Cerebral glucose metabolism abnormalities in patients with major depressive symptoms in pre-dialytic chronic kidney disease: Statistical parametric mapping analysis of F-18-FDG PET, a preliminary study

Authors


Abstract

Aims:  The aim of the present study was to investigate the relationship between depressive symptoms and cerebral glucose metabolism in pre-dialytic chronic kidney disease (PDCKD) patients.

Methods:  Twenty-one patients with stage 5 CKD and 21 healthy volunteers underwent depressive mood assessment and statistical parametric mapping (SPM) using F-18-fluorodeoxyglucose (FDG) positron emission tomography (PET).

Results:  Several voxel clusters of significantly decreased cerebral glucose metabolism were found in PDCKD patients. The largest cluster was left prefrontal cortex (Brodmann area [BA] 9). The second largest cluster was also left prefrontal cortex (BA 9). The third largest clusters were right prefrontal cortex (BA 10) and right basolateral prefrontal cortex (BA 46). Other brain areas also showed decreased cerebral glucose metabolism including left anterior cingulate gyrus (BA 32), left premotor cortex (BA 6), left transverse temporal gyrus (BA 41), left superior temporal gyrus (BA 42), right basolateral prefrontal cortex (BA 44), right inferior parietal lobule (BA 39), left middle temporal gyrus (BA 19), and left angular gyrus (BA 39). Hypermetabolized brain areas, however, were not found in PDCKD patients compared to normal controls. For the right orbitofrontal cortex there was a negative correlation of cerebral glucose metabolism with Hamilton Depression Rating Scale (HDRS) in PDCKD patients (BA 11).

Conclusion:  PDCKD patients with depressive symptoms had decreased cerebral glucose metabolism in several brain areas. For the right orbitofrontal cortex there was a negative correlation with HDRS in PDCKD patients. The present findings provide functional neuroimaging support for abnormal cerebral glucose metabolism in PDCKD patients with depressive symptoms.

CHRONIC DISEASES SUCH as heart disease, cancer, and diabetes are leading causes of disability and death in the USA, and every year they claim the lives of more than 1.7 million Americans.1 Chronic diseases cause major limitations in daily living for more than one of every 10 Americans, or 25 million people.1

Depression is an important global public health problem due to both its relatively high lifetime prevalence and the significant disability that it causes. In 2002, depression accounted for 4.5% of the worldwide total burden of disease (in terms of disability-adjusted life years).2 It is also responsible for the greatest proportion of burden attributable to non-fatal health outcomes, accounting for almost 12% of total years lived with disability worldwide.

Depression has been identified as the most common psychiatric illness in patients with end-stage renal disease (ESRD), but its prevalence has varied widely.3,4 In chronic kidney disease (CKD), the neuropsychiatric effects range from subtle, non-specific changes to gross, distinct abnormalities according to stage and chronicity of kidney disease and condition of patients. Among patients with ESRD, depression and cognitive impairment are the most common causes of neuropsychiatric illness.5 In particular, depression is considered as an important factor in patient survival.6,7

A recent study has shown that the 12-month prevalence of major depression in individuals with chronic medical conditions ranges from 7.9% to 17% and the age/sex-adjusted odds of major depression range from 1.96 to 3.56.8 Among the various chronic medical conditions, individuals with ESRD had the highest 12-month prevalence and odds of major depression.8

Functional neuroimaging techniques including single photon emission tomography (SPECT) and positron emission tomography (PET) have been used to investigate the influence and mechanism of various mental factors such as depression, adjustment disorders, and anxiety. Recent advances in neuroimaging techniques have led to some interesting data concerning alteration in brain structure and functions in several chronic illnesses with psychiatric symptoms.9–11

In some studies of depression secondary to chronic medical diseases or neurological diseases, particular patterns of cerebral blood flow and metabolism were implicated in anterior cingulate, paralimbic areas, parietal, and temporal areas.12–14 The major findings of those studies were that in secondary depression, orbitofrontal metabolism is reduced or normal; and dorsolateral prefrontal blood flow and metabolism are apparently normal.

The aim of the current study was to investigate the relationship between depressive symptoms and cerebral glucose metabolism in pre-dialytic chronic kidney disease (PDCKD) patients using statistical parametric mapping (SPM) of brain PET, and to examine the usefulness of brain PET for early detection of depression in PDCKD.

METHODS

Patients

Twenty-one patients with stage 5 PDCKD (established kidney failure, glomerular filtration rate [GFR] <15 mL/min per 1.73 m2, or permanent renal replacement therapy) who attended the Department of Internal Medicine at Pusan National University Hospital, underwent depressive mood assessment and brain PET. Twenty-one healthy volunteers were enrolled for comparison. Healthy volunteers were screened for neuropsychiatric and medical illness. Of the PDCKD patients, nine were male and 12 were female. The causes of CKD were as follows: hypertension (n = 8), diabetes mellitus (n = 10), and glomerulonephritis (n = 3). Patients were excluded who had previous psychological disease, neurologic disease, previous stroke or transient ischemic attack, cognitive impairment, head trauma, atypical headache, substance abuse, alcoholism, malnutrition, and electrolyte imbalance such as hyponatremia, hypernatremia, hypocalcemia, or hypercalcemia. Depressive mood was assessed on Beck Depression Inventory (BDI) and psychiatric interview using Hamilton Depression Rating Scale (HDRS).3,15 Each patient was assessed for depressive mood on the enrollment date and the psychiatrist completing the HDRS was blinded to the medical status of the patient and healthy volunteers. All the PDCKD patients were in an anti-depressant drug-naïve state at the time of enrollment. Ethics approval was obtained from the Ethics committee of Pusan National University Hospital. All CKD patients and healthy controls voluntarily agreed to participate in the study and written informed consent was obtained from the participants.

Clinical parameters

We measured blood urea nitrogen, creatinine, and GFR according to the Modification of Diet in Renal Disease (MDRD) study as clinical parameters.16

Neuropsychiatric assessments

Before brain PET imaging, psychiatrists performed an interview including Structural Clinical Interview for DSM-IV-Axis I Disorder to determine whether the PDCKD patients had a depressive symptom. The HDRS is a 17-item scale that evaluates depressed mood, vegetative and cognitive symptoms of depression, and comorbid anxiety symptoms. Also, HDRS was used to measure depression severity. The total score ranges from 0 to 53 as follows: normal, 0–6; mild depression, 7–17; moderate depression, 18–24; and severe depression, 25–53. The BDI is a 21-item scale that has been validated and used commonly among patients with ESRD. The BDI has been shown to be a sensitive screening tool for depression as well as a clinically useful scale for measuring severity of depression.

F-18-fluorodeoxyglucose (FDG) brain PET

Brain PET scans of a single frame of 15 min were acquired starting 60 min after i.v. injection of 370 MBq (10 mCi) F-18-fluorodeoxyglucose (FDG) using a Gemini PET/computed tomography (PET/CT) scanner (Philips, Milpitas, CA, USA). The subjects were scanned in a resting condition with their eyes closed and ears unplugged, comfortably lying in a darkened and quiet room. Subjects fasted for at least 6 h before PET imaging. PET images were reconstructed using 3-D Row Action Maximum Likelihood Algorithm (RAMLA) (two repetition, 0.006 relaxation parameter) and displayed in a 128 × 128 matrix (pixel size, 2 × 2 mm, with a slice thickness of 2 mm). Attenuation correction was performed with a uniform attenuation coefficient (µ = 0.096/cm). In-plane and axial resolution of the scanner were 4.2 and 5.6 mm full width at half maximum (FWHM), respectively.

Image analysis

Spatial pre-processing and statistical analysis were performed using the SPM2 implemented in Matlab 7.2 (MathWorks, Natick, MA, USA). All the reconstructed F-18-FDG brain PET images were spatially normalized into Montreal Neurological Institute (MNI, McGill University, Montreal, Quebec, Canada) standard templates using an affine transformation (12 parameters for rigid transformations) and a non-linear transformation, then smoothed with an FWHM 8-mm Gaussian kernel to increase the signal-to-noise ratio and to account for subtle variations in anatomic structures. To remove the effects of the difference in the overall counts, the voxel counts were normalized to the mean voxel count of the gray matter in each PET image using proportional scaling. Images of the PDCKD patients were compared with those from normal controls in a voxel-wise manner using SPM2 for between-group analysis (P < 0.001, uncorrected; extent threshold, k = 100). The clusters that passed this threshold were considered significant at P < 0.05 corrected for multiple comparisons using false discovery rate (FDR). The Talairach brain coordinates were estimated from a non-linear transformation from MNI space to Talairach space (Talairach Daemon Client, Version 1.1, Research Imaging Center, University of Texas Health Science Center at San Antonio, TX, USA). For the visualization of the t-score statistics (SPM{t} map), significant voxels were projected onto the 3-D rendered brain or a standard high-resolution magnetic resonance (MR) image template provided by SPM2, thereby allowing anatomic identification. In addition, linear correlations of the regional brain glucose metabolism with HDRS and BDI scores were investigated using single subject–covariates only analysis and simple regression analysis, which are statistical models in SPM2 based on the general linear model.

Statistical analysis

All continuous variables were expressed as mean (minimum–maximum). The comparison of groups was performed using Mann-Whitney test, Kruskal–Wallis test, and Fisher's exact test (MedCalc version 8.1.0.0, Mariakerke, Belgium), as appropriate.

RESULTS

Patient characteristics

Table 1 lists the characteristics of the CKD patients and normal controls. The causes of CKD were as follows: hypertension (n = 8), diabetes mellitus (n = 10), and glomerulonephritis (n = 3).

Table 1. Patient characteristics
CharacteristicsCKDControls P
  1. BDI, Beck Depression Inventory; CGN, chronic glomerulonephritis; DM, diabetes mellitus; GFR, glomerular filtration rate; HDRS, Hamilton Depression Rating Scale; HT, hypertension; PDCKD, pre-dialytic chronic kidney disease.

Age (years)51 (32–66)21 (25–80)NS
Gender (M : F)9:1210:11NS
PDCKD cause   
 HT8 
 DM10 
 CGN3 
GFR (mL/min/1.73 m2)6.7 (4.1–11.6)112.6 (88.9–143.8)<0.01
BDI20.1 (2–48)4.1 (2–7)<0.01
HDRS18.5 (5–32)3.4 (2–5)<0.01

Regional cerebral glucose metabolism abnormalities

As shown in Table 2 and Fig. 1, several voxel clusters of significantly decreased cerebral glucose metabolism were found in PDCKD patients. The largest clusters were in the left prefrontal cortex (Brodmann area [BA] 9; voxel number 1750, peak Z value = 5.01). The second largest cluster was in the left prefrontal cortex (BA 9; voxel number 756, peak Z value = 5.71). The third largest clusters were in the right prefrontal cortex (BA 10; voxel number 653, peak Z value = 5.63) and right basolateral prefrontal cortex (BA 46; voxel number 653, peak Z value = 5.32). Other brain areas also showed decreased cerebral glucose metabolism including the left anterior cingulate gyrus (BA 32; voxel number 562, peak Z value = 5.26), left premotor cortex (BA 6; voxel number 459, peak Z value = 6.86), left transverse temporal gyrus (BA 41; voxel number 328, peak Z value = 5.57), left superior temporal gyrus (BA 42; voxel number 328, peak Z value = 4.71), right basolateral prefrontal cortex (BA 44; voxel number 294, peak Z value = 5.16), right inferior parietal lobule (BA 39; voxel number 165, peak Z value = 4.96), left middle temporal gyrus (BA 19; voxel number 106, peak Z value = 4.97), and left angular gyrus (BA 39; voxel number 106, peak Z value = 4.91). Hypermetabolized brain areas, however, were not found in PDCKD patients compared to normal controls.

Table 2. Abnormally decreased cerebral glucose metabolism in PDCKD
ClustersHemisphereCorrected PPeak ZTalairach coordinateStructureBA
XYZ
  1. BA, Brodmann area; PDCKD, pre-dialytic chronic kidney disease.

1750Left0.0095.01−382838Prefrontal cortex9
756Left0.0005.71−50438Prefrontal cortex9
653Right0.0015.634852−8Prefrontal cortex10
Right0.0025.32464018Basolateral prefrontal cortex46
562Left0.0035.26−44216Anterior cingulate32
459Left0.0006.86−26−1254Premotor cortex6
328Left0.0015.57−50−2610Transverse temporal gyrus41
Left0.0344.71−66−3418Superior temporal gyrus42
294Right0.0055.16621218Basolateral prefrontal cortex44
165Right0.0124.9648−7038Inferior parietal lobule39
106Left0.0124.97−50−8018Middle temporal gyrus19
Left0.0154.91−46−7632Angular gyrus39
Figure 1.

Rendered images of hypometabolic regions in patients with pre-dialytic chronic kidney disease compared with healthy controls.

As shown in Table 3 and Fig. 2, the right orbitofrontal cortex had a negative correlation with HDRS in PDCKD patients (BA 11; voxel number 133, peak Z value = 4.96).

Table 3. Negatively correlated cerebral glucose metabolism with HDRS in PDCKD
ClustersHemispherecorrected PPeak ZTalairach coordinateStructureBA
XYZ
  1. BA, Brodmann area; HDRS, Hamilton Depression Rating Scale; PDCKD, pre-dialytic chronic kidney disease.

133Right0.0214.961452−32Orbitofrontal cortex11
Figure 2.

Correlation between cerebral glucose metabolism and Hamilton Depression Rating Scale in pre-dialytic chronic kidney disease patients. (d) Right orbitofrontal cortex (Brodmann area 11). (▪) brain response at orbitofrontal cortex; (●) regression line in SPM results.

DISCUSSION

The current study first investigated the presence of regional cerebral glucose metabolism abnormalities in PDCKD patients who developed depressive symptoms compared to normal controls. In the present study PDCKD patients had similar pattern of regional cerebral glucose abnormalities that found in patients with major depression. Also, additional abnormal cerebral glucose metabolism abnormalities were seen in temporal areas, which is found in patients with depression secondary to chronic illnesses and neurologic disorders.

Previously, depression has been reported to be associated with reduced brain perfusion, but conflicting results, showing both increases and discordance in the location of reduced regional cerebral blood flow (rCBF), have also been reported.17–20 Perico et al. reported that severity of depressive mood and rCBF was inversely correlated in the left amygdala, lentiform nucleus, and parahippocampal gyrus, and positively correlated with right postero-lateral parietal cortex.21 Grasby suggested that regional deficits of functional activity have been consistently detected in the brains of subjects with affective symptoms, particularly in prefrontal areas.22 Videlbech reported that there is evidence that patients with major depression have reduced blood flow and metabolism in the prefrontal cortex, anterior cingulate gyrus, and basal ganglia.18

As well as preferential sites of regional cerebral glucose metabolism abnormalities in patients of major depression, the present study also found regional cerebral glucose metabolism abnormalities in the temporal cortex, inferior parietal lobule, and angular gyrus in PDCKD patients with depressive symptoms. Similar to the current study, in depressive pseudodementia patients, brain SPECT showed decreased regional cerebral blood flow in temporo-parietal region. Also, patients with dementia of Alzheimer's disease showed decreased cerebral blood flow in similar brain areas.14

The finding that the comorbidity between depressive symptom and PDCKD is associated with reduced regional cerebral glucose metabolism in the prefrontal cortex, basolateral prefrontal cortex, and anterior cingulate cortex, which are the preferential sites for the presence of brain abnormalities in patients with major depression, is consistent with previous studies. Until now, many researchers have suggested a depressive mood-related lesion, which is possibly located in the prefrontal cortex, cingulate gyrus, amygdala, thalamus, striatum, and so on.9,23,24 Dysfunction of the dorsomedial/dorsal anterolateral prefrontal cortex may impair ability to modulate emotional responses in mood disorders.25

Nagamachi et al. reported depression-related lesions in transient hypothyroidism.26 On SPM-96 analysis of 99mTc Hexamethyl propylene amine oxime (HMPAO) brain SPECT, there was significant cerebral hypoperfusion in the posterior part of the bilateral parietal lobes and in part of the bilateral occipital lobes, including the cuneus. Those decreased rCBF areas extended to the bilateral prefrontal cortices as deterioration became more profound. The dorsal prefrontal cortex has been found to be activated in response to emotional stimuli in normal controls by both positive and negative emotional stimuli.27,28 This region may be involved in the modulation of emotional responses, suggesting that dysfunction of this region will produce heightened or abnormal physiological and or psychological responses to stressful stimuli.29

The present study found decreased glucose metabolism in the anterior cingulate cortex (BA 32). Contrary to the current study, in patients with secondary depressive episodes after pancreatic cancer diagnosis, higher metabolism was found in the subgenual anterior cingulate cortex and this higher metabolism may be associated with the pathophysiology of secondary depressive episodes in patients following pancreatic cancer diagnosis.12 Individual studies, however, have reported a decrease in activity in this site that may be partially explained by reduced volume of this site: this effect of volume could easily confound evaluation of the degree of resting change across studies.10 Activity relative to volume may actually be increased, which would be consistent with the present finding of an overall reduction in activity with selective serotonin re-uptake inhibitor treatment.29,30

Interestingly, the present study found that there was a negative correlation of cerebral glucose metabolism with HDRS for the right orbitofrontal cortex in PDCKD patients. In major depression, orbitofrontal hypometabolism has been demonstrated on FDG PET.31 Structural magnetic resonance imaging studies have found evidence of reduction in gray matter volumes specifically in the orbitofrontal cortex.32,33 A recent study has suggested the hypothesis that orbitofrontal hypometabolism may act as a predisposing risk factor for the development of depression in patients with temporal lobe epilepsy.13 Also, a previous study has suggested that a reduction in volume of the orbitofrontal cortex provides additional evidence suggesting that this region of the brain may be important in the pathophysiology of depression.34 It has been shown that depression severity correlates positively with blood flow and glucose metabolism in the amygdala and negatively with blood flow or glucose metabolism in the prefrontal, cingulate, and temporoparietal cortex.24,25

A potentially important drawback of the present study was the lack of a non-depressed group of patients suffering PDCKD; also, it is difficult to identify the exact cause of cerebral glucose metabolism change in PDCKD patients. These problems should be clarified in future studies.

CONCLUSION

PDCKD patients with depressive symptoms had decreased cerebral glucose metabolism in several brain areas including both prefrontal cortices, right basolateral prefrontal cortex, left anterior cingulate gyrus, left premotor cortex, left transverse temporal gyrus, left superior temporal gyrus, right inferior parietal lobule, left middle temporal gyrus, and left angular gyrus. The right orbitofrontal cortex was found to have a negative correlation with HDRS in PDCKD patients. The present findings provide functional neuroimaging support for abnormal cerebral glucose metabolism in PDCKD patients with depressive symptoms.

ACKNOWLEDGMENTS

The authors thank the psychiatrists for assistance with psychiatric interview. This study was supported by a grant from the Medical Research Institute of Pusan National University (Grant No, 2007-28).

Ancillary