MRI High-Intensity Signals in Late-Life Depression and Alzheimer's Disease: A Comparison of Subjects Without Major Vascular Risk Factors

The authors examined periventricular white matter, deep white matter, and subcortical magnetic resonance imaging (MRI) high-intensity signals in subjects with late-life depression, probable Alzheimer's disease (DAT), and healthy, age-matched controls. All subjects were healthy and free of major vascular risk factors, including hypertension. MRIs were performed using a 1.5-tesla GE Signa scanner. T2 and proton-density-weighted images were analyzed by a neuroradiologist blind to the clinical status of all subjects. There were no statistically significant differences on any of the MRI indices between the groups studied. These data demonstrate that late-life depression, like DAT, in the absence of major vascular risk factors, is not associated with a significant increase in MRI high-intensity signals when compared to healthy, control subjects. The authors examined periventricular white matter, deep white matter, and subcortical magnetic resonance imaging (MRI) high-intensity signals in subjects with late-life depression, probable Alzheimer's disease (DAT), and healthy, age-matched controls. All subjects were healthy and free of major vascular risk factors, including hypertension. MRIs were performed using a 1.5-tesla GE Signa scanner. T2 and proton-density-weighted images were analyzed by a neuroradiologist blind to the clinical status of all subjects. There were no statistically significant differences on any of the MRI indices between the groups studied. These data demonstrate that late-life depression, like DAT, in the absence of major vascular risk factors, is not associated with a significant increase in MRI high-intensity signals when compared to healthy, control subjects. Major depression (MDD) is one of the most common psychiatric illnesses of old age.1Blazer DG Hughes DC George LK The epidemiology of depression in an elderly community population.Gerontologist. 1987; 27: 281-287Crossref PubMed Scopus (543) Google Scholar Studies of noninstitutionalized older patients reveal that although the prevalence of MDD is approximately 1%, significant and clinically relevant depressive symptomatology is present in 10–15% of this group.1Blazer DG Hughes DC George LK The epidemiology of depression in an elderly community population.Gerontologist. 1987; 27: 281-287Crossref PubMed Scopus (543) Google Scholar, 2Parmelee PA Katz IR Lawton MP Depression of institutionalized aged: assessment and prevalence.J Gerontol. 1989; 44: M22-M29Crossref PubMed Scopus (316) Google Scholar, 3Rovner BW German PS Brant LJ et al.Depression and mortality in nursing homes.JAMA. 1991; 265: 993-996Crossref PubMed Scopus (458) Google Scholar Among nursing home residents, the rates are even higher. It is estimated that 20–25% of cognitively intact nursing home residents meet DSM-III-R criteria for MDD, and an even higher prevalence rate exists for other forms of depression in this population.23Roth M Depression and affective disorder in late life.in: Angst J The Origins of Depression: Current Concepts and Approaches. Springer-Verlag, Berlin, Germany1986: 39-75Google Scholar Depression in late life often occurs concurrently with a variety of chronic illnesses, such as cardiovascular disease and neurological illness (e.g., poststroke depression) and cancer.4Ruegg RG Zisook S Swerdlow NR Depression in the aged: an overview.Psychiatr Clin Norm Am. 1988; 11: 83-99PubMed Google Scholar In a review of late-onset depression (i.e., first episode after age 60 years), Alexopoulos noted that medical illness was felt to be more influential than genetic factors in the pathogenesis of late-onset depression, whereas the reverse may be true with depression of earlier onset.5Alexopoulos GS Young RC Meyers BS et al.Late-onset depression.Psychiatr Clin North Amer. 1988; 11: 101-115PubMed Google Scholar Meyers and Greenberg suggest that late-onset depression is associated with a higher prevalence of delusions.6Meyers BS Greenberg R Late-life delusional depression.J Affect Disord. 1986; 11: 133-137Abstract Full Text PDF PubMed Scopus (96) Google Scholar Several investigators have postulated that late-onset depression may represent a specific depressive subgroup mediated by structural changes in the caudate-frontal areas.5Alexopoulos GS Young RC Meyers BS et al.Late-onset depression.Psychiatr Clin North Amer. 1988; 11: 101-115PubMed Google Scholar, 7Figiel GS Krishnan KRR Doraiswamy PM et al.Subcortical hyperintensities on brain magnetic imaging: a comparison between late-age-onset and early-onset elderly, depressed subjects.Neurobiol Aging. 1991; 12: 245-247Abstract Full Text PDF PubMed Scopus (200) Google Scholar The advent of highly sensitive imaging techniques like computerized tomography (CT) and magnetic resonance imaging (MRI) has led to many reports regarding structural abnormalities in MDD as well as other neuropsychiatric illnesses, including Alzheimer's disease (AD).8Sze G De Armond SJ Brant-Zawadzki M et al.Foci of MRI signal (pseudo lesions) anterior to the frontal horns: histologic correlation of a normal finding.American Journal of Radiology. 1986; 147: 331-337Google Scholar, 9Coffey CE Figiel GS Djang WT et al.Subcortical hyperintensity on magnetic resonance imaging: a comparison of normal and depressed elderly subjects.Am J Psychiatry. 1990; 147: 187-189Crossref PubMed Scopus (355) Google Scholar, 10Krishnan KRR Neuroanatomic substrates of depression in the elderly.J Geriatr Psychiatry Neurol. 1993; 1: 39-58Crossref Scopus (143) Google Scholar Among these abnormalities are high-intensity signals seen primarily in the periventricular white matter, deep white matter, and subcortical regions of the brain. Heretofore, many of the papers on this topic noted an increased prevalence of these signals in depressed older patients compared to age-matched control subjects.9Coffey CE Figiel GS Djang WT et al.Subcortical hyperintensity on magnetic resonance imaging: a comparison of normal and depressed elderly subjects.Am J Psychiatry. 1990; 147: 187-189Crossref PubMed Scopus (355) Google Scholar, 10Krishnan KRR Neuroanatomic substrates of depression in the elderly.J Geriatr Psychiatry Neurol. 1993; 1: 39-58Crossref Scopus (143) Google Scholar, 11Coffey CE Figiel GS Djang WT et al.Leukoencephalopathy in elderly depressed patients referred for ECT.Biol Psychiatry. 1988; 24: 143-161Abstract Full Text PDF PubMed Scopus (176) Google Scholar, 12Coffey CE Figiel GS Djang WT et al.Subcortical white matter hyperintensity on magnetic resonance imaging: clinical and neuroanatomic correlates in the depressed elderly.J Neuropsychiatry Clin Neurosci. 1989; 1: 135-144Crossref PubMed Scopus (197) Google Scholar, 13Krishnan KRR Goli V Ellinwood EH et al.Leukoencephalopathy in patients diagnosed as major depressive.Biol Psychiatry. 1988; 23: 519-522Abstract Full Text PDF PubMed Scopus (197) Google Scholar In a replication and extension of earlier studies, Coffey et al.14Coffey CE Wilkinson WE Weiner RD et al.Quantitative cerebral anatomy in depression: a controlled magnetic resonance imaging study.Arch Gen Psychiatry. 1993; 50: 7-16Crossref PubMed Scopus (405) Google Scholar found a higher prevalence of subcortical hyperintensities (particularly in the periventricular white matter) in both younger and elderly depressed patients compared to normal, control subjects. Rabins and associates found that these MRI hyperintensities extended beyond the white matter and periventricular areas to include the cortex, and the degree of subcortical change in depressed patients was related only to the presence of an affective disorder and not to age at onset of illness. Other investigators, however, reported that there was a higher prevalence of high-intensity signals in patients with late-onset depression (> 60 years and > 45 years, respectively) when compared to early-onset depression.7Figiel GS Krishnan KRR Doraiswamy PM et al.Subcortical hyperintensities on brain magnetic imaging: a comparison between late-age-onset and early-onset elderly, depressed subjects.Neurobiol Aging. 1991; 12: 245-247Abstract Full Text PDF PubMed Scopus (200) Google Scholar, 13Krishnan KRR Goli V Ellinwood EH et al.Leukoencephalopathy in patients diagnosed as major depressive.Biol Psychiatry. 1988; 23: 519-522Abstract Full Text PDF PubMed Scopus (197) Google Scholar Although there is an emerging consensus that a subgroup of subjects with late-life depression present with increased high-intensity signals on MRI, the exact etiology of these signals and their role in the pathogenesis of depression remain unclear.16Morris P Rapoport SI Neuroimaging and affective disorders in late life: a review.Can J Psychiatry. 1990; 35: 347-354PubMed Google Scholar In an earlier report, we showed that the prevalence of MRI high-intensity signal abnormalities was not significantly different between our two study groups—patients with AD and healthy, elderly control subjects—when both were free of cerebrovascular risk factors (e.g., hypertension, stroke, transient ischemic attack, and diabetes).17Kumar A Yousem D Souder E et al.Highintensity signals in Alzheimer's disease without cerebrovascular risk factors: a magnetic resonance imaging evaluation.Am J Psychiatry. 1992; 149: 248-250Crossref PubMed Scopus (19) Google Scholar Other studies suggest an association between increasing age, vascular risk factors, and MRI high-intensity signals in both healthy aging and MDD.12Coffey CE Figiel GS Djang WT et al.Subcortical white matter hyperintensity on magnetic resonance imaging: clinical and neuroanatomic correlates in the depressed elderly.J Neuropsychiatry Clin Neurosci. 1989; 1: 135-144Crossref PubMed Scopus (197) Google Scholar, 18Fazekas F Chawluk JB Alavi A et al.MR signal abnormalities at 1.5 tesla in Alzheimer's disease and normal aging.American Journal of Neuroradiology. 1987; 8: 421-426Google Scholar Because of the association between high-intensity signals and vascular risk factors, we set out to examine the relationship between late-life depression without major vascular risk factors and MRI high-intensity signals. It was our hypothesis that the prevalence of MRI high-intensity signals in MDD not associated with vascular risk factors would not be significantly different from age-matched control subjects. We also hypothesized that the prevalence of MRI high-intensity signals in late-life depression would be similar to AD patients, when both were free of vascular risk factors. Study groups were divided as follows: 19 subjects (13 women/6 men; age range = 62–87 years; mean age = 69 years; standard deviation [SD] = 6) who met DSM-III-R criteria for an MDD as determined by history and a diagnostic interview by two board-certified psychiatrists; 16 subjects (8 women/8 men; age range = 56–82 years, mean age = 67 years; SD = 8) who met NINCDS-ADRDA criteria for probable AD [reported on previously]17Kumar A Yousem D Souder E et al.Highintensity signals in Alzheimer's disease without cerebrovascular risk factors: a magnetic resonance imaging evaluation.Am J Psychiatry. 1992; 149: 248-250Crossref PubMed Scopus (19) Google Scholar); and 23 healthy, aged-matched control subjects (12 women/11 men; age range = 57–83 years; mean age = 68 years; SD = 8). The depressed and AD patients were recruited from both the inpatient and ambulatory care services in the Department of Psychiatry at the Hospital of the University of Pennsylvania. Control subjects were obtained in response to advertisements placed at senior centers and retirement centers. All subjects in this study were free of major vascular risk factors such as hypertension, stroke, transient ischemic attack, diabetes, and atrial fibrillation. The depressed subjects all had a minimum score of 15 on the 17-item Hamilton Depression Rating Scale19Hamilton M Development of a rating scale for primary depressive illness.British Journal of Social and Clinical Psychology. 1967; 6: 278-296Crossref PubMed Scopus (7304) Google Scholar (mean = 22; SD = 6). All but two of the AD subjects had Mini-Mental State Exam (MMSE)20Folstein 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: 189-198Abstract Full Text PDF PubMed Scopus (72288) Google Scholar scores greater than 16 (mean = 23; SD = 5; range = 12–28). In addition to the MMSE, all patients diagnosed with AD had a clinical history consistent with both intellectual and functional decline and met DSM-III-R and NINCDS-ADRDA criteria for the disease. Depressed patients were screened for dementia based on history and a detailed mental status assessment (that included the MMSE) on which all study subjects scored at least 26/30 (mean = 29; SD = 2). A complete medical history was elicited from each subject (with input from other sources when indicated). All subjects also underwent a complete physical and neurological examination. Control subjects were screened using the same health criteria as both the depressed and AD subjects. Also, the control subjects were free from significant psychiatric, systemic, and neurologic disease, including head injury. The laboratory indices—complete and differential blood counts; hepatic, renal, and thyroid panels; B12 and folate levels; and syphilis serologies were normal in all study participants. The MRI studies were performed with a 1.5-tesla General Electric Signa scanner (Milwaukee, WI) and a quadrature head coil. Scans included sagittal short-repetition time and axial long-repetition time series with 5-mm contiguous slices. A qualified neuroradiologist (DY), who was blind to the clinical status of all subjects, evaluated the transaxial T2-weighted and proton-density images (repetition time = 3,000 msec; echo times = 30 msec and 80 msec) for high-intensity signals in the periventricular white matter, deep white matter, and subcortical gray nuclei regions. The neuroradiologist rated the degree of high-intensity signals in each region, using a 0–3 severity scale that is widely used in clinical studies.18Fazekas F Chawluk JB Alavi A et al.MR signal abnormalities at 1.5 tesla in Alzheimer's disease and normal aging.American Journal of Neuroradiology. 1987; 8: 421-426Google Scholar In the periventricular area, 0 = absent, 1 = frontal or occipital caps or pencil-thin lining of the lateral ventricle, 2 = smooth halo around the lateral ventricle, and 3 = irregular periventricular high-intensity signals extending into the deep white matter. Deep white-matter high intensities were similarly graded: 0 = absent, 1 = punctate foci, 2 = early confluence of foci, 3 = large confluence of foci.18Fazekas F Chawluk JB Alavi A et al.MR signal abnormalities at 1.5 tesla in Alzheimer's disease and normal aging.American Journal of Neuroradiology. 1987; 8: 421-426Google Scholar High-intensity signals in the caudate and lenticular nuclei and thalami were rated 0 = absent, 1 = punctate, 2 = multi-punctate, and 3 = diffuse.9Coffey CE Figiel GS Djang WT et al.Subcortical hyperintensity on magnetic resonance imaging: a comparison of normal and depressed elderly subjects.Am J Psychiatry. 1990; 147: 187-189Crossref PubMed Scopus (355) Google Scholar Differences in ratings between the depressed and AD patients and the control group with regard to periventricular white matter, deep white matter, and subcortical gray nuclei high-intensity signals were assessed using the Robust Rank-Order Test. The Spearman rank-order correlation coefficient was used to determine whether there was any association among the ratings of periventricular, deep white matter, and subcortical high-intensity signals, the depressed patients' Hamilton Depression scores, and their age. These findings categorize each group with regard to the prevalence and severity of MRI high-intensity signals. Assessing periventricular white matter among the depressed subjects, 84% (n = 16) were rated Grade 0–1, and 16% (n = 3) were Grade 2 or higher. In subjects with AD, 63% (n = 10) were graded 0–1, and 37% (n = 6), 2 or more. The control subjects showed 83% (n = 19) at Grade 0–1 and 17% (n = 4) at Grade 2 or 3. For deep white matter: among the depressed subjects, 79% (n = 15) were rated Grade 1 or less and 21% (n = 4) were 2 or more. For the AD group, 81% (n = 13) were graded 0–1, and 19% (n = 3) were graded 2–3. In the control group, 91% (n=21) were Grade 0 or 1, and 9% (n = 2), Grade 2 or 3. For subcortical high-intensity signals, no depressed subjects, 6% of AD subjects (n = 1), and 9% of control subjects (n = 2) had a grade of 2 or more. The exact percentages in each category and grade are shown in Table 1. On all three measures, the differences between the depressed and AD subjects and the control subjects were not statistically significant (Robust Rank Order Test, P > 0.05).TABLE 1Percentage of subjects meeting criteria for MRI high-intensity signal Grades 0–3 by area and conditionDWMPVWSCMDDDATCTRLMDDDATCTRLMDDDATCTRLGrades 0–17981918463831009491Grade 216139161213069Grade 35600254000Note: DWM = deep white matter; PVW = periventricular white matter; SC = subcortical gray nuclei; MDD = major depressive disorder; DAT = probable Alzheimer's disease; CTRL = control subjects. Open table in a new tab Note: DWM = deep white matter; PVW = periventricular white matter; SC = subcortical gray nuclei; MDD = major depressive disorder; DAT = probable Alzheimer's disease; CTRL = control subjects. When the depressed subjects were divided between early- and late-onset groups (onset < 60 years old vs. > 60 years old), 0% of those with early onset showed a periventricular white-matter grade greater than 2, compared with 25% (n = 3) of the late-onset subgroup. For deep white matter, none of the early-onset group had a grade above 2, vs. 33% (n = 4) of those with late onset. Neither of these differences, however, reached statistical significance. By use of the Spearman rank-order correlation test (after correcting for ties), we found that among those with late-life depression (both early- and late-onset) there was a positive correlation between deep white-matter hyperintensities and age (r = 0.68; P < 0.002). Within the AD group, there was a positive correlation between age and both periventricular white-matter high intensity (r = 0.55; P < 0.01) and deep white-matter high intensity (r = 0.55; P < 0.01) and between subcortical high-intensity signals and (r = P < hypothesis is the data that when free of major vascular risk factors such as hypertension, diabetes, and cerebral ischemic our study group of older depressed patients not a statistically significant higher prevalence of MRI high-intensity signals than seen in our healthy, age-matched control subjects. Among the depressed patients, there was a positive correlation between the severity of deep high-intensity signals and age (r = 0.68; P < 0.002). In those subjects with AD, both periventricular and deep white-matter high-intensity signals increased significantly with age. Although the prevalence of deep high-intensity signals among the depressed group is higher than in our control subjects, this not statistical in our When periventricular and subcortical high-intensity signals were there was a and also in prevalence between depressed patients and control subjects. The prevalence of MRI high-intensity signals in our healthy, age-matched control subjects was and consistent with reports that a similarly healthy group of older WE et matter hyperintensities in dementia of Alzheimer's and in healthy subjects without cerebrovascular risk factors: a magnetic resonance imaging study.Arch Neurol. 1990; PubMed Scopus Google Scholar Although the prevalence of high-intensity signals in brain regions in our group with late-life depression was greater than that seen in control subjects, the not and may be in by the of our It be noted that the prevalence of these signals seen in the deep white matter and subcortical gray nuclei not in the periventricular white matter) areas in our of older depressed patients was to that of our AD patients without vascular risk A Yousem D Souder E et al.Highintensity signals in Alzheimer's disease without cerebrovascular risk factors: a magnetic resonance imaging evaluation.Am J Psychiatry. 1992; 149: 248-250Crossref PubMed Scopus (19) Google Scholar Several investigators have that late-life depressive illness is to changes or brain by cerebrovascular F The of affective in old age. University Scholar, M Depression and affective disorder in late life.in: Angst J The Origins of Depression: Current Concepts and Approaches. Springer-Verlag, Berlin, Germany1986: 39-75Google Scholar reports MRI white-matter high-intensity signals in of depressed older subjects referred for CE Figiel GS Djang WT et al.Subcortical white matter hyperintensity on magnetic resonance imaging: clinical and neuroanatomic correlates in the depressed elderly.J Neuropsychiatry Clin Neurosci. 1989; 1: 135-144Crossref PubMed Scopus (197) Google Scholar Also, the same group found of subcortical gray nuclei in more than of a similar study CE Figiel GS Djang WT et al.Subcortical white matter hyperintensity on magnetic resonance imaging: clinical and neuroanatomic correlates in the depressed elderly.J Neuropsychiatry Clin Neurosci. 1989; 1: 135-144Crossref PubMed Scopus (197) Google Scholar of these patients had risk factors for (e.g., hypertension, and ischemic and had neurological Figiel et reported an increased prevalence of deep white-matter high-intensity signals in late-onset depression compared to those with early GS Krishnan KRR Doraiswamy PM et al.Subcortical hyperintensities on brain magnetic imaging: a comparison between late-age-onset and early-onset elderly, depressed subjects.Neurobiol Aging. 1991; 12: 245-247Abstract Full Text PDF PubMed Scopus (200) Google Scholar It be noted that in the vascular risk factors were not and were present in of those with late but only of those with early GS Krishnan KRR Doraiswamy PM et al.Subcortical hyperintensities on brain magnetic imaging: a comparison between late-age-onset and early-onset elderly, depressed subjects.Neurobiol Aging. 1991; 12: 245-247Abstract Full Text PDF PubMed Scopus (200) Google Scholar the association between vascular risk factors and MRI high-intensity signals, is to study of depressed and AD patients as well as control subjects who all risk factors that the of the association of MRI high-intensity signal and vascular MRI high-intensity signals may be by ischemic and vascular other for these signals include and of to the frontal G De Armond SJ Brant-Zawadzki M et al.Foci of MRI signal (pseudo lesions) anterior to the frontal horns: histologic correlation of a normal finding.American Journal of Radiology. 1986; 147: 331-337Google Scholar, et subcortical on magnetic resonance imaging in the elderly, correlation with age and cerebrovascular risk 1986; PubMed Scopus Google Scholar (n = of our depressed subjects to have had their first episode of depression after the age of 60 a variety of factors may this Among are of depression that not to psychiatric were diagnosed as or not diagnosed at and in one to have an earlier The large in prevalence of periventricular and deep white-matter high-intensity signals between early- and late-onset depressed patients not statistical and is by the of the depressive Although depression in older is often associated with medical the depressed older subjects in our study represent a subgroup of late-life depression that The AD patients reported were to have based on their It is that subjects with increased severity of dementia would increased prevalence of MRI high-intensity signals. our this is the first study to examine MRI high-intensity signals in depressed patients free of vascular risk factors. findings earlier studies that found a prevalence of MRI high-intensity signals in healthy, older control subjects free of major vascular risk factors. The presence of MRI high-intensity signals in elderly subjects is significantly with age and vascular risk factors, and cerebral from of the earlier in It that late-life depressed patients are a In all late-onset depression is not a of late-life factors may be for but not all of late-onset depression, and be when depressed data suggest that late-life depression, as is clinically and a vascular etiology associated with MRI high-intensity signals occurs in but not all of studies on the relationship between of late-life depression, vascular risk factors, and of high-intensity signals on