Lateral ventricle volume and psychotic features in adolescents and adults with bipolar disorder

Abstract

This magnetic resonance imaging study demonstrates increased lateral ventricle volume (LVV) in adolescents and adults with bipolar disorder (BD) with psychotic symptoms, but not without psychosis, compared to healthy adolescents and adults. This suggests LVV is a morphologic feature associated with psychosis in BD, present by adolescence.

1. Introduction

Varying reports of increases or no differences in lateral ventricle volume (LVV) in adults with bipolar disorder (BD) (Andreasen et al., 1990; Ali et al., 2001; Brambilla et al., 2001; McDonald et al., 2006; Kempton et al.; 2008; Rosa et al., 2010) could result from clinical heterogeneity (Strasser et al., 2005). Increased LVV is a consistent finding in psychotic disorders (Wright et al., 2000), long implicated in psychosis development (Johnstone et al., 1976). This study tested the hypothesis LVV is increased in adolescents and adults with BD with psychotic symptoms (PBD), but not in those who have not experienced psychosis (NPBD).

2. Methods

Thirty-six individuals with PBD (ages 14–56yrs, 61% female, 33% adolescents ≤21years), 48 with NPBD (10–59yrs, 58% female, 35% adolescents) and 79 healthy comparison (HC) participants without personal history or first-degree relative with an Axis I disorder (10–57yrs, 54% female, 42% adolescents) (non-LVV data on 64 included in Kalmar et al., 2009; Womer et al., 2009) were recruited from Yale University and Veterans Affairs medical centers and advertisement in the community. Written informed consent was obtained from parents/guardians of minors and participants ≥18yrs, and written assent from minors, in accordance with institutional review boards of the Yale School of Medicine and Department of Veterans Affairs.

The presence or absence of DSM IV Axis I Disorders and mood state at scanning were confirmed by administration of the revised Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime to participants ≤18yrs and their parents/guardians (Kaufman et al., 1997), or Structured Clinical Interview for DSM-IV Axis I Disorders for participants over 18yrs (First et al., New York State Psychiatric Institute, 2002). PBD was defined by history of hallucinations or delusions determined by consensus of subject self-report, structured and clinical interviews, and discussion with treaters. No participant had head trauma with loss of consciousness over 5min, or a major neurological or medical disorder, except 5 BD participants with treated hypothyroidism.

Twenty-one percent (7 PBD, 11 NPBD) of BD participants were unmedicated at scanning; the remainder was prescribed lithium carbonate (10 PBD, 10 NPBD), anticonvulsants (16 PBD, 22 NPBD), atypical antipsychotics (17 PBD, 19 NPBD), antidepressants (11 PBD, 19 NPBD) or benzodiazepines (5 PBD, 7 NPBD). One BD subject was never-medicated. Sixty-one percent of BD participants were euthymic (21 PBD, 30 NPBD), 25% were experiencing an elevated (manic/mixed/hypomanic) mood episode (7 PBD, 14 NPBD) and 14% were depressed (8 PBD, 4 NPBD). Forty-eight percent of BD subjects had rapid-cycling or chronic mood symptoms (17 PBD, 23 NPBD). Thirty-five percent of BD participants (15PBD, 14NPBD) had a history of alcohol or other substance abuse or dependence in remission. Five BD subjects had comorbid post traumatic stress disorder [4PBD (1 also with social phobia and 1 specific phobia), 1 NPBD], 2 panic disorder (1PBD, 1 NPBD), 2 generalized anxiety disorder (1PBD, 1NPBD) and 2PBD a specific phobia. Attention deficit hyperactivity disorder and oppositional defiant disorder were assessed in adolescents and present in 2PBD/4NPBD and 1 NPBD participant respectively.

MRI scans were obtained on a 3T Trio MR Scanner (Siemens, Erlangen, Germany) using a three-dimensional Magnetization Prepared Rapid Acquisition Gradient Echo T1-weighted sequence (TR=1500ms, TE=2.83ms, FOV=256 × 256 mm², matrix=256 × 256, 160 1.0mm contiguous sagittal slices, NEX=2). Images were aligned along the anterior commissure-posterior commissure plane. Gray matter, white matter, and cerebrospinal fluid segmentation were performed using the Statistical Parametric Mapping 99 (SPM99) (www.fil.ion.ucl.ac.uk) tissue classification algorithm. The lateral ventricles were delineated by hand on axial slices and confirmed in orthogonal planes using BioImage Suite software (www.bioimagesuite.org) by operators (EEE, EH, LS) blind to participant characteristics: interrater intraclass reliability coefficient >0.99, intrarater intraclass reliability coefficient >0.97. Gray and white matter were summed for total brain volume (TBV) (Womer et al., 2009).

Analyses were conducted using an analysis of covariance (ANCOVA) model that included LVV as the dependent measure, diagnosis (HC, PBD, NPBD) as a between-subjects factor, and age and TBV as covariates. The distributions of LVVs were skewed (McDonald et al., 2006); values were logarithmically transformed successfully before analyses. Hypothesis-testing was performed by pairwise comparison between the HC and PBD group; additional posthoc pairwise comparisons were also performed. All significant main effects and 2-way interactions (p<0.05) are reported below.

3. Results

Groups did not differ significantly in age (HCmean27.5yrs±SD 14.1, NPBD 29.7±13.7, PBD 29.0±12.7, p=0.69). Age and TBV were both positively associated with LVV (p<0.0001). LVV was larger in the PBD group, compared to the HC group (p=0.034) (Figure), but did not differ between the NPBD and either the HC or PBD groups (unadjusted p=0.66, p=0.12, respectively). There was no significant interaction of age and group on LVV (p=0.75). No significant effects of mood state, rapid-cycling or medication subclasses (lithium/anticonvulsants/antipsychotics/antidepressants) were detected.

Figure 1.

Lateral Ventricle Volume in Psychotic Bipolar Disorder (PBD), Nonpsychotic BD (NPBD) and Healthy Comparison (HC) Groups

Least square means adjusted for age and total brain volume±standard errors *p<0.05.

4. Discussion

This is the first report of which we are aware of increased LVV in adolescents and adults with PBD, but not NPBD, relative to HC participants. The findings suggest LVV may be a morphologic feature associated with psychotic symptoms in BD present by adolescence. Increased LVV has been reported previously in one study of adults with PBD (Strasser et al., 2005). Its presence in adolescents and adults could suggest a developmental mechanism. Alternatively, increased ventricle size might reflect the presence of a degenerative mechanism, early brain insult or progressive periventricular volume losses resulting from mood episodes (Strakowski et al., 2002). However, longitudinal studies of larger samples are needed to assess age-related effects. Combined imaging and genetic studies may help elucidate etiologies, particularly given reports relating risk alleles in catechol-O-methyl transferase and neuregulin 1 both to susceptibility to PBD and schizophrenia and to increases in LVV in these disorders (Crespo-Facorro et al., 2007; Mata et al., 2009).

Effects of medication subclasses were not detected but power was limited; studies of medication-naive individuals would provide more definitive evidence. Though there were comparable rates of alcohol/substance comorbidity, as for medication use, retrospective reporting was not considered sufficiently reliable to investigate effects of duration and severity of use. An effect of rapid-cycling was not detected, suggesting LVV is not associated with mood episodes; however, analyses for episode numbers were not performed as these could not be reliably assessed, especially in a sample with high rapid-cycling rates. Future studies that employ longitudinal designs with larger samples of adolescents and adults and examine clinical factors, such as number of mood episodes, psychosis symptom severity, duration and severity of alcohol/substance comorbidity, amount and duration of medication exposure, as well as functional outcomes, may elucidate the mechanisms and effects of increased LVV in PBD.