, the hospital LOS did not vary significantly between males (20.7 ± 2.1) and females (22.6 ± 2.3) and about 2/3 of patients presented with psychotic features. Similarly, the age of first hospitalization did not differ between males and females. However, females had a nonsignificant trend (p = .08) toward more previous hospitalizations, but no differences in the number of prior manic or depressive episodes or the presence of psychotic features during past manic episodes or in the history of aggressive behaviors.
Females, however, reported significantly more previous suicidal attempts compared to males (p = .002).

| Baseline medical history, vital signs, and laboratory results
Other than bipolar illness, the cohort was generally healthy, and only few patients had comorbid medical conditions: diabetes (n = 1), hypertension (n = 1), and morbid obesity (n = 1). Males had slight but significant increase in heart rate and diastolic blood pressure and elevated, but within normal range, liver functions (ALT, AST, and direct bilirubin), alongside reduced vitamin B 12 concentration. On the other hand, females had mild anemia (low Hgb, HCT, and RBC counts) and lower free T3 concentration. No significant sex differences were observed in other vital signs or laboratory measures (Table 3). Table 4 describes the details of medications used. All patients were either on antipsychotic medication upon admission or were started on one during hospitalization. Quetiapine and haloperidol were the most commonly prescribed antipsychotics. We used classical mean dose (Leucht et al., 2015) and the minimum effective dose methods (Leucht et al., 2014) to convert individual antipsychotic daily doses to chlorpromazine equivalent dose. The mean dose equivalent did not differ between males and females upon admission or at discharge. Besides antipsychotics, lithium, VPA, and benzodiazepines, other medications were used such as anticholinergics (n = 27), antidepressants: Sertraline (n = 1), Lamotrigine and Venlafaxine (n = 1), Propranolol (n = 2), and vitamin B complex (n = 1).

| Course of manic manifestations (YMRS)
Daily YMRS ratings were collected and verified by two experienced psychiatrists (GS and OS) assisted with two senior residents. All rating results were discussed by the team to ensure reasonable agreement. The admission total YMRS scores ranged between 10 and 51 (mean ± SEM: 31.2 ± 2.2) for females and be-

| Course of psychotic features (SAPS)
A wide range of psychotic symptom severity was observed upon ad-

| D ISCUSS I ON
To the best of our knowledge, this is the first study that prospectively followed acute manic patients throughout hospitalization in a nonhealth maintenance organization (HMO). Our results show significant sex differences in four specific manic domains. Males scored higher in sexual interest, disruptive, and aggressive behaviors, while females had poor appearance but better insight into their illness.
However, whether there are sex differences in impulsive behaviors among manic patients remain largely unknown. In agreement with our findings, one cross-sectional study reported that male manic patients scored significantly higher on the sexual interest item than female patients upon admission to inpatient unit (Young et al., 2007).
In our preclinical LHK rat model of mania, male rats showed a nonsignificant trend (p = .0516) toward increased sexual behavior compared with female rats (Abulseoud et al., 2015). Further studies are needed to investigate the phenomenology and neurobiological basis of sex differences in impulsive behaviors during manic episode.
Interestingly, manic females in our group were less appropriately dressed and groomed compared to males but they also had more insight into their illness. It is conceivable that self-awareness drives both the motivation for better appearance and insight into mental illness. However, poor motivation and self-care are also common during depression. Females in this study reported a nonsignificant trend (p = .09) toward more depressive symptoms during hospitalization which could explain, at least in part, the discrepancy between insight and overall appearance. In contrast to our results, a recent study (Guclu et al., 2011) Reynolds, 2006). Methionine is transformed into S-Adenosyl Methionine (SAM) which is converted by methyltransferase into S-Adenosyl Homocysteine (SAH) by the enzyme methyltransferase an important step in methylation reactions which is essential for genomic and nongenomic methylation. Inside the mitochondria, vitamin B 12 functions as a cofactor for MMCo-A mutase which converts methylmalonic acid to succinyl co-enzyme A, which subsequently enters in the Krebs cycle for ATP production (Froese & Gravel, 2010;Gueant et al., 2013) the YMRS. These two scales could be probing different aspects of disease awareness with SUMD more focused on insight into psychotic symptoms while YMRS question 11 is a general questionnaire about admitting or denying mental illness or behavioral changes. A meta-analysis of eight studies looking at the change in insight ratings during hospitalization for acute mania found impaired insight upon admission and significant improvement at time of discharge (Ghaemi & Rosenquist, 2004). However, this meta-analysis reported no data on sex difference. Insight and self-awareness in general are higher cortical functions that are compromised in several psychiatric disorders (Sass & Parnas, 2003) and have a profound impact on psychosocial outcome (Ghaemi, Boiman, & Goodwin, 2000) but are challenging to study in preclinical models. Functional imaging studies on insight in patients with schizophrenia show different activation pattern in response to self-and/or other-referential probes in medial orbitofrontal cortex, prefrontal cortex, and posterior cingulate (Blackwood et al., 2004;Holt et al., 2011;Shad et al., 2012).
Only a few studies have looked at differences in activation pattern between manic patients and healthy controls (reviewed in Chen, Suckling, Lennox, Ooi, & Bullmore, 2011) without specifically addressing the neural network behind impaired insight during manic episode for sex difference.
Despite the differences in specific domains of mania between males and females, depressive and psychotic symptoms did not show sex difference and only a nonsignificant trend toward higher total MADRS scores was observed in females without any subject meeting DSM-5 criteria of mixed episode (American Psychiatric Association, 2013). Future studies designed specifically to probe sex differences in the depressive pole of the illness are needed.
In addition, we found significant effect of sex on VPA dose where males received significantly higher doses compared to females.
However, VPA plasma concentrations were not significantly different between the two groups, suggesting difference in body weight.
Similarly, benzodiazepine dose equivalent showed slight overall effect for sex but did not survive post hoc multiple comparison test. In contrast, lithium dose, plasma concentration, and antipsychotic dose equivalent showed no sex differences. Taken together, it seems that both males and females received similar treatment and had comparable hospital LOS.
Remarkably, none of the patients in this study met DSM-5 criteria for mixed state and despite that about half of the females in our cohort reported history of suicide attempts compared to only 10% of males, we observed only a nonsignificant trend (p = .09) toward higher MADRS scores in females throughout hospitalization. In general, mixed mania presentation is not common. The prevalence estimations for mixed episodes according to DSM-IV were about 13% (Vieta & Morralla, 2010).
In addition, bipolar women were not found to be different from men in the number of depressive episodes or the proportion of time spent in depression (reviewed in Diflorio & Jones, 2010). Furthermore, several studies found limited influence on male sex on suicide risk (reviewed in Diflorio & Jones, 2010) and one study reported that women were more likely to report history of suicidal attempts compared to men F I G U R E 3 Brain thyroid hormone. (a) Hypothalamic-pituitary-thyroid (HPT) axis; thyrotropin-releasing hormone (TRH) is released from the hypothalamus to stimulate anterior pituitary to secrete thyroid stimulating hormone (TSH) which in turn stimulates thyroid gland to produce mostly tetraiodothyronine (T4) and to less extent triiodothyronine (T3). (b) Transportation and brain action of plasma T3 and T4: Both T3 and T4 are present in free and protein-bound forms. Free T4 is transported through the blood-brain barrier (BBB) by the organic anion-transporting peptide 1c1 (OATP1c1) to astrocytes and tanycytes where it is converted into T3 by deiodinase type 2 (DIO-2) enzyme. Astrocyte T3 is transported to neurons through monocarboxylate transporter (MCT)8. Free T3 and to less extent T4 can also be transported directly into neurons through gap junctions and MCT8. Another unknown thyroid hormone transporter (UTHT) transporters astrocytic T3 to oligodendrocytes to activate myelination genes. Within neurons, T3 (and to possibly T4) bind to nuclear thyroid hormone receptors (nTR) to influence gene expression critical for cell growth and differentiation and synaptic plasticity. Neuronal T4 and T3 are metabolized by deiodinase type 3 (DIO-3) enzyme into inactive reverse T3 (rT3) and T2, respectively (Cheng, Leonard, & Davis, 2010;Lee & Petratos, 2016;Schroeder & Privalsky, 2014) (Benedetti et al., 2007). Whether the sex difference in history of suicide attempts in this study is subject to recall bias or true vulnerability for bipolar women needs to be studied further.
To our surprise, the slight, but significant, elevation in heart rate and diastolic blood pressure was observed only in male patients. Bipolar type I patients are known to have higher mortality due to cardiovascular risk factors such as hypertension (Murray, Weiner, Prabhakar, & Fiedorowicz, 2009;Prieto et al., 2016). This increase in heart rate and diastolic blood pressure could stem from alterations in the autonomic nervous system (Nardelli, Valenza, Gentili, Lanata, & Scilingo, 2014) or activation of the hypothalamic-pituitary-adrenal axis (Watson, Gallagher, Ritchie, Ferrier, & Young, 2004). However, sex-specific effects need to be further investigated in subsequent studies.
Interestingly, we found significant sex differences in several physiological measures (Table 3)  Other studies found increased free T4 index in mixed mania (Joffe, Young, Cooke, & Robb, 1994) and blunted TSH response to TRH challenge in manic patients (Extein et al., 1980). Collectively, these results suggest that acute mania is associated with altered peripheral thyroid indices. However, we cannot role an effect for antipsychotic medications on the low FT3 levels we observed (Bunevicius, Steibliene, & Prange, 2014). Regardless to the basis for the change in FT3, and given the significant role of optimal thyroid economy on brain function (see Figure 3 for details), further studies are needed to investigate thyroid dynamic changes during acute manic episode and its relationship with sex-specific variability in mania phenotype.
The results of this study should be viewed in the light of its limitations and strength. The study was conducted in Turkey where alcohol and drug comorbidities are less reported compared to western countries. The number of female participants who reported history of alcohol use was not large enough to use it as a covariate in the analysis. However, future studies are needed to test the validity of our results in other non-HMO western European countries. Here, it is relevant to note that we aimed to design this study in a non-HMO facility where the hospital LOS is typically longer (Blader, 2011).
Manic patients in our cohort were hospitalized for an average of about 3 weeks which is consistent with the average LOS in European countries and Turkey (Karamustafalioglu et al., 2014).
Interestingly, patients in our cohort received relatively high doses of antipsychotic medications (about 900 mg/day of Chlorpromazine equivalent) in conjunction with lithium and VPA. This complex pharmacotherapy strategy is not unusual when treating severe mania during acute hospitalization (Zarate & Quiroz, 2003). Zarate and Quiroz reviewed controlled trials looking at treatment modalities for acute mania and reported that the most useful treatment is the combination of antipsychotics with mood stabilizers: lithium and VPA (Zarate & Quiroz, 2003). Typically, the medication regimen is simplified during post discharge outpatient follow-up visits.
The effort to enroll 57 acutely manic patients in this pilot naturalistic study and track the progression of their symptoms on daily basis is remarkable compared to the few other studies reporting sex differences, which were mostly retrospective and cross-sectional in nature (Hendrick et al., 2000;Kumar & Ram, 2001;Viguera et al., 2000;Young et al., 2007). However, future studies should achieve larger cohort sizes to permit for examining the relationship between individual treatments and outcome of specific symptoms domains.
In conclusion, we found significant sex differences in sexual interest, aggressive and disruptive behavior, appearance and grooming, and insight item scores on the YMRS in acute manic patients.
Total YMRS scores did not differ between males and females.
Further studies are needed to address the neurobiological basis for sex difference in mania phenotype.

ACK N OWLED G M ENT
The authors wish to thank all the patients and family members who participated in the study. This research was supported, in part, by the Intramural Research Program of the NIH, NIDA.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.

AUTH O R S ' CO NTR I B UTI O N
OAA conceived the study. OAA, GS, OS, and UC designed the study.
GS and OS collected the data. OAA and BS analyzed the data and wrote the first draft. All authors contributed to data interpretation, manuscript writing, and revisions.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data files are available upon request.