Obstetric complications and postpartum psychosis: a follow-up study of 1.1 million first-time mothers between 1975 and 2003 in Sweden.

OBJECTIVE: The severity of postpartum psychosis calls for further research on the association between obstetric variables and this psychiatric disorder. METHOD: A total of 1,133368 Swedish first-time mothers were included during a 29-year period yielding 1413 hospitalized cases of postpartum psychosis. Several obstetric variables were analysed separately after adjustment for possible confounders. RESULTS: Respiratory disorder in the neonate, severe birth asphyxia, preterm birth, caesarean section, perinatal death and SGA infant were associated with an increased risk of postpartum psychosis. After adjustment for previous hospitalization for psychiatric disorder only preterm birth and acute caesarean section remained significant risk factors for postpartum psychosis (relative risks were 1.20 and 1.31 respectively). The relative risk of postpartum psychosis among first-time mothers with previous hospitalization for psychiatric disorder was increased more than 100-fold. CONCLUSION: Careful clinical risk assessments of postpartum psychosis are crucial among women with a history of psychiatric disorder whereas obstetric variables have a minor importance.

Increased glutathione levels in neurochemically identified fibre systems in the aged rat lumbar motor nuclei.

The spinal cord motor nuclei have been the focus of a number of investigations exploring neurodegenerative mechanisms, e.g. excitotoxicity mediated by glutamate and oxidative stress. Here, high-resolution quantitative post-embedding immunocytochemistry with antibodies to oxidized and reduced glutathione (GSH), an ubiquitously expressed scavenger of free radicals, was used to examine if GSH synthesis is upregulated pre- and/or postsynaptically in the lumbar motor nuclei of aged (30 month old) rats. The purpose was, moreover, to resolve the extent of correlation between GSH expression, transmitter identity and degenerative changes. Tissue from young adult rats was co-processed for comparison. The quantitative immunogold analysis revealed an increase in GSH-immunoreactivity in both pre- and postsynaptic compartments in the lumbar motor nuclei of aged rats. Presynaptically, the enrichment of GSH-immunoreactivity was seen in axonal boutons of normal appearance, and was furthermore restricted to the extra-mitochondrial compartment. Postsynaptically, the aged rats disclosed, in comparison with young adults, higher values for GSH-immunoreactivity both over mitochondria (+49%) and cytoplasmic matrix (+130%). When analysing the transmitter identity of the bouton profiles, it turned out that close to 50% of all glutamate-immunoreactive boutons in the aged rats contained very high levels (> 40 gold particles/microm2) of GSH-immunoreactivity. Strong GSH-immunoreactivity was also a typical feature of a subset of axon terminal- and axon fibre-like profiles in the aged rat that showed signs of axon dystrophy and degeneration. When comparing with normally appearing axon fibre profiles located in close vicinity, the population of aberrant axons had higher average levels of glutamate-immunoreactivity (+93%), and lower average levels of glycine-immunoreactivity (-88%). No difference was seen regarding the levels of GABA. The results of this study lend support to the idea that aging in the spinal cord motor nuclei is associated with an increased oxidative stress and indicate that different transmitter systems are differentially affected by the degenerative process.

Different modes of expression of AMPA and NMDA receptors in hippocampal synapses.

Postembedding immunogold labeling was used to determine the relationship between AMPA and NMDA receptor density and size of Schaffer collateral-commissural (SCC) synapses of the adult rat. All SCC synapses expressed NMDA receptors. AMPA and NMDA receptors were colocalized in at least 75% of SCC synapses; the ratio of AMPA to NMDA receptors was a linear function of postsynaptic density (PSD) diameter, with AMPA receptor number dropping to zero at a PSD diameter of approximately 180 nm. These findings indicate that 'silent' SCC synapses are smaller than the majority of SCC synapses at which AMPA and NMDA receptors are colocalized. Thus synapse size may determine important properties of SCC synapses.

Phosphate activated glutaminase is concentrated in mitochondria of sensory hair cells in rat inner ear: a high resolution immunogold study.

Glutamate has been implicated in signal transmission between sensory hair cells and afferent fibers in the inner ear. However, the mechanisms responsible for glutamate replenishment in these cells are not known. Here we provide evidence that phosphate activated glutaminase, which is thought to be the predominant glutamate-synthesizing enzyme in the brain, is concentrated in all types of hair cell in the organ of Corti and vestibular epithelium. By use of two different antibodies (directed to the N and C terminus, respectively) it was shown that glutaminase is largely restricted to mitochondria and that part of the enzyme pool is associated with the inner membrane of this organelle. Quantitative analysis of immunogold labelled Lowicryl sections revealed that the level of glutaminase immunoreactivity in mitochondria of supporting cells is less than 15% of that in hair cell mitochondria. Using triple labelling for glutaminase, glutamate, and glutamine, evidence was provided of a positive correlation between the glutamate/glutamine ratio and the level of glutaminase immunoreactivity, suggesting that the glutaminase antibodies identify a functional enzyme pool. Our results strengthen the idea that glutamate is a hair cell transmitter and indicate that the sensory epithelia in the inner ear show a metabolic compartmentation analogous to that in the brain.

Decreased axosomatic input to motoneurons and astrogliosis in the spinal cord of aged rats.

An increasing body of evidence indicates that aging-related impairments of nervous functions are caused by damage to neuron integrity rather than by loss of neurons. By using electron microscopy, we have examined axosomatic boutons on spinal cord motoneurons derived from aged and young adult Sprague-Dawley rats. The main finding was that about half of the examined motoneuron somata from aged rats had a reduced (50%) bouton coverage, which seemed to be caused by a smaller number of axosomatic bouton profiles. Long stretches of the cell body plasma membrane were apposed by pale processes, and immunolabeling for glial fibrillary acidic protein (GFAP) disclosed that a number of the aged motoneurons appeared embedded in GFAP immunopositive processes. Lumbar motoneurons seemed to be more severely affected than cervical motoneurons. At the ultrastructural level, affected motoneurons disclosed plasma membrane irregularities with appendages/sprout-like extensions that in some cases were sites for axosomatic contacts.

Enkephalin-, thyrotropin-releasing hormone- and substance P-immunoreactive axonal innervation of the ventrolateral dendritic bundle in the cat sacral spinal cord: an ultrastructural study.

The distribution and synaptic arrangement of thyrotropin-releasing hormone-, substance P- and enkephalin-immunoreactive axonal boutons have been studied in the ventrolateral nucleus (Onuf's nucleus) of the upper sacral spinal cord segments in the cat. For this purpose, the peroxidase-antiperoxidase immunohistochemical technique was used. Immunoreactive axonal boutons were traced in complete series of sections in order to reveal synaptic contacts with the bundled dendrites of the ventrolateral nucleus. As judged from the cross-sectional diameter of the postsynaptic dendrites, the distribution of immunoreactive boutons was non-random. Enkephalin-immunoreactive axonal boutons, presumed to be mostly of segmental origin, displayed a rather restricted distribution to mainly (> 80%) medium-to-large dendrites. Thyrotropin-releasing hormone-immunoreactive boutons, that derive from supraspinal levels, were also found to impinge on medium-to-large dendrites (> 80%), indicating a proximal location within the dendritic trees. The skewness toward large postsynaptic dendrites was even more marked for thyrotropin-releasing hormone- than for enkephalin-immunoreactive boutons. Substance P-immunoreactive boutons, that are of either supraspinal or spinal origin, showed a more even distribution throughout the dendritic trees, including both thin distal branches and thick proximal dendrites. In view of the well-known fact that virtually all thyrotropin-releasing hormone-immunoreactive boutons in the ventral horn co-contain substance P (and serotonin) it was assumed that substance P-immunoreactive boutons in synaptic contact with the finest-calibre dendrites as well as most of those with a very proximal juxtasomatic location on the dendritic trees were of segmental origin, while those impinging on medium-to-large dendrites could be of either spinal or supraspinal origin. Fine-calibre dendrites (< 1 micron) represent about 25% of the dendritic branches in the ventrolateral nucleus, but receive, with the exception of substance P (8%), very little (< 3%) peptidergic or GABAergic (Ramírez-León and Ulfhake, 1993) input, although the degree of dendritic membrane covering by bouton profiles in the ventrolateral nucleus does not seem to vary much with the calibre of the postsynaptic dendrite (Ramírez-León and Ulfhake, 1993). Both substance P- and enkephalin-immunoreactive axonal boutons established synaptic contact with more than one dendrite. Furthermore, one and the same bouton could be found in contact with two dendrites that were coupled to each other by a dendro-dendritic contact of desmosomal or puncta adherentia type.(ABSTRACT TRUNCATED AT 400 WORDS)

Serotoninergic, peptidergic and GABAergic innervation of the ventrolateral and dorsolateral motor nuclei in the cat S1/S2 segments: an immunofluorescence study.

Indirect single- and double-staining immunofluorescence techniques were used to study the serotoninergic, peptidergic and GABAergic innervation of the ventrolateral (Onuf's nucleus) and dorsolateral (innervating intrinsic foot sole muscles) nuclei, located in the S1/S2 segments of the cat spinal cord. The relative density of 5-hydroxytryptamine-, thyrotropin-releasing hormone-, substance P- and gamma-aminobutyric acid-immunoreactive axonal varicosities was similar in both nuclei. The highest relative density was recorded for varicosities immunoreactive to gamma-aminobutyric acid, while those immunoreactive to 5-hydroxytryptamine or thyrotropin-releasing hormone yielded the lowest values. The density of enkephalin-immunoreactive varicosities was higher in the ventrolateral than in the dorsolateral nucleus. Calcitonin gene-related peptide-like immunoreactivity could be seen in neurons of the ventrolateral and dorsolateral nuclei. Occasionally, calcitonin gene-related peptide-immunoreactive axonal fibers were also encountered in these nuclei. Virtually all thyrotropin-releasing hormone-immunoreactive varicosities in the ventrolateral and dorsolateral nuclei also contained 5-hydroxytryptamine-like immunoreactivity, while a somewhat smaller number of them were co-localized with substance P. About 5-10% of the 5-hydroxytryptamine-immunoreactive varicosities were devoid of peptide-like immunoreactivity, and the number of 5-hydroxytryptamine-immunoreactive varicosities lacking thyrotropin-releasing hormone-like immunoreactivity was higher in the dorsolateral than in the ventrolateral nucleus. Finally, the free fraction of substance P-immunoreactive varicosities, i.e., those lacking both 5-hydroxytryptamine and thyrotropin-releasing hormone, was about 39% in the ventrolateral and 26% in the dorsolateral nucleus. Spinal cord transection at the lower thoracic level induced a depletion of 5-hydroxytryptamine and thyrotropin-releasing hormone-immunoreactive fibers from the ventrolateral and dorsolateral nuclei, indicating an exclusive supraspinal origin for these fibers. A reduction in substance P-like immunoreactivity following spinal cord transection alone or spinal cord transection combined with unilateral dorsal rhizotomy was also detected in both nuclei, suggesting a dual origin for substance P-immunoreactive fibers, i.e., both supra- and intraspinal. The decrease in number of substance P-immunoreactive fibers was however smaller than expected from the analysis of the fraction of substance P-immunoreactive fibers co-localized with 5-hydroxytryptamine, indicating thus that the experimental lesions may have triggered a sprouting of substance P-immunoreactive axons originating from spinal cord sources. The distribution of gamma-aminobutyric acid in the ventrolateral and dorsolateral nuclei was not affected by the different lesion paradigms. It is therefore assumed that these inputs are intrinsic to the spinal cord.(ABSTRACT TRUNCATED AT 250 WORDS)

GABA-like immunoreactive innervation and dendro-dendritic contacts in the ventrolateral dendritic bundle in the cat S1 spinal cord segment: an electron microscopic study.

The motoneurons (MNs) in the ventrolateral nucleus (VLN) of the upper sacral spinal cord segments in the cat supply the external sphincters and the ischiocavernosii muscles. The dendrites of the MNs in the VLN are arranged into rostro-caudally oriented bundles (ventrolateral dendritic bundle, VLB). In this study we describe the distribution and synaptic arrangement of gamma-aminobutyric acid-immunoreactive (GABA-IR) axonal bouton profiles innervating the VLB. This was accomplished using the peroxidase-antiperoxidase technique and a polyclonal antibody raised against glutaraldehyde-conjugated GABA. The VLN receives an extensive innervation of GABA-IR axonal bouton profiles that surround both cell bodies and dendrites. Twenty-five per cent of the total number of vesicle-containing axonal profiles in the VLN neuropil were estimated to be GABA-IR. On cell bodies in the alpha-motoneuron size-range, the membrane covering of GABA-IR bouton profiles was about 18% and they constituted about 29% of the total membrane covering of axonal bouton profiles. Quantitative analysis of GABA-IR bouton profiles on dendrites revealed membrane covering figures rather similar to those on the cell bodies. They were not randomly distributed within the dendritic arborisations. Instead, they were very infrequent (2.5% of the covering) on small calibre dendrites (< 1 microns) as compared to larger dendrites (> 1 microns, 14-18.5% of the covering), although the total membrane covering of axonal bouton profiles was rather similar for all dendrites (42-52%). The data on membrane covering by GABA-IR boutons presented here may be low estimates due to technical limitations, indicating that the GABAergic input to this region might be even more extensive. A frequent finding was that one and the same GABA-IR bouton made synaptic contact with two to three adjacent dendrites. This type of synaptic arrangement among the VLN MNs indicates a divergence of the GABAergic input at the terminal level. In addition, the postsynaptic dendrites involved in such arrangements often disclosed dendro-dendritic contacts. In total, 44% of the bundled dendrites in the VLN disclosed direct dendro-dendritic contact regions. These contacts were most often of the puncta adherentia type, while desmosome-type contacts were less frequent. None of the dendro-dendritic contacts studies had the characteristics of a gap junction. Taken together, the present results indicate that GABA may be a transmitter substance in a large fraction of the synaptic input to the VLN MNs.(ABSTRACT TRUNCATED AT 400 WORDS)