Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 May 1.
Published in final edited form as: Int J Geriatr Psychiatry. 2018 Feb 15;33(5):763–768. doi: 10.1002/gps.4856

Associations between CSF cortisol and CSF norepinephrine in cognitively normal controls and patients with amnestic MCI and AD dementia

Lucy Y Wang 1,3, Murray A Raskind 1,3, Charles W Wilkinson 2,3, Jane B Shofer 3, Carl Sikkema 2, Patricia Szot 1, Joseph F Quinn 4, Douglas R Galasko 5, Elaine R Peskind 1,3
PMCID: PMC5882504  NIHMSID: NIHMS944030  PMID: 29446123

Abstract

Objective

This study evaluated the effects of Alzheimer’s disease (AD) on the relationship between the brain noradrenergic system and hypothalamic pituitary adrenocortical axis (HPA). Specifically, relationships between cerebrospinal fluid (CSF) norepinephrine (NE) and CSF cortisol were examined in cognitively normal participants and participants with AD dementia and amnestic mild cognitive impairment (aMCI). We hypothesized there would a positive association between these two measures in cognitively normal controls, and that this association would be altered in AD.

Methods

421 CSF samples were assayed for NE and cortisol in controls (n=305), participants with aMCI (n=22) and AD dementia (n=94). Linear regression was used to examine the association between CSF cortisol and NE, adjusting for age, sex, education, and body mass index.

Results

Contrary to our hypothesis, CSF cortisol and NE levels were not significantly associated in controls. However, higher cortisol levels were associated with higher NE levels in AD and aMCI participants. Regression coefficients ± standard errors for the change in cortisol per 100 pg/ml increase in NE are as follows: controls 0.0±0.2, p=1.0; MCI, 1.4±0.7, p=0.14; and AD 1.1±0.4, p=0.032. Analysis with MCI and AD participants combined strengthened statistical significance (1.2±0.3, p=0.007).

Conclusions

Enhanced responsiveness of the HPA axis to noradrenergic stimulatory regulation in AD and disruption of the blood brain barrier may contribute to these findings. Because brainstem noradrenergic stimulatory regulation of the HPA axis is substantially increased by both acute and chronic stress, these findings are also consistent with AD participants experiencing higher levels of acute and chronic stress.

Keywords: cortisol, norepinephrine, Alzheimer’s disease, Mild Cognitive Impairment, cerebrospinal fluid, stress

INTRODUCTION

The relationships between the central nervous system (CNS) noradrenergic system and the hypothalamic pituitary adrenocortical (HPA) axis are complex, and the levels of at which these systems interact are multiple. Noradrenergic and adrenergic input to HPA axis activation can involve direct projections and indirect paths between brain structures and signaling through peripheral sympathetic input 1,2. One significant pathway includes brainstem projections thought to provide the substantial proportion of regulatory input to corticotropin releasing factor (CRF) producing neurons in the paraventricular nucleus (PVN) of the hypothalamus. These include noradrenergic projections, the larger proportion from the nucleus of the solitary tract (NTS) but also from the locus coeruleus (LC) 1, likely mediated through alpha-1 adrenoreceptors in the PVN 3.

Previous cerebrospinal fluid (CSF) studies suggest dysregulation of the HPA axis and the CNS noradrenergic system in Alzheimer’s disease (AD). Cross-sectional studies examining CSF cortisol have shown higher levels in older compared to younger participants, with even higher levels in older participants with AD 4–6. Neurodegeneration of the LC with substantial neuronal loss and neurofibrillary tangle formation occurs early in the course of AD 7,8. Compensatory upregulation of norepinephrine (NE) biosynthetic capacity in surviving LC neurons and upregulation of noradrenergic postsynaptic receptors in AD 9,10 may contribute to the pathophysiology of agitation and aggression in AD 11,12.

Given the role of CNS noradrenergic regulation of the HPA axis, we hypothesized that in normal controls there would be a positive relationship between CSF cortisol and CSF NE. We also hypothesized that this relationship would be altered in AD dementia and aMCI. Therefore, we examined the cross-sectional association between CSF cortisol and NE from samples in a large CSF repository. Our aims were to examine this relationship in cognitively normal controls and whether CSF cortisol and NE associations differed by presence of aMCI and AD dementia. We also added a sensitivity analysis combining AD dementia and aMCI in one group, based on the conceptualization that aMCI is the earliest clinical manifestation of AD.

METHODS

Study participants

We analyzed CSF samples and data from an existing CSF repository from an academic multicenter study. Study procedures pertaining to this repository and multicenter study have been previously described 13,14. The sites involved were the University of Washington, University of California, San Diego, Oregon Health and Science University, University of Pennsylvania, Indiana University, and University of California, Davis. The repository includes samples from participants aged 20 to 100 years with normal cognition, aMCI, and AD. (This repository also contains samples from participations with other neurodegenerative conditions including Dementia with Lewy bodies, Parkinson’s Disease, and Frontotemporal dementia, and traumatic brain injury).

For this analysis, we extracted data from the repository for participants who had normal cognition, aMCI, or AD dementia. We excluded participants taking medications that are known to affect CSF cortisol or NE concentrations, including corticosteroids, stimulants, serotonin-norepinephrine reuptake inhibitors, tricyclic antidepressants, monoamine oxidase inhibitors, and Parkinson disease medications (including carbidopa/levodopa and dopamine agonists).

Diagnoses for normal cognition, aMCI, and AD dementia, were based on the following criteria. Criteria for normal cognition included a Mini-Mental State Examination (MMSE) 15 score of 26 to 30 and Clinical Dementia Rating (CDR) Scale 16 score of 0, Logical Memory 17 score of 6 or greater (a test of delayed verbal recall) and no evidence or history of cognitive decline. Criteria for patients with aMCI included a CDR of 0.5 and a Logical Memory score of less than 6. Participants with AD dementia met NINCDS-ADRDA criteria for probable AD and had a Clinical Dementia Rating Scale of 1 or more.

The repository protocol required all participants be medically stable and have stable medications for 4 weeks prior to the lumbar puncture. Exclusionary criteria included current substance use disorders, a current major depressive episode, and chronic psychiatric conditions such as schizophrenia, bipolar disorder and PTSD.

All procedures were approved by the institutional review boards for participating institutions, and all participants provided written informed consent before study enrollment.

CSF collection

Sample collections occurred between 0900 and 1100 hours with subjects fasting. CSF was divided into sequential 0.5 ml aliquots into polypropylene tubes, frozen immediately on dry ice at the bedside, and then stored at −70 C° until assayed. Tubes assigned to the norepinephrine assay also contained the additive glutathione.

Norepinephrine and cortisol assays

For measurement of CSF NE concentration, CSF was extracted using the alumina extraction method optimized by Holmes et al 18. Extracted samples were separated by high-pressure liquid chromatography (HPLC) using a reverse phase C-18 column and measured by electrochemical detection (Coulechem II, ESA, Inc., Chelmsford, MA) with 3,4-dihydroxybenzylamine (DHBA) as an internal standard 14. The intra-assay coefficient of variation (CV) is 7.1% (based on measurements of the internal reference standard DHBA) and the inter-assay CV is 10.8% (based on measurement of pooled CSF samples repeated in each assay).

CSF cortisol was measured with ImmuChem Cortisol Coated Tube RIA Kits (ICN/MP Biomedical, Santa Ana, CA) following the commercial protocol modified for their use with CSF. Because of the low protein content of CSF cortisol relative to plasma, steroid-free Stripped Human Serum (ICN/MP Biomedical, Santa Ana, CA) was added to samples to bring their protein content roughly equivalent to the protein in the kit standards that are in a “human serum matrix.”

Statistical analysis

Linear regression was used to determine the association between CSF NE or CSF cortisol (the dependent variables) and age, and the association between the CSF variables and sex, education and body mass index (BMI) adjusting for age (i.e. with linear age as a model co-variate).

To determine if mean CSF cortisol or CSF NE levels differed by AD/aMCI vs. cognitively normal participants, a three diagnostic category variable (controls (i.e., cognitively normal), aMCI or AD) was added as an independent covariate to the above model (using 2 dummy variables). Pair-wise comparisons of AD, aMCI and controls were tested for significance using simultaneous inference which adjust for multiple comparisons within a model 19.

Linear regression was used to assess relationships between CSF cortisol (the dependent variable) and CSF NE, adjusting for age, sex, education, and BMI. Interaction terms for NE by diagnostic category or age were added to the regression models to assess whether associations between NE and cortisol differ by age or diagnostic category. Slopes of change in CSF cortisol by NE for each diagnostic category and predicted mean cortisol for each group at selected values of NE were estimated from this model. Hypothesis testing to determine if each slope was different from zero and the computation of 95% confidence intervals (CI) for predicted mean cortisol values across selected NE values and group were carried out using simultaneous inference as above.

Three sets of sensitivity analyses were also carried out: 1) regressions where age, as an independent covariate, was modeled as nonlinear using a restricted cubic spline; 2) regressions of cortisol or NE either excluding potential outliers or transforming the data to address possible violations of linear regression assumptions, and 3) regressions using two category variable for diagnostic category, combining AD and aMCI.

Type 1 error was set at p=0.05. Summary statistics in the text are reported as means ± standard errors (SE), [95% confidence intervals (CI)], unless where otherwise indicated. Analyses were carried out using R 3.2.3 20, with regressions carried out using the rms package 21, simultaneous inference statistics estimated using the multcomp package 19

RESULTS

Study participant characteristics

Study participant characteristics are summarized in Table 1. The study sample included 305 individuals with normal cognition, 22 individuals with MCI, and 94 individuals with AD. Study groups had generally similar characteristics with regards to sex, BMI and education.

Table 1.

Subject characteristics by diagnostic category: mean ± SD (range) or frequency (%)

Controls MCI (n=22) AD (n=94)
All (n=305) Age < 50 (n=97) Age ≥ 50 (n=208)
Age 56.1 ± 18.3 (20–100) 33.6 ± 9.4 (20–49) 66.7 ± 10.0 (50–100) 72.3 ± 8.2 (56–84) 71.2 ± 9.3 (52–88)
Male 139 (46%) 53 (55%) 86 (41%) 13 (59%) 49 (52%)
Education 16.1 ± 2.7 (10–27) 16.4 ± 2.7 (11–27) 16.0 ± 2.7 (10–25) 16.5 ± 2.7 (12–21) 15.8 ± 3.1 (6–25)*
Body Mass Index 26.0 ± 4.0 (18–41) 25.7 ± 4.3 (19–41) 26.1 ± 3.8 (18–40) 24.7 ± 3.4 (18–33)* 25.9 ± 3.4 (18–36)†
Open in a new tab
*

unavailable for 1 participant

†

unavailable for 3 participants

To characterize this sample, we examined these participant characteristics in relation to CSF cortisol and CSF NE. Linear regression analysis showed both CSF cortisol and CSF NE increased significantly with subject age (p<0.0001) by a mean ± SE of 0.7 ± 0.1 ng/ml (95% CI [0.5, 0.9] and 8.4 ± 2.1 pg/ml (95%CI [4.3, 12.4]) per 10 years of participant age increase, respectively. Adjusting for age, mean CSF cortisol was slightly higher for men than women by 0.8 ± 0.3 ng/ml (95%CI [0.2, 1.3], p=0.012) but mean CSF NE did not differ significantly by sex (p=0.28). Age adjusted mean cortisol was not associated with education, but age adjusted mean NE decreased by 2.8 ± 1.3 pg/ml per increase in one year of education ([0.2, 5.4], p=0.033). (This association between CSF NE and education was no longer significant when diagnostic category was added to the regression model as a covariate, p=0.22.) No association for either CSF cortisol or NE was found with BMI (p>0.11).

CSF Cortisol and NE concentrations by diagnostic category

Figure 1 presents scatterplots of CSF cortisol (A) and CSF NE (B) by age with separate symbols and slopes by diagnostic category. There were significant differences in mean cortisol concentrations among controls, aMCI and AD dementia groups adjusting for age, sex, education and BMI. Compared to controls, CSF cortisol was higher by 1.6 ± 0.4 ng/ml (95% CI [0.7, 2.5], p<0.001) in the AD dementia group. Mean CSF cortisol for the aMCI group was higher than controls by 0.8 ± 0.7 ng/ml (95% CI [−0.8, 2.5]) and lower than the AD dementia group by the same amount (0.8 ± 0.7 ng/ml, 95% CI [−0.9, 2.5]), but neither difference was significant (p>0.44). There was no significant difference in mean CSF NE among the control, aMCI and AD dementia groups (p>0.60).

Figure 1.

Figure 1

Figure 1

Open in a new tab

Scatterplot of CSF cortisol (A) and norepinephrine (B) concentrations by age and diagnostic category. Slopes are estimated from linear regression of cortisol on age and diagnostic category.

Association between CSF cortisol and NE

Linear regression, adjusting for age, sex, education, BMI and diagnostic category showed no significant association between cortisol (the dependent variable) and NE overall (regression coefficient ± SE per 100 pg/ml increase in NE [95% CI], 0.3 ± 0.2 [−0.1, 0.7], p=0.11). There was no evidence that the association between cortisol and NE differed by age (linear age by NE interaction, p=0.45).

However, when testing for the effect of diagnostic category on the cortisol versus NE association, a significant interaction was found (interaction term: NE by category; p = 0.021). In cognitively normal controls (Figure 2), no association between CSF cortisol and NE was found (regression coefficient ± SE per increase in 100 pg/ml [95%CI]: 0.0 ± 0.2, 95%CI [−0.6, 0.5], p=1.0). In aMCI and AD dementia participants, higher cortisol levels were associated with higher NE levels (regression coefficient [95%CI]: aMCI, 1.4 ± 0.7 [−0.3, 3.2], p=0.14; and AD 1.1 ± 0.4 [0.1, 2.2], p=0.032).

Figure 2.

Figure 2

Open in a new tab

Scatterplot of CSF cortisol by CSF norepinephrine and diagnostic category. Slopes are estimated from linear regression of CSF cortisol on CSF NE, diagnostic category, age, sex, education and BMI, and averaged across age, sex, education and BMI.

There is a significant positive correlation between CSF cortisol and CSF NE in AD/aMCI.

The linear regression model above was used to calculate mean cortisol levels per increased NE levels (Table 2). In AD dementia participants, an increase in CSF NE from 100 to 300 pg/ml was associated with an increase in cortisol from 13.7 to 16.0 ng/ml. In aMCI participants, the same increase in CSF NE was associated with an increase in cortisol from 12.5 to 15.4 ng/ml. In contrast, cognitively normal participants’ mean cortisol level remained constant with increasing NE concentrations.

Table 2.

Mean CSF cortisol level in ng/ml (95% CI) per NE concentration, predicted from regression analysis of CSF cortisol on CSF NE by diagnostic category adjusting for age, sex, education and BMI*

NE (pg/ml) Cortisol (ng/ml)
Controls Age ≥ 50 (n=305) aMCI (n=22) AD (n=94)
100 12.9 (12.1, 13.7) 12.5 (10.0, 15.1) 13.7 (12.3, 15.0)
200 12.9 (12.2, 13.6) 14.0 (12.1, 15.9) 14.8 (13.8, 15.9)
300 12.9 (11.9, 13.9) 15.4 (12.5, 18.3) 16.0 (14.2, 17.8)
Open in a new tab
*

From linear regression of cortisol concentration (the dependent variable) on NE concentration by category interaction with co-variates age, sex, education and BMI.

Sensitivity Analyses

Sensitivity analysis using age as a non-linear variable demonstrated similar results as the primary findings above, further confirming age does not explain the relationship between CSF cortisol and NE found in AD dementia patients.

Another sensitivity analysis using regression models using log(NE) instead of NE as the dependent variable to address right skewness in the NE variable also yielded similar findings. When a participant with cortisol > 30 ng/ml was excluded, differences in mean cortisol between the controls and aMCI and AD groups remained significant. However, the exclusion of this participant attenuated the difference in the cortisol vs. NE slopes by diagnostic category, (NE by diagnostic category interaction p=0.079).

When diagnostic category was modeled with two categories (AD dementia and aMCI combined vs. controls), diagnostic category differences in mean cortisol and in the cortisol vs.NE slopes were strengthened. Mean CSF cortisol for the AD/aMCI group was higher than controls by 1.3 ± 0.4 ng/ml (p<0.001). Mean CSF cortisol increased by 1.2 ± 0.3 ng/ml per increase in in CSF NE of 100 pg/ml, (p=0.002, NE by category interaction; p = 0.007). These associations remained significant with the exclusion of the high cortisol outlier.

DISCUSSION

This analysis is the first, to our knowledge, to examine the association between CSF cortisol and CSF NE in cognitively normal controls, participants with aMCI, and participants with AD dementia. We show that contrary to our hypothesis, there was not a significant positive association between CSF cortisol and CSF NE in control participants. In contrast, we did find a positive association between CSF cortisol and CSF NE in AD dementia participants, a finding strengthened when the aMCI and AD dementia participants were combined in the analysis.

Although stimulatory regulation of the HPA access by central noradrenergic system input is well established, this mechanism is one among multiple neural and endocrine regulators of the HPA axis. These range from direct projections and indirect paths between brain structures, signaling through peripheral sympathetic input, and glucocorticoid negative feedback 2,22. Given the complexity of the relationships between these two stress systems, it is not surprising that we did not find the association between CSF cortisol and CSF NE in controls we had anticipated. These multiple regulatory mechanisms likely obscure finding a significant relationship between CSF NE and CSF cortisol, particularly as the latter hormone is derived from the periphery and must cross the blood brain barrier.

There are several phenomena in AD that could increase the likelihood of detecting a significant relationship between NE and cortisol, one of which is the blood brain barrier. Normally, access of cortisol to the central nervous system is restricted by tight junctions and by efflux transporters, the primary of which is P-glycoprotein 23. In AD, P-glycoprotein has been shown to be expressed at lower levels and have decreased function; tight junctions are loosened, and focal necrosis also increase permeability of the blood brain barrier 24–26. Together, these changes provide greater access to the brain and CSF but decrease the efficiency of cortisol efflux, likely resulting to higher CSF cortisol in those with AD. This process in AD would favor detection of a significant association between NE and cortisol in the CSF compartment.

Another possibility is raised by demonstration in preclinical studies that chronic stress enhances PVN excitability, through multiple mechanisms including reduction of feedback inhibition and increased input from afferent connections 22. One such mechanism involves the noradrenergic system, as chronic stress enhances noradrenergic innervation 1. NE has been suggested to play a key role in chronic stress enhancement of HPA axis excitability 27. It is likely that the impaired memory, executive function, and reasoning ability of AD patients make everyday challenges more stressful. Such elevations in chronic stress could give the noradrenergic component of HPA axis stimulatory regulation a more dominant role, and thus make a relationship between CSF NE and CSF cortisol more detectable in this study.

Finally, there are known alterations in the noradrenergic system in AD 28. Marked degeneration of the LC is one of the earliest changes that occur in the course of AD 7,29–31, and more recently, accumulation of tau in the LC has been found to occur early in AD as well 8. Post-mortem studies indicate increased biosynthetic capacity in remaining LC neurons 9 and upregulation of numbers of post-synaptic alpha-1 adrenoreceptors present in AD patients compared to controls in the hippocampus and frontal cortex 10,32. We have previously proposed that this apparent compensatory upregulation contributes to symptoms such as agitation and aggression in AD 12,33. How these changes in AD may influence HPA axis activation is less clear. In animal studies, LC projections to the PVN are present, although the substantial proportion noradrenergic input comes from the NTS 22. There also may not be detectable differences in NE measured in the hypothalami of AD subjects compared to controls 34. However, given the early and substantial changes in the noradrenergic system in AD, as well as our previous work suggesting compensatory upregulation of this system, an alteration in the relationship between CSF NE and cortisol compared to controls is a consistent finding, notwithstanding uncertainty over the exact mechanism by which this association occurs.

One separate area of significance of this study is that it replicates several other previous CSF studies. We show that higher cortisol concentrations occur in older versus younger adults and in AD patients versus controls 4,5,35,36. We had previously demonstrated in this sample that CSF NE increases with age; the analysis described in this report also replicates smaller studies revealing similar CSF NE concentrations in controls and AD participants 33,37.

There are several limitations to this study. It is a cross-sectional examination of stress systems known to fluctuate depending on the time of day and environment. To minimize this variability, the study design standardized lumbar puncture procedures including time of day, and included a large number of participants in relation to most CSF studies, which better accommodates variability. This study also may not have taken into account potential medication effects, although we excluded participants taking medications known to CSF NE or cortisol levels (e.g., certain antidepressants, Parkinson’s disease medications). Another limitation is that this analysis presents associations only and does not elucidate mechanisms. However, associations seen in this study provide rationale for future work investigating underlying pathophysiology.

In summary, this study reports a positive association between CSF cortisol and CSF NE levels in AD patients but not in cognitively normal controls. This association likely reflects alterations in the HPA axis and noradrenergic systems in the CNS of AD patients. Given the complexity of the relationships between these two systems, we propose several possible mechanisms ranging from how cortisol is cleared from the CSF to neurobiological changes in chronic stress and AD. The underlying mechanism of this association may not be clear from this study alone, but findings support future studies examining the relationships between these two stress systems in AD.

Acknowledgments

This work was supported by the VA CSR&D Career Development Award Program (Project ID: 3125); the Mental Illness Research, Education, and Clinical Center (MIRECC) VISN 20 NW Network; Geriatric Research, Education and Clinical Center, VA Puget Sound Health Care System; and NIH/NIA grants AG005136 and AG008017.

Footnotes

This work was presented in part at the XV Annual Meeting of the International College of Geriatric Psychoneuropharmacology, Palo Alto, CA, October 2015.

The authors have no conflicts of interest to disclose.

References

  • 1.Flak JN, Myers B, Solomon MB, McKlveen JM, Krause EG, Herman JP. Role of paraventricular nucleus-projecting norepinephrine/epinephrine neurons in acute and chronic stress. The European journal of neuroscience. 2014;39(11):1903–1911. doi: 10.1111/ejn.12587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Herman JP, Figueiredo H, Mueller NK, et al. Central mechanisms of stress integration: hierarchical circuitry controlling hypothalamo–pituitary–adrenocortical responsiveness. Frontiers in Neuroendocrinology. 2003;24(3):151–180. doi: 10.1016/j.yfrne.2003.07.001. [DOI] [PubMed] [Google Scholar]
  • 3.Plotsky PM, Cunningham ET, Jr, Widmaier EP. Catecholaminergic modulation of corticotropin-releasing factor and adrenocorticotropin secretion. Endocrine reviews. 1989;10(4):437–458. doi: 10.1210/edrv-10-4-437. [DOI] [PubMed] [Google Scholar]
  • 4.Peskind ER, Wilkinson CW, Petrie EC, Schellenberg GD, Raskind MA. Increased CSF cortisol in AD is a function of APOE genotype. Neurology. 2001;56(8):1094–1098. doi: 10.1212/wnl.56.8.1094. [DOI] [PubMed] [Google Scholar]
  • 5.Popp J, Schaper K, Kölsch H, et al. CSF cortisol in Alzheimer’s disease and mild cognitive impairment. Neurobiology of Aging. 2009;30(3):498–500. doi: 10.1016/j.neurobiolaging.2007.07.007. [DOI] [PubMed] [Google Scholar]
  • 6.Wilkinson CW. Correlations Among Simultaneous Basal Measurements of Human Cerebrospinal Fluid, Saliva, and Total and Free Plasma Cortisol as a Function of Age and Alzheimer’s Disease/Mild Cognitive Impairment. Gerontologist; Poster session presented at the Gerontological Society of America’s 66th Annual Scientific Meeting; New Orleans, LA. 2013. p. 278. [Google Scholar]
  • 7.Bondareff W, Mountjoy CQ, Roth M. Selective loss of neurones of origin of adrenergic projection to cerebral cortex (nucleus locus coeruleus) in senile dementia. Lancet. 1981;1(8223):783–784. doi: 10.1016/s0140-6736(81)92657-x. [DOI] [PubMed] [Google Scholar]
  • 8.Braak H, Del Tredici K. The pathological process underlying Alzheimer’s disease in individuals under thirty. Acta Neuropathol. 2011;121(2):171–181. doi: 10.1007/s00401-010-0789-4. [DOI] [PubMed] [Google Scholar]
  • 9.Szot P, White SS, Greenup JL, Leverenz JB, Peskind ER, Raskind MA. Compensatory changes in the noradrenergic nervous system in the locus ceruleus and hippocampus of postmortem subjects with Alzheimer’s disease and dementia with Lewy bodies. The Journal of neuroscience : the official journal of the Society for Neuroscience. 2006;26(2):467–478. doi: 10.1523/JNEUROSCI.4265-05.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Szot P, White SS, Greenup JL, Leverenz JB, Peskind ER, Raskind MA. Changes in adrenoreceptors in the prefrontal cortex of subjects with dementia: evidence of compensatory changes. Neuroscience. 2007;146(1):471–480. doi: 10.1016/j.neuroscience.2007.01.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Peskind ER, Elrod R, Dobie DJ, et al. Cerebrospinal fluid epinephrine in Alzheimer’s disease and normal aging. Neuropsychopharmacology. 1998;19(6):465–471. doi: 10.1016/S0893-133X(98)00054-2. [DOI] [PubMed] [Google Scholar]
  • 12.Wang LY, Shofer JB, Rohde K, et al. Prazosin for the treatment of behavioral symptoms in patients with Alzheimer disease with agitation and aggression. The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry. 2009;17(9):744–751. doi: 10.1097/JGP.0b013e3181ab8c61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Peskind ER, Li G, Shofer JB, et al. INfluence of lifestyle modifications on age-related free radical injury to brain. JAMA Neurology. 2014;71(9):1150–1154. doi: 10.1001/jamaneurol.2014.1428. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Wang LY, Murphy RR, Hanscom B, et al. Cerebrospinal fluid norepinephrine and cognition in subjects across the adult age span. Neurobiol Aging. 2013;34(10):2287–2292. doi: 10.1016/j.neurobiolaging.2013.04.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189–198. doi: 10.1016/0022-3956(75)90026-6. [DOI] [PubMed] [Google Scholar]
  • 16.Morris JC. Clinical dementia rating: a reliable and valid diagnostic and staging measure for dementia of the Alzheimer type. Int Psychogeriatr. 1997;9(Suppl 1):173–176. doi: 10.1017/s1041610297004870. discussion 177–178. [DOI] [PubMed] [Google Scholar]
  • 17.Wechsler D, Stone C. Manual: Wechsler memory Scale. New York: Psychological Corporation; 1973. [Google Scholar]
  • 18.Holmes C, Eisenhofer G, Goldstein DS. Improved assay for plasma dihydroxyphenylacetic acid and other catechols using high-performance liquid chromatography with electrochemical detection. J Chromatogr B Biomed Appl. 1994;653(2):131–138. doi: 10.1016/0378-4347(93)e0430-x. [DOI] [PubMed] [Google Scholar]
  • 19.Hothorn T, Bretz F, Westfall P. Simultaneous inference in general parametric models. Biom J. 2008;50(3):346–363. doi: 10.1002/bimj.200810425. [DOI] [PubMed] [Google Scholar]
  • 20.R Core Team. A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2015. [Google Scholar]
  • 21.Harrell FE. rms: Regression Modeling Strategies. R package version 4.3-0. 2015 https://CRAN.R-project.org/package=rms.
  • 22.Herman JP, Tasker JG. Paraventricular Hypothalamic Mechanisms of Chronic Stress Adaptation. Front Endocrinol (Lausanne) 2016;7:137. doi: 10.3389/fendo.2016.00137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Uhr M, Holsboer F, Muller MB. Penetration of endogenous steroid hormones corticosterone, cortisol, aldosterone and progesterone into the brain is enhanced in mice deficient for both mdr1a and mdr1b P-glycoproteins. J Neuroendocrinol. 2002;14(9):753–759. doi: 10.1046/j.1365-2826.2002.00836.x. [DOI] [PubMed] [Google Scholar]
  • 24.Bartels AL. Blood-brain barrier P-glycoprotein function in neurodegenerative disease. Curr Pharm Des. 2011;17(26):2771–2777. doi: 10.2174/138161211797440122. [DOI] [PubMed] [Google Scholar]
  • 25.Chiu C, Miller MC, Monahan R, Osgood DP, Stopa EG, Silverberg GD. P-glycoprotein expression and amyloid accumulation in human aging and Alzheimer’s disease: preliminary observations. Neurobiol Aging. 2015;36(9):2475–2482. doi: 10.1016/j.neurobiolaging.2015.05.020. [DOI] [PubMed] [Google Scholar]
  • 26.Marques F, Sousa JC, Sousa N, Palha JA. Blood-brain-barriers in aging and in Alzheimer’s disease. Mol Neurodegener. 2013;8:38. doi: 10.1186/1750-1326-8-38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Flak JN, Ostrander MM, Tasker JG, Herman JP. Chronic stress-induced neurotransmitter plasticity in the PVN. The Journal of comparative neurology. 2009;517(2):156–165. doi: 10.1002/cne.22142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Gannon M, Che P, Chen Y, Jiao K, Roberson ED, Wang Q. Noradrenergic dysfunction in Alzheimer’s disease. Frontiers in neuroscience. 2015:9. doi: 10.3389/fnins.2015.00220. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Chan-Palay V, Asan E. Alterations in catecholamine neurons of the locus coeruleus in senile dementia of the Alzheimer type and in Parkinson’s disease with and without dementia and depression. The Journal of comparative neurology. 1989;287(3):373–392. doi: 10.1002/cne.902870308. [DOI] [PubMed] [Google Scholar]
  • 30.German DC, Manaye KF, White CL, 3rd, et al. Disease-specific patterns of locus coeruleus cell loss. Annals of neurology. 1992;32(5):667–676. doi: 10.1002/ana.410320510. [DOI] [PubMed] [Google Scholar]
  • 31.Zarow C, Lyness SA, Mortimer JA, Chui HC. Neuronal loss is greater in the locus coeruleus than nucleus basalis and substantia nigra in Alzheimer and Parkinson diseases. Archives of neurology. 2003;60(3):337–341. doi: 10.1001/archneur.60.3.337. [DOI] [PubMed] [Google Scholar]
  • 32.Sharp SI, Ballard CG, Chen CP, Francis PT. Aggressive behavior and neuroleptic medication are associated with increased number of alpha1-adrenoceptors in patients with Alzheimer disease. The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry. 2007;15(5):435–437. doi: 10.1097/01.JGP.0000237065.78966.1b. [DOI] [PubMed] [Google Scholar]
  • 33.Elrod R, Peskind ER, DiGiacomo L, Brodkin KI, Veith RC, Raskind MA. Effects of Alzheimer’s disease severity on cerebrospinal fluid norepinephrine concentration. The American journal of psychiatry. 1997;154(1):25–30. doi: 10.1176/ajp.154.1.25. [DOI] [PubMed] [Google Scholar]
  • 34.Sparks DL, DeKosky ST, Markesbery WR. Alzheimer’s disease. Aminergic-cholinergic alterations in hypothalamus. Archives of neurology. 1988;45(9):994–999. doi: 10.1001/archneur.1988.00520330084014. [DOI] [PubMed] [Google Scholar]
  • 35.Murakami K, Nakagawa T, Shozu M, Uchide K, Koike K, Inoue M. Changes with aging of steroidal levels in the cerebrospinal fluid of women. Maturitas. 1999;33(1):71–80. doi: 10.1016/s0378-5122(99)00040-7. [DOI] [PubMed] [Google Scholar]
  • 36.Guazzo EP, Kirkpatrick PJ, Goodyer IM, Shiers HM, Herbert J. Cortisol, dehydroepiandrosterone (DHEA), and DHEA sulfate in the cerebrospinal fluid of man: relation to blood levels and the effects of age. The Journal of clinical endocrinology and metabolism. 1996;81(11):3951–3960. doi: 10.1210/jcem.81.11.8923843. [DOI] [PubMed] [Google Scholar]
  • 37.Raskind MA, Peskind ER, Holmes C, Goldstein DS. Patterns of cerebrospinal fluid catechols support increased central noradrenergic responsiveness in aging and Alzheimer’s disease. Biological psychiatry. 1999;46(6):756–765. doi: 10.1016/s0006-3223(99)00008-6. [DOI] [PubMed] [Google Scholar]

RESOURCES