Introduction
Inter-individual differences in brain structure and function are partly due to environmental factors (Davidson & McEwen, Reference Davidson and McEwen2012). Psychosocial stress is one such factor that may lead to neural changes, and these alterations can have both adaptive and maladaptive consequences (McEwen & Gianaros, Reference McEwen and Gianaros2011). Although most people show resilience to psychosocial stress (Gillespie et al. Reference Gillespie, Phifer, Bradley and Ressler2009), the experience of stress may result in traumatic memories and development of neuropsychiatric disorders (Kessler et al. Reference Kessler, Davis and Kendler1997; Wang et al. Reference Wang, Wahlberg, Karp, Winblad and Fratiglioni2012). Furthermore, psychosocial stress can lead to long-term morphological and functional changes in the brain (Bremner, Reference Bremner2006; Tottenham et al. Reference Tottenham, Hare, Quinn, McCarry, Nurse, Gilhooly, Millner, Galvan, Davidson, Eigsti, Thomas, Freed, Booma, Gunnar, Altemus, Aronson and Casey2010), which may underlie cognitive and behavioural symptoms observed in mood and anxiety disorders.
Two limbic brain regions that are thought to be particularly vulnerable to psychosocial stress are the hippocampus (McEwen, Reference McEwen2007) and the amygdala (Roozendaal et al. Reference Roozendaal, McEwen and Chattarji2009). Experimental animal studies have shown opposing effects of chronic stress on these two structures: after the experience of a stressor, the hippocampus shows impaired neurogenesis and atrophy or remodelling of the apical dendrites of pyramidal neurons in the CA3 region (McEwen, Reference McEwen2001), whereas the same stressor causes dendritic growth in basolateral amygdala (Vyas et al. Reference Vyas, Mitra, Shankaranarayana Rao and Chattarji2002). In humans, smaller hippocampal volume is a recurrent finding in post-traumatic stress disorder (PTSD; Bremner et al. Reference Bremner, Elzinga, Schmahl and Vermetten2008) and in young and middle-aged adults who experienced stressful events during childhood (Teicher et al. Reference Teicher, Anderson and Polcari2012). With regard to amygdala volume, inconsistent findings have been reported, where both smaller (Woon & Hedges, Reference Woon and Hedges2009; Morey et al. Reference Morey, Gold, Labar, Beall, Brown, Haswell, Nasser, Wagner and McCarthy2012) and larger (Tottenham et al. Reference Tottenham, Hare, Quinn, McCarry, Nurse, Gilhooly, Millner, Galvan, Davidson, Eigsti, Thomas, Freed, Booma, Gunnar, Altemus, Aronson and Casey2010; Kuo et al. Reference Kuo, Kaloupek and Woodward2012) volumes have been observed in patients with PTSD and in relation to psychosocial stress.
There is increasing evidence that the timing of a stressor determines how much impact it will have on the brain (Teicher et al. Reference Teicher, Tomoda and Andersen2006). Experience of stress is thought to have a detrimental effect on the brain especially during stages of life when the brain undergoes major changes, i.e. development in early life and deterioration in late life (Wolkowitz et al. Reference Wolkowitz, Epel, Reus and Mellon2010). Moreover, the effect of childhood stress can influence the brain well into adulthood, although only a few studies have investigated such long-term associations (van Harmelen et al. Reference van Harmelen, van Tol, van der Wee, Veltman, Aleman, Spinhoven, van Buchem, Zitman, Penninx and Elzinga2010).
Finally, stress may have different neural effects in men and women. As psychiatric disorders are more common in women than in men, it has been proposed that women are more vulnerable to the negative effects of stress (Troisi, Reference Troisi2001). Contrary to this assertion, however, some studies have found that men show greater structural brain effects after the experience of psychosocial stress than women (De Bellis, Reference De Bellis2001), whereas others have failed to find such gender differences (Woon & Hedges, Reference Woon and Hedges2011).
The current study aims to explore (a) the effect of stressful life events experienced from childhood to late life on hippocampal and amygdala volumes in a large cohort of non-demented older adults, (b) whether the experience of stressful life events has different effects on the brain in men and women and (c) whether the experience of stressful life events exacerbates age-related volume differences of the hippocampus and amygdala.
Method
Study sample
The study population consisted of a sample of 555 persons who were recruited to participate in a Magnetic Resonance Imaging (MRI) study, from the larger Swedish National Study of Ageing and Care in Kungsholmen (SNAC-K), a longitudinal, multidisciplinary study on ageing and health initiated in 2001 that included 3363 participants aged ⩾60 years at baseline (Lagergren et al. Reference Lagergren, Fratiglioni, Hallberg, Berglund, Elmståhl, Hagberg, Holst, Rennemark, Sjölund, Thorslund, Wiberg, Winblad and Wimo2004). In short, the SNAC-K samples were stratified by age and year of assessment. Eleven age cohorts were chosen with different retest intervals: 6 years for the younger cohorts (age 60–72 years), and 3 years for the older cohorts (age ⩾78 years) due to more rapid changes and higher attrition in older ages. During the baseline examination, 555 individuals were randomly selected from all non-institutionalized and non-disabled participants to undergo a MRI examination. Individuals with severe cerebral diseases that affect brain integrity were excluded (e.g. stroke, n = 21). Participants with a clinical diagnosis of depression or an anxiety disorder and/or who used antidepressant medication were excluded from the analyses (N = 45). Full brain and skull coverage was required for the MRI datasets and detailed quality control was carried out on all MR images, according to established criteria (Simmons et al. Reference Simmons, Westman, Muehlboeck, Mecocci, Vellas, Tsolaki, Kłoszewska, Wahlund, Soininen, Lovestone, Evans and Spenger2011). After exclusion of bad-quality MR images (n = 25), complete data were available for 466 non-demented subjects.
MRI acquisition
MRI scanning was undertaken on a 1.5 T scanner (Philips Intera, The Netherlands) on which 3D FFE (fast-field echo) T1, axial SE (spin echo) proton density/T2, axial FLAIR (fluid-attenuated inversion recovery) and axial DTI (diffusion tensor imaging) were acquired. In this study, the 3D FFE T1 images (TR = 15 ms, TE = 7 ms, flip angle = 15°, number of axial slices = 128, thickness = 1.5 mm, in-plane resolution = 0.94 × 0.94 mm, no gap, field of view = 240 × 240, matrix = 196 × 256) were used for assessment of hippocampal and amygdala volumes.
Image processing
The T1-weighted images were segmented into grey matter, white matter, and cerebrospinal fluid (CSF) in native space using the unified segmentation approach (Ashburner, Reference Ashburner2007), as implemented in SPM8 (http://www.fil.ion.ucl.ac.uk/spm/software/spm8/). Segmentation accuracy was carefully verified and intracranial volume (ICV) was computed by adding grey matter, white matter, and CSF volumes, which included all structures inside the inner table of the skull (total brain tissue, ventricles and extra-ventricular CSF, brain stem, and cerebellum).
Automatic segmentation of amygdala and hippocampal volumes
Automatic subcortical volumetric segmentation was performed using the Freesurfer image analysis suite (v. 5.0.1, Martinos Center for Biomedical Imaging, Harvard-MIT, Boston, USA; http://surfer.nmr.mgh.harvard.edu/). The technical details of these procedures have been described previously (Fischl et al. Reference Fischl, Salat, Busa, Albert, Dieterich, Haselgrove, van der Kouwe, Killiany, Kennedy, Klaveness, Montillo, Makris, Rosen and Dale2002). Briefly, this fully automated process includes motion correction, removal of non-brain tissue, automated Talairach transformation, segmentation of subcortical white-matter and deep grey-matter volumetric structures (including hippocampus, amygdala, caudate, putamen, and ventricles), intensity normalization, and cortical reconstruction.
This segmentation procedure assigns a neuroanatomical label to every voxel in the MR image volume. The method is based on probabilistic information estimated from a manually labelled training set. Markov Random Field Theory is applied, where the probability of a label at a given voxel is computed not just in terms of the greyscale intensities and prior probabilities at that voxel, but also as a function of the labels in the neighbourhood around the voxel in question. This feature is important for correct separation of hippocampus and amygdala, which have similar greyscale values.
Psychosocial stress measures
Psychosocial stress was represented by negative life events and measured using a structured questionnaire consisting of 18 pre-defined events (Miller & Rahe, Reference Miller and Rahe1997). For each event, participants indicated whether it had happened and, if so, the age at which it happened (Table 1). Some events could have occurred more than once over the life course; each occurrence was counted as a single event. The reported negative life events were summed to create a total life-events score.
Age of events
To examine the effect of age at which events occurred, we computed the following variables: early-life events (events that occurred at or before 18 years); early-adulthood events (events that occurred between 18 and 40 years); middle-adulthood events (events that occurred between 40 and 65 years) and late-life events (events that occurred at or after 65 years).
Covariates
Data on demographics (age, gender, education) were collected through interviews according to standardized protocols by trained nurses. Educational level was measured by the maximum years of formal schooling and recoded into eight levels, from not finishing primary school (1) to having a PhD degree (8). In our sample the mean level of education was 5, which corresponds to having a community college degree. Body mass index (BMI) was calculated as weight (kg) divided by height (m2). Arterial blood pressure was measured twice by trained nurses with a sphygmomanometer and the mean of the two measurements was used. Smoking and alcohol use were assessed using self-reports. Other vascular and psychiatric disorders at baseline were identified through the inpatient register system, including depression, anxiety disorders, PTSD, diabetes, heart disease, and stroke (International statistical Classification of Diseases and related health problems – 10th revision; ICD-10 codes). Cognitive screening was performed by a physician using the Mini-Mental State Examination (MMSE; Folstein et al. Reference Folstein, Folstein and McHugh1975). Depressive symptoms were assessed using the Montgomery–Åsberg scale (Montgomery & Asberg, Reference Montgomery and Asberg1979).
Data analysis
Volumes for left and right hippocampus (r = 0.84, p < 0.001) and left and right amygdala (r = 0.77, p < 0.001) were highly related, and aggregated into total volumes for both regions. Hippocampal and amygdala volumes were positively related (r = 0.70, p < 0.001), the association did not differ across hemispheres All analyses were conducted in separate models for hippocampal and amygdala volumes.
Before summing all events in stress variables, we ensured that there were no effects of specific single life events, by examining the relationship between each category of life events (e.g. widowhood, severe illness, etc.) and hippocampal and amygdala volumes, in linear regression models, adjusted for age, gender and ICV.
To investigate the effects of the computed stress variables on hippocampal and amygdala volumes we conducted the following analysis:
First, we conducted a linear regression with total number of events reported as determinant of brain volumes. Second, we investigated the relationship between age of negative life events occurrence and brain volumes by creating generalized linear models with reported stress in each age period (early-life, early-adulthood, middle-adulthood, and late-life events) as independent variables.
To test for the interactive effect of stressful life events with age and gender on the brain, in a third step, interaction terms of the stress variables with age and with gender were added to the models. In case of significant interactions, analyses were repeated in separate groups (e.g. men v. women or events v. no events).
All analyses were adjusted for age, gender, educational level, ICV, BMI, alcohol use, current smoking, diabetes mellitus, systolic and diastolic blood pressure, depressive symptoms, and MMSE score.
To ensure that the effects of life events were selective for amygdala and hippocampal volumes we repeated all analyses with caudate nucleus volume and total brain volume as outcome.
Results
Data were available for 466 subjects (age range 60–96 years), of whom 58% were female. Compared to subjects who were excluded from the study (n = 89), the study sample was younger, had lower BMI and larger hippocampal volumes (p < 0.05), but no differences were found on any other variable (Table 2). The mean number of total reported stressful life events was five, with a maximum of 14 (Table 1). None of the single-event categories was associated with hippocampal or amygdala volumes (all p > 0.05).
Table 1. Frequencies of reported negative life events per age category
Table 2. Subjects’ characteristics
a Medians with 10–90% range in Mini-Mental State Examination (MMSE) score and MADRS depressive symptoms were reported because of non-normally distributed data.
Association between stressful life events and amygdala volume
In the total sample, total number of life events was positively associated with amygdala volume (F 2,450 = 3.259, p = 0.017) (Fig. 1a ). When the separate stress measures per lifespan period were examined, we found that larger amygdala volume was related to late-life events (F 2,443 = 3.946, p = 0.020), but not to early- or middle-adulthood events (F < 1.5, p > 0.2 (Fig. 1b ). There was no interaction between the life-events measures and gender for amygdala volume (p > 0.1).
Fig. 1. Adjusted amygdala volume in relation to total number of negative life events over the lifespan (a) and number of late-life events (b). Error bars represent standard errors (*p < 0.05). Adjusted for age, gender, educational level, intracranial volume (ICV), body mass index, alcohol use, current smoking, diabetes mellitus, systolic and diastolic blood pressure, depressive symptoms, and Mini-Mental State Examination score.
Association between stressful life events and hippocampal volume
None of the age of event measures was associated with hippocampal volume (F < 1.5, p > 0.2). There was trend-level association between number of early-life events and hippocampal volume (F 2,443 = 2.04, p = 0.10), where Fisher's least significant difference post-hoc testing showed that subjects who reported two or more early-life events had smaller hippocampal volume than subjects who reported no early-life events [mean hippocampal volume difference: −198.3 μl; 95% confidence interval (CI) −2.8 to −393.7, p = 0.05].
Association between stressful life events and ‘control’ brain areas
Total brain volume and caudate nucleus volume were not associated with any of the life-event variables (F < 1.5, p > 0.2).
Interaction between gender and stress variables
We found an interaction between gender and early-life events for hippocampal volume (B = 27.6, 95% CI 17.7–37.6, p < 0.001). To trace the source of this interaction, we repeated the analyses separately for men and women. We found smaller hippocampal volume in men who reported ⩾2 early-life events compared to men who reported none or one event (F 2,180 = 6.688, p = 0.002). This effect was not seen for early-adulthood, middle-adulthood or late-life events, and there were no effects for women (all F < 1.5, p > 0.2) (Fig. 2).
Fig. 2. Mean ICV-adjusted hippocampal volume in relation to number of early-life events by gender. Error bars represent standard errors (*p < 0.05). Adjusted for age, educational level, intracranial volume, body mass index, alcohol use, current smoking, diabetes mellitus, systolic and diastolic blood pressure, depressive symptoms, and Mini-Mental State Examination score.
Interaction between age and stress variables
We found an interaction between age and early-life events for amygdala volume (B = − 4.14, 95% CI −8.08 to −0.19, p = 0.04) and a borderline significant interaction for hippocampal volume (B = − 8.43, 95% CI −17.44 to 0.56, p = 0.06). To trace the source of this interaction, we plotted the estimated amygdala and hippocampal volumes against age with separate lines for persons who reported no early-life events, one early-life event and ⩾2 early-life events (Fig. 3). This finding was specific for early-life events, as we found no interaction with adulthood, mid-life or late-life events (all p for interactions > 0.05).
Fig. 3. Estimated volume differences with age in (a) amygdala and (b) hippocampus for early-life stress groups. Adjusted for gender, educational level, intracranial volume, body mass index, alcohol use, current smoking, diabetes mellitus, systolic and diastolic blood pressure, depressive symptoms, and Mini-Mental State Examination score.
Compared to persons who reported none or only one early-life event, persons with ⩾2 early-life events showed smaller amygdala volume (B for difference = −18.03, 95% CI −34.11 to −1.95, p = 0.03) and smaller hippocampal volume (B for difference = − 36.42, 95% CI −71.72 to −1.117, p = 0.04) with increasing age.
Because of the observed interaction between gender and early-life stress, we examined whether there was a significant three-way interaction between early-life stress, gender and age, but none of these interaction terms reached significance (p > 0.05).
Discussion
In this large sample of adults aged ⩾60 years, self-reported negative life events, particularly late-life events, were associated with larger amygdala volume in both men and women, whereas smaller hippocampal volumes were associated only with early-life events and in men. Furthermore, we found that ⩾2 early-life events exacerbated the age-related volume decline in both the hippocampus and the amygdala.
The experience of stress triggers the activation of the hypothalamic–pituitary–adrenal axis and the end product of this activation is the secretion of glucocorticoids and catecholamines (adrenaline and noradrenaline). Adrenergic receptors are abundant in the amygdala, a region critical to fear processing and memory for emotional information (Roozendaal et al. Reference Roozendaal, McEwen and Chattarji2009). From an evolutionary perspective, increases in amygdala volume in response to stress may lead to enhanced detection of threatening stimuli and increased survival probability. Experimental studies with animals show that the experience of acute stress strengthens the structural basis of synaptic connectivity, manifested by newly formed spines in the basolateral portion of the amygdala (Vyas et al. Reference Vyas, Mitra, Shankaranarayana Rao and Chattarji2002). If the same acute stress is repeated for several days, it eventually leads to more extensive spinogenesis. This strengthening of the structural basis of synaptic connectivity could even lead to persistent dendritic elongation (Zhang & Rosenkranz, Reference Zhang and Rosenkranz2012). However, in other nuclei of the amygdala, such as the medial and central portions, these stress-related effects have not been observed (Bennur et al. Reference Bennur, Shankaranarayana Rao, Pawlak, Strickland, McEwen and Chattarji2007). These opposing effects within the amygdala might explain why both larger and smaller amygdala volumes have been reported in human studies on stress and stress-related disorders (Woon & Hedges, Reference Woon and Hedges2009; Tottenham et al. Reference Tottenham, Hare, Quinn, McCarry, Nurse, Gilhooly, Millner, Galvan, Davidson, Eigsti, Thomas, Freed, Booma, Gunnar, Altemus, Aronson and Casey2010; Kuo et al. Reference Kuo, Kaloupek and Woodward2012; Morey et al. Reference Morey, Gold, Labar, Beall, Brown, Haswell, Nasser, Wagner and McCarthy2012). Although the current amygdala findings are consistent with some animal work, higher spatial resolution in MRI (e.g. 7 T) is needed to unravel the differential effects of stress on the subnuclei of the amygdala.
The hippocampus is important in regulating stress responses and a major feedback site for glucocorticoids. This structure is also highly sensitive to neurotoxic effects of increased glucocorticoid levels (Knoops et al. Reference Knoops, Gerritsen, van der, Mali and Geerlings2010). Loss of hippocampal neurons due to stress exposure has been reported in animal studies (Sapolsky, Reference Sapolsky2003) and in neuropsychiatric disorders, such as PTSD (Bremner et al. Reference Bremner, Elzinga, Schmahl and Vermetten2008) and major depressive disorder (Macqueen & Frodl, Reference Macqueen and Frodl2011). Neural development early in life undergoes many phases and it has been suggested that brain structures differ in the age at which they are fully developed (Andersen & Teicher, Reference Andersen and Teicher2008). For instance, trauma experienced between the ages of 3 and 5 has been associated with much more decreased hippocampal integrity than similar events experienced at other ages (Andersen et al. Reference Andersen, Tomada, Vincow, Valente, Polcari and Teicher2008). Exposure to stressful events during the developmental period may increase secretion of glucocorticoids, which could interfere with the transcriptional mechanisms that control the expression of brain-derived neurotrophic factor and thereby increase vulnerability to attrition, resulting in volume loss (Nestler et al. Reference Nestler, Barrot, DiLeone, Eisch, Gold and Monteggia2002). Accordingly, there are more studies reporting a detrimental effect of early-life stress (Frodl et al. Reference Frodl, Reinhold, Koutsouleris, Reiser and Meisenzahl2010; van Harmelen et al. Reference van Harmelen, van Tol, van der Wee, Veltman, Aleman, Spinhoven, van Buchem, Zitman, Penninx and Elzinga2010; Teicher et al. Reference Teicher, Anderson and Polcari2012) than of late-life stress (Yehuda et al. Reference Yehuda, Golier, Tischler, Harvey, Newmark, Yang and Buchsbaum2007) on hippocampal volume. Unfortunately, we were not able to examine the effect of different age categories early in life due to the relatively low number of reported early-life events. To get a better understanding of the exact sensitive periods early in life, future studies should examine in more detail the age at which stressful events occur early in life.
In addition, our findings suggest that the effect of stress experienced early in life on hippocampal volume persists into old age in men, but not in women. Many previous studies on the effects of psychosocial stress on the brain reported no gender differences (Frodl et al. Reference Frodl, Reinhold, Koutsouleris, Reiser and Meisenzahl2010; van Harmelen et al. Reference van Harmelen, van Tol, van der Wee, Veltman, Aleman, Spinhoven, van Buchem, Zitman, Penninx and Elzinga2010; Teicher et al. Reference Teicher, Anderson and Polcari2012). However, these studies were all conducted in younger adult populations and used other methods to investigate brain volumes (e.g. voxel-based morphometry) making it difficult to compare findings to those obtained in the present study. Other studies examining gender differences in the effects of stress on brain structures found results similar to those observed in the current study. A review of the rodent literature indicates that, in female rats, stress failed to produce dendritic retraction in hippocampus, whereas in male rats clear atrophic effects of chronic stress were observed (McLaughlin et al. Reference McLaughlin, Baran and Conrad2009). Relatedly, in human studies smaller cerebral volumes and corpus callosum regions in PTSD boys as well as larger lateral ventricular volume in maltreated boys with PTSD than in maltreated girls with PTSD have been observed (De Bellis & Keshavan, Reference De Bellis and Keshavan2003). Furthermore, men have greater stress-hormone responses to experimentally induced stress than women (Kirschbaum et al. Reference Kirschbaum, Kudielka, Gaab, Schommer and Hellhammer1999) and high levels of cortisol predicted poorer memory performance in men, but not in women (Wolf et al. Reference Wolf, Schommer, Hellhammer, McEwen and Kirschbaum2001). Taken together, these findings suggest that men may show a greater response to stress than women.
The observed gender differences may also reflect a protective effect of oestrogen on stress-related influences on hippocampal integrity (Takuma et al. Reference Takuma, Matsuo, Himeno, Hoshina, Ohno, Funatsu, Arai, Kamei, Mizoguchi, Nagai, Koike, Inoue and Yamada2007). Expression of growth hormones is greater in the hippocampus in female than in male rats and growth hormones are sensitive to the presence of oestrogen (Tanapat et al. Reference Tanapat, Hastings, Reeves and Gould1999). Although the women in the current study were post-menopausal, the observed gender differences in the influence of stress on hippocampal volume were seen for events occurring in early life, a period in which there are marked gender differences in oestrogen levels (McEwen et al. Reference McEwen, Akama, Spencer-Segal, Milner and Waters2012).
Besides the gender difference, we also found an interaction between early-life events and age at which the brain volumes were assessed. Both amygdala and hippocampal volumes decrease with age, but in the group who reported ⩾2 early-life events we observed a much steeper decline of volume as a product of age. Even though our study has a cross-sectional design, the results suggest that the effect of early-life stress becomes more apparent with increasing age. Possibly, persons with a history of early-life stress are at greater risk for pathological ageing effects on the brain (e.g. neurodegenerative disorders).
All our findings taken together, there was a clear effect of timing of stressor on the brain, with stronger effects observed for events reported in early and late life. Because of morphological changes occurring in the brain during these stages, it is plausible that environmental factors have a greater influence on the brain than during more stable stages of life, especially on stress-responsive regions such as the hippocampus and amygdala (Lupien et al. Reference Lupien, McEwen, Gunnar and Heim2009). Moreover, the fact that the hippocampus and amygdala were differentially affected by early- and late-life stress suggests that the hippocampus is more vulnerable to early detrimental effects, whereas the amygdala is more reactive to stressful events occurring during later stages of life.
Even though the current findings fit well with past research (McEwen, Reference McEwen2001; Bremner, Reference Bremner2006; Yehuda et al. Reference Yehuda, Golier, Tischler, Harvey, Newmark, Yang and Buchsbaum2007; Tottenham et al. Reference Tottenham, Hare, Quinn, McCarry, Nurse, Gilhooly, Millner, Galvan, Davidson, Eigsti, Thomas, Freed, Booma, Gunnar, Altemus, Aronson and Casey2010), the possibility of reversed causality cannot be excluded. For instance, individuals with smaller hippocampal volume and larger amygdala volume might report more negative life events as a result of impaired emotion regulation (Gerritsen et al. Reference Gerritsen, Rijpkema, van Oostrom, Buitelaar, Franke, Fernández and Tendolkar2012). Longitudinal studies are needed to substantiate these intriguing timing differences.
A major strength of the current study is the large number of participants, in which hippocampal volume, amygdala volume, and stressful events throughout the entire lifespan were measured. Limitations of the study include assessment of life events by means of self-report, which may have resulted in underreporting of the actual stress experienced. Selection bias may be present, because the MRI sample was slightly younger, more educated, and cognitively advantaged compared to the remaining participants of the SNAC-K study. Further, we excluded people with dementia and those with large brain infarcts. Moreover, survival bias may be operating due to the cross-sectional nature of the study. That said, both the possible underreporting of stress-related events and selection biases may, if anything, have resulted in an underestimation of the actual effects.
In summary, our findings support the view that, depending on the timing of a negative life-event occurrence, stress affects amygdala and hippocampal volumes differently and the detrimental effects of early-life stress is not only maintained into old age but also exacerbates age-related volume decline.
Acknowledgements
L.G. was supported by a Marie Curie intra-European Fellowship of the European Community's Seventh Framework Programme under contract number PIEF-GA-2011-300355 and a Veni grant (ZonMW 916-14-016) from The Netherlands Organisation for Scientific Research. The Swedish National study on Ageing and Care (SNAC; www.snac.org) is financially supported by the Ministry of Health and Social Affairs, Sweden, the participating County Councils, Municipalities and University Departments, and the Swedish Research Council. In addition, specific grants were obtained from the Swedish Council for Working Life and Social Research, the Swedish Brain Power, and private foundations (the Konung Gustaf V:s och Drottning Victorias Frimurare Foundation, Gun and Bertil Stohnes Foundation, Gamla Tjänarinnor Foundation, Loo och Hans Ostermans funding for geriatric research). A.S. was supported by funds from the NIHR Biomedical Research Centre for Mental Health and the Biomedical Research Unit for Dementia, at the South London and Maudsley NHS Foundation Trust and Institute of Psychiatry, Kings College London, as well as Alzheimer's Research UK. L.B. was also supported by an Alexander von Humboldt Research Award and by a donation from the af Jochnick Foundation.
Declaration of Interest
None.
