Late Life Depression

Evaluation and Treatment Center

www.latelifedepression.org

For Researchers:

What research is carried out at the Late-Life Depression Evaluation and Treatment Program? 

Our center has been funded by the NIMH since 1995; our major research mission has been to map pathways to and from depression in late life and to model variation in outcomes such as wellness, recurrence of depression, suicide, and placement in long-term care.

Strengths of our Center include the demographics of Pittsburgh (Allegheny is the number one county in the U.S. in terms of per capita elderly persons), the availability of a large number of senior and junior investigators, a longstanding commitment to aging research, and a large and diverse funding of research, career development, and research training grants.  

Here are six main goals of research being carried out in our center:

1.  To improve the recognition and treatment of depression in elderly patients seen in the general medical sector.

2. To accelerate the onset of antidepressant treatment response.

3. To improve the early recognition of treatment resistance in late life depression and to develop strategies for improving response and attaining recovery in such patients.

4. To demonstrate maintenance treatments with long term efficacy in late life depression, especially for patients over age 70 (many of who are also cognitively impaired) and for those who need ECT to recover.

5. To develop preventive interventions to reduce the risk for late-onset depression.

6. To develop strategies to minimize residual disability and to facilitate full recovery, especially after medically serious events (such as hip fracture or coronary artery bypass surgery).

See below for a list of our research findings and an abridged list of center publications.

The Department of Psychiatry at the University of Pittsburgh has post-doctoral training programs in geriatric psychiatry, including depression and Alzheimer's disease.  These programs have had a high success rate, in terms of trainees going on to successful academic careers.

If you are interested in research in geriatric psychiatry, you can call, write, or email for information.  M.D.s, Ph.D.s, and others with doctoral degrees are welcome.

Call: Kathy Slomka (412)-246-6455

email: slomkaka@upmc.edu

or write: Kathy Slomka, WPIC E1132, 3811 O'Hara Street, Pittsburgh, PA 15213

Main Research Findings at the Center:

2001

We have organized 2001 findings thematically, grouping them under the six major goals of the IRC/LLMD, as listed in the Overview.

1. To improve the recognition and treatment of depression in elderly patients seen in the general medical sector.

1. Enrolling suicidal patients into clinical trials demonstrates the need for experimental design to be informed by ethical demands for beneficence. PROSPECT’s use of treatment as usual control practices illustrates how the tension between research design and ethics can be managed (Reynolds CF, Degenholtz H, Parker L, Schulberg HC, Mulsant BH, Post EP, Rollman BL, PROSPECT Study Group. International Journal of Geriatric Psychiatry, 16(6): 602-680, 2001).
2. PROSPECT's depression care manager is an effective model for improving four-month depression outcomes in elderly primary care patients, but the effectiveness is better in white then in black patients (Katz IR, Alexopoulos GS, Reynolds CF: ACNP Study Group, December 10, 2001).
3. An analysis from data from the National Ambulatory Medical Care Surveys (NAMCS) indicate that after controlling for symptom presentation (including complaints of depressed mood or other symptoms of depression), primary care physicians were 56% less likely to record a diagnosis of depression during visits made by elderly patients, 37% less likely during visits by African Americans, and 35% less likely during visits by Medicaid patients. Visits with a depression diagnosis were, on average, 2.9 minutes longer in duration (16.4 vs. 19.3) than visits without a depression diagnosis. If rates of diagnosis are to improve, interventions that go beyond getting physicians to recognize the symptoms of depression are needed (Harman JS, Schulberg HC, Mulsant BH, Reynolds CF. Effect of patient and visit characteristics on diagnosis of depression in primary care. Journal of Family Practice, in press). Another analyis of NAMCS data reveals that significant differences in rates of treatment for depression during office visits made by African American patients, elderly patients, or patients on Medicaid that occurred in 1993-1994, are no longer evident in 1996-1997, reflecting improved rates of depression treatment in these vulnerable populations (Harman JS, Mulsant BH, Kelleher KJ, Schulberg HC, Kupfer DJ, Reynolds CF. Narrowing the gap in treatment of depression. International Journal of Psychiatry in Medicine, in press).

2. To accelerate the onset of antidepressant treatment response

1. Thus far, we have noted no significant differences in [¹¹C]WAY100635 binding to 5-HT1A receptors between patients and controls in the autoreceptor brainstem region or postsynaptic areas of high receptor binding (data shows a trend toward slightly lower binding in all regions in the patients) However, these data indicate a significant relationship between binding potential values in the hippocampus and time to remission among patients (r=0.76*; Spearman p<0.05) (Meltzer CC).
2. There is a paucity of data addressing the outcome of electroconvulsive therapy (ECT) in persons over 75 years of age. In a prospective study including 268 patients with primary, unipolar, major depressive episode, we found that despite a higher burden of physical illness and cognitive impairment, even the oldest patients (i.e., those 75 and older) with severe major depression tolerate ECT in a manner similar to younger patients and demonstrate similar or better acute response (Tew JD, Mulsant BH, Haskett RF, Prudic J, Thase ME, Crowe R, Dolata D, Begley AE, Reynolds CF, Sackeim HA. Acute efficacy of ECT in the treatment of major depression in the old-old. American Journal of Psychiatry, 156:1865-1870, 1999). Also, given the limited capacity to predict seizure threshold in ECT, we found that empirical titration remains the only accurate method to determine electrical dosage in unilateral ECT (Boylan LS, Haskett RF, Mulsant BH, Greenberg R, Prudic J, Spignall K, Lisanby SH, Sackeim HA. Determinants of seizure threshold in ECT: benzodiazepine use, anesthetic dosage, and other factors. Journal of ECT 16:3-18, 2000). In the same data-set, patients randomized to continuation pharmacotherapy after ECT had a lower relapse rate than patients randomized to receive placebo (relapse rate: 84%). However, only patients who received a combination of nortriptyline and lithium did relatively well (relapse rate: 39%); patients who received only nortriptyline had a high relapse rate (60%) (Sackeim HA, Haskett RF, Mulsant BH, Thase ME, Mann JJ, Pettinati HM, Greenberg RM,Crowe RR, Cooper TB, Prudic J. Continuation pharmacotherapy in the prevention of relapse following electroconvulsive therapy. Journal of the American Medical Association, 285(10):1299-1307, 2001).

3. To improve the early recognition of treatment resistance in late life depression and to develop strategies for improving response and attaining recovery in such patients

1. Decreased working memory and processing speed mediate cognitive impairment in geriatric depression (Nebes RD, Butters MA, Mulsant BH, Pollock BG, Zmuda M, Houck PR, Reynolds CF: Decreased working memory and processing speed mediate cognitive impairment in geriatric depression. Psychological Medicine, 30(3): 679-691, 2000).
2. The presence of comorbid anxiety does not appear to reduce the likelihood of successful outcomes in the standardized treatment of late-life depression in the mental health sector. (Lenze EJ, et al, presented at the AAGP annual meeting, Feb. 2002).
3. Anxiety disorders are common in adults with depressive disorders, but several studies have suggested a relatively low prevalence of anxiety disorders in older subjects with depression. Contrary to previous reports, in assessing lifetime and point prevalence rates and associated clinical features of anxiety disorders in 182 depressed elderly patients, we found a relatively high rate of current and lifetime anxiety disorders in elderly depressed individuals. Comorbid anxiety disorders and symptoms of GAD were associated with a more severe presentation of depressive illness in the elderly (Lenze EJ, Mulsant BH, Shear MK, Schulberg HC, Dew MA, Begley AE, Reynolds CF. Comorbid anxiety disorders in depressed elderly patients. American Journal of Psychiatry 157:722-728, 2000). In an open-trial, a 75% response rate was observed in older patients with GAD, panic disorder, or obsessive-compulsive disorder who completed treatment with the SSRI fluvoxamine median dose: 200 mg/day). However, drop-out rate was high in this population (37%), highlighting the difficulties treating older patients with anxiety (Wylie ME, Miller MD, Shear MK, Little JT, Mulsant BH, Pollock BG, Reynolds CF. Fluvoxamine pharmacotherapy of anxiety disorders in late life: Preliminary open-trial data. Journal of Geriatric Psychiatry and Neurology 13(1): 43-48, 2000). Pre-existing and co-existing anxiety disorders are highly prevalent in depressed elderly patients and are associated with measures of greater disability and increased rates of suicidal ideation (Lenze EJ, Rogers JC, Martire LM, Mulsant BH, Rollman BL, Dew MA, Schulz R, Reynolds CF: The association of late-life depression and anxiety with physical disability: A review of the literature and prospectus for future research. American Journal of Geriatric Psychiatry, 9(2): 113-135, 2001).
4. Endogenous concentrations of DHEA and DHES-S decrease with remission of depression in older adults (Fabian TJ, Dew MA, Pollock BG, Reynolds CF, Mulsant BH, Butters MA, Zmuda MD, Linares AM, Trottini M, Korboth PD. Biological Psychiatry, in press).
5. The course and rate of antidepressant response in the very old is as good as in the young old. Increasing age need not impair antidepressant treatment response (Gildengers AG, Houck PR, Mulsant BH, Pollock BG, Mazumdar S, Miller MD, Dew MA, Frank E, Kupfer DJ, Reynolds CF. Journal of Affective Disorders, in press).
6. CYP 2D6 genotyping with oligonucleotide microarrays predicts nortriptyline levels in geriatric depression (Murphy GM, Pollock BG, Kirshner M, Pascoe N, Cheuk W, Mulsant BH, Reynolds CF. Neuropsychopharmacology, in press).
7. Remission rates in depressed suicidal elderly patients are as good as those of non-suicidal patients, but remission is more brittle, with a higher rate of relapse during continuation treatment and greater need for adjunctive pharmacotherapy for anxiety and agitation (Szanto K, Mulsant BH, Houck PR, Miller MD, Mazumdar S, Reynolds CF. American Journal of Geriatric Psychiatry, 9(3): 261-268, 2001.
8. Analyses comparing the distribution of e2, e3, and e4 alleles in groups of LLD, AD, and elderly control subjects revealed that neither LLD, accompanying cognitive impairment, nor late age-of-onset were associated with an increased e4 allele frequency. This suggests that the risk of developing AD in the context of these syndromes is no greater than for the general population. The finding that age-of-onset of LLD was significantly reduced in e4 carriers is similar to the effect of e4 on age-of-onset in AD. This work will be presented at the Annual Meeting of the American Association for Geriatric Psychiatry to be held in Orlando, FL in 2/02. (Butters MA, Sweet RA, Mulsant BH, Kamboh MI, Pollock BG, Nebes RD, Begley AE, DeKosky ST, and Reynolds CF: APOE is associated with age-of-onset, but not cognition, in late-life depression).
9. Prior studies have shown that elders with psychiatric syndromes such as depression and dementia have poorer medical rehabilitation outcomes. In a study was carried out at a rehabilitation hospital, significant associations were found between symptoms and behaviors reflective of cognitive and motivational impairments and rehabilitation outcome. These findings suggest that psychiatric syndromes adversely affect rehabilitation outcomes primarily through impairments in cognition and motivation (Dorra HH, Lenze EJ, Yookyung K, Mulsant BH, Munin MC, Dew MA, Reynolds CF. Disability and Rehabilitation, submitted).

4. To demonstrate maintenance treatments with long term efficacy in late life depression, especially for patients over age 70 (many of who are also cognitively impaired) and for those who need ECT to recover

1. Combined pharmacotherapy and psychotherapy in maintenance treatment for late life depression is associated with better preservation of gains in social adjustment than is the use of monotherapy (Lenze EJ, Dew MA, Mazumdar S, Begley AE, Cleon C, Miller MD, Imber SD, Frank E, Kupfer DJ, Reynolds CF. American Journal of Psychiatry, in press).
2. We attempted to replicate and expand the important finding that impaired executive functioning is associated with higher rate of relapse and recurrence of late-life depression. We examined the effect of executive function and memory, at both baseline and post-treatment, on time to relapse in patients with late-life depression. We found no reliable evidence of an executive function-depression relapse relationship. Moreover, the effect sizes across studies were relatively modest, suggesting limitations in the practical significance of any relationship. (Butters MA, Mulsant BH, Pollock BG, Dew MA, Mazumdar S, Begley AE, Nebes RD, Reynolds CF. American Journal of Geriatric Psychiatry, under editorial review.
3. In preliminary, open-trial studies, the SSRI antidepressant paroxetine appears to be as effective as the older tricyclic antidepressant nortriptyline in preventing or delaying relapse and recurrence of major depressive episodes in the elderly (Walters G, Reynolds CF, Mulsant BH, Pollock BG: Continuation and maintenance pharmacotherapy in geriatric depression: an open-trial comparison of paroxetine and nortriptyline in patients over age 70. Journal of Clinical Psychiatry, 60(suppl): 38-44, 1999; Bump GM, Mulsant BH, Pollock BG, Mazumdar S, Begley AE, Dew MA, Reynolds CF: Paroxetine versus nortriptyline in the continuation and maintenance treatment of depression in the elderly. Depression and Anxiety, 13: 38-44, 2001).
4. Elderly patients with less severe depression and those who recover quickly from their index episodes remain well with monthly interpersonal psychotherapy and do not need antidepressant medication to prevent relapse and recurrence of major depressive episodes (Taylor MP, Reynolds CF, Frank E, Cornes C, Miller MD, Stack JA, Begley AE, Mazumdar S, Dew MA, Kupfer DJ: Which elderly depressed patients remain well on maintenance interpersonal psychotherapy alone? A report from the Pittsburgh study of maintenance therapies in late-life depression. Depression and Anxiety, 10(2): 55-60, 1999; Dew MA, Reynolds CF, Mulsant BH, Frank E, Houck PR, Mazumdar S, Begley AE, Kupfer DJ: Initial recovery patterns may predict which maintenance therapies for depression will keep older adults well. Journal of Affective Disorders, 65: 155-166, 2001).
5. The effect of nortriptyline and paroxetine on extrapyramidal signs and symptoms: a prospective double-blind study in depressed elderly patients. American Journal of Geriatric Psychiatry 8:226-231, 2000) or excessive weight gain (Weber E, Stack J, Pollock BG, Mulsant BH, Begley A, Mazumdar S, Reynolds CF. Weight change in older depressed patients during acute pharmacotherapy with paroxetine and nortriptyline: a double-blind randomized trial. American Journal of Geriatric Psychiatry 8:245-250, 2000).

5. To develop preventive interventions to reduce the risk for late-onset depression

1. Restricting time in bed by 30 minutes nightly helps to protect sleep quality over one year in elderly subjects (Hoch CC, Reynolds CF, Buysse DJ, Monk TH, Nowell PD, Begley AE, Hall F, Dew MA: Protecting sleep quality in later life: A pilot study of bed restriction and sleep hygiene. Journals of Gerontology: Series B, Psychological Sciences and Social Sciences, 56(1): 52-59, 2001).
2. From an analysis of the Cardiovascular Health Study dataset, we have found that 1) persistent depressive symptoms predict greater decline in functional ability, compared to brief depressive symptoms; 2) white matter hyperintensities and depressive symptoms exert additive, but not interactive, effects on functional decline (thus not supportive of vascular dep hypothesis); 3) White matter hyperintensities are predictive of chronicity of depressive symptoms, partly mediated by cognitive and functional decline (Lenze EJ, et al, presented in part at the AAGP annual meeting, February 2001).
3. Elderly control subjects with greater burden of white matter hyperintensity on MRI report more depressive symptoms than subjects with lesser WMH (Nebes RD, Vora IJ, Meltzer CC, Fukui MB, Williams RL, Kamboh MI, Saxton J, Houck PR, DeKosky ST, Reynolds CF: Relation of deep white matter hyperintensities and APOE genotype to depressive symptomatology in nondepressed older adults. American Journal of Psychiatry, 158(6): 878-884, 2001).
4. Among older persons, the impact of losing one's spouse varies as a function of the caregiving experiences that precede the death. For individuals who experience caregiver strain prior to the death of their spouse, the death itself does not increase their level of distress. Instead, they show reductions in health risk behaviors. Among noncaregivers, losing one's spouse results in increased depression and weight loss (Schulz R, Beach SR, Lind B, Martire LM, Zdaniuk B, Hirsch C, Jackson S, & Burton L, JAMA, 285:3123-3129,2001).
5. Poorer perceived quality of received spousal care is associated with more depressive symptoms in elder care-recipients one year later, even after controlling for sociodemographic factors, baseline depressive symptoms for both care-recipient and caregiver, and care-recipient disability, marital quality, and care-receiving strain (Martire L, Schulz R, Wrosch C, Newsom JT, Gerontological Society of America 54th Annual Meeting, Nov. 16, 2001).
6. Using event-related functional MRI to identify the brain regions engaged during explicit and implicit sequence learning has yielded interesting preliminary findings. In young control subjects, engagement of the frontal and striatal circuit in demonstrated in both implicit and explicit sequence learning. Preliminary fMRI results in late-life depression show decreased frontal and striatal activation during both implicit and explicit sequence learning. We have also observed a significant association between white matter hyperintensities and decreased implicit sequence learning performance in the elderly (Aizenstein et al. Neurology, under editorial review).
7. Applying voxel-based morphometry to volumetric MRI data, we have shown that depressed subjects had significantly reduced gray matter volume of the right hippocampus and bilateral middle frontal gyrus relative to healthy elders. The depressed subjects also had significantly less white matter volume in the regions of the left anterior cingulate gyrus and right middle frontal gyrus. In addition, the volume of the anterior most region of the hippocampus, possibly including the entorhinal cortex, was inversely associated with the number of years since first lifetime episode of depression, even after controlling for chronological age (McGinty et al. Am J Psychiatry, under editorial review).
8. We previously demonstrated a strong effect of age on [18F]altanserin binding to 5-HT2A receptors (Meltzer CC, Price J, Mathis CA, Greer PJ, Cantwell MN, Houck PR, Mulsant BH, Ben-Eliezer D, Lopresti B, DeKosky ST, Reynolds CF. American Journal of Psychiatry, 156(12):1871-1878, 1999). Further examination of the male and female patterns of binding suggests that among older subjects (above 60 years) women (HRT non-users) have lower binding values than men. These data suggest that the influence of age on binding measures may differ between men and women. Our hypothesis of a linear inverse relationship between 5-HT2A binding and age in men and non-linear relationship in women (with a more rapid fall in binding measures following menopause) will be tested in a pending R01 (MH63353; PI: Meltzer). This hypothesis is consistent with data from our laboratory supporting a protective role of hormone replacement therapy on 5-HT2A binding (Moses EL, Drevets WC, Smith G, Mathis CA, Kalro BN, Butters MA, Leondires MP, Greer PJ, Lopresti B, Loucks TL, and Berga SL. Biological Psychiatry, 48:854-860, 2000.
9. A review of dementia caregiver intervention studies conducted in the past 5 years, focused on issues of clinical significance, indicates that most studies meet criteria of social validity; study participants consistently rate the interventions as beneficial, helpful, or valuable. Some studies are able to achieve clinically significant outcomes by improving caregiver psychiatric symptomatology, but few are able to achieve clinically meaningful effects in improving the overall quality of life of caregivers. A small number of studies are able to demonstrate impressive socially significant outcomes (Schulz et al., under review).
10. White matter hyperintensities on MRI may contribute to a ‘vascular depression’ characterized by later age of illness onset; prominent psychomotor retardation, cognitive impairment, and disability. We examined the correlation of WMH with age of illness onset, and symptom profile in 101 elders meeting DSM criteria for a major depressive episode. Using a modified Cardiovascular Health Study scale, we found a positive correlation between age of illness onset and WMH scores. Cognitive measures (MMSE and DRS scores) negatively correlated with WMH burden even after controlling for current age. These preliminary data support the role of WMH in defining a specific illness course and symptom profile among elderly depressed patients. (Whyte E, et al: presented at the AAGP annual meeting, Feb. 2002).

6. To develop strategies to minimize residual disability and to facilitate full recovery, especially after medically serious events (such as hip fracture or coronary artery bypass surgery)

1. Higher levels of cerebrovascular risk factors do not impede depression treatment response in later life (Miller MD, Lenze EJ, Dew MA, Whyte E, Weber E, Begley AE, Reynolds CF. American Journal of Geriatric Psychiatry, in press).
2. Although pre-operative level of depressive symptoms did not affect post-CABG hospital length of stay (median 5.0 days), depressed CABG patients are far more likely to be rehospitalized for any cause over the following 12 months than nondepressed CABG patients (90% vs. 33%, p<0.002) (Rollman BL, unpublished observation).
3. Traumatic grief psychotherapy effectively reduces the symptoms of traumatic or complicated grief (Shear MK, Frank E, Foa EB, Cherry CR, Reynolds CF, Vander Bilt J, Masters S: Traumatic grief therapy: A pilot study. American Journal of Psychiatry, 158(9): 1506-1508, 2001.
4. Paroxetine is as effective as nortriptyline for bringing about remission of depression in older inpatients and outpatients (Mulsant BH, Sweet RA, Rosen J, Pollock BG, Flynn T, Begley A, Mazumdar S, Reynolds CF. Journal of Clinical Psychiatry, 62:597-604, 2001).
5. Daytime sleepiness predicts mortality and cardiovascular disease in older adults (Newman AB, Spiekerman CF, Enright P, Lefkowitz D, Manolio T, Reynolds CF, Robbins J: Daytime sleepiness predicts mortality and cardiovascular disease in older adults. Journal of the American Geriatrics Society, 48(2): 115-123, 2000).
6. Preliminary analyses of fMRI data during performance of a working memory task in patients with LLD and elderly controls suggest that the dorsolateral prefrontal-anterior cingulate circuit hypothesized to be involved in the cognitive symptoms of LLD can be activated and studied. (Aizenstein H: presented at the AAGP annual meeting, Feb. 2002).
7. 5HTTLPR-polymorphism influences platelet activation in geriatric depression (Whyte EM, Pollock BG, Wagner WR, Mulsant BH, Ferrell RE, McGinley P, Begley AE, Bensasi S, Mazumdar S, Reynolds CF. American Journal of Psychiatry, 158:2074-2076, 2001).
8. Elderly patients who are rehabilitating after a hip fracture have a poorer recovery if they are depressed. Those with high levels of depressive symptoms (Ham-D >15) had a lower improvement in functional ability, greater length of stay in acute rehabilitation, and higher likelihood of institutionalization. (Lenze EJ, et al, presented at the AAGP annual meeting, Feb. 2002).
9. A review of family-focused treatments for midlife mood disorder reveals an advantage of this approach over pharmacotherapy alone, for decreased or delayed symptom relapse and increased medication adherence. Research on late-life depression suggests that patients are less likely to drop out of continuation treatment when family members are provided with education and support. Together, these findings indicate that family-focused treatment may be a valuable method to implement and evaluate in the treatment of late-life mood disorder (Martire LM, Schulz R, & Reynolds CF; presented at the AAGP annual meeting, Feb. 2002).
10. Two commonly used tests of executive function, Trail Making Test- Part B and the Wisconsin Card Sorting Test, significantly predict functional status among elderly adults, accounting for 54% of the variance in performing instrumental activities of daily living. These executive measures contributed significantly to the prediction of functional status even after statistically controlling for age, sex, and education (Bell-McGinty S, Podell K, Franzen M, Baird A, Williams JM, International Journal of Geriatric Psychiatry, under review).

1999-2000

• Cerebral blood flow does not decline with successful aging. Quantitative [15O]water PET imaging was used to measure regional cerebral blood flow (CBF) in 27 healthy individuals aged 19-76 and corrected for partial volume effects using an MR-based algorithm. These data demonstrated an apparent aging decline in CBF that resolved after correcting the data for partial volume effects (Meltzer CC, Cantwell MN, Greer PJ, Ben-Eliezer D, Smith G, Frank G, Kaye WH, Houck PR, and Price JC: J Nucl Med. in press).
• Estrogen replacement therapy in postmenopausal women was found to inhibit the drug metabolizing enzyme CYP1A2, which is responsible for the metabolism of fluvoxamine, theophylline and caffeine (Pollock BG, Wylie M, Stack JA, Sorisio DA, Thompson DS, Kirshner MA, Folan MM, and Condifer KA: J Clin Pharmacol 39:936-940, 1999).
• In preliminary, open-trial studies, the SSRI antidepressant paroxetine appears to be as effective as the older tricyclic antidepressant nortriptyline in preventing or delaying relapse and recurrence of major depressive episodes in the elderly (Walters, Reynolds, et al., Journal of Clinical Psychiatry 60 (Suppl): 38-44, 1999; Bump GM, Mulsant BH, Pollock BG, Mazumdar S, Begley AE, Dew MA, and Reynolds CF: Depress Anxiety. in press).
• Elderly patients with less severe depression remain well with monthly interpersonal psychotherapy and do not need antidepressant medication to prevent relapse and recurrence of major depressive episodes (Taylor MP, Reynolds CF, Frank E, Cornes C, Miller MD, Stack JA, Begley AE, Mazumdar S, Dew MA, and Kupfer DJ: Depress Anxiety 10:55-60, 1999). Similarly, elderly depressed patients whose index episodes remit within 4-6 weeks are able to survive depression free on maintenance monotherapy with either nortriptyline or monthly interpersonal psychotherapy (Dew MA, Reynolds CF, Mulsant BH, Frank E, Houck PR, Mazumdar S, Begley AE, and Kupfer DJ: J Affect Disord. in press).
• Electroconvulsive treatment (ECT) is an effective treatment of major depression in the very old (Tew JD, Mulsant BH, Haskett RF, Prudic J, Thase ME, Crowe R, Dolata D, Begley AE, Reynolds CF, and Sackeim HA: Am J Psychiatry 156:1865-1870, 1999).
• Daytime sleepiness predicts mortality and cardiovascular disease in older adults (Newman AB, Spiekerman CF, Enright P, Lefkowitz D, Manolio T, Reynolds CF, and Robbins J: J Am Geriatr Soc 48:115-123, 2000).
• Co-existing anxiety disorders are highly prevalent in depressed elderly patients and are associated with  greater disability and increased suicidal ideation (Lenze EJ, Mulsant BH, Shear MK, Schulberg HC, Dew MA, Begley AE, Pollock BG, and Reynolds CF: Am J Psychiatry 157:722-728, 2000).
• Decreased working memory and processing speed mediate cognitive impairment in geriatric depression (Nebes RD, Butters MA, Mulsant BH, Pollock BG, Zmuda M, Houck PR, and Reynolds CF: Psychol Med 30:679-691, 2000).
• Restricting time in bed by 30 minutes nightly helps to protect sleep quality over one year in healthy elderly subjects (Hoch CC, Reynolds CF, Buysse DJ, Monk TH, Nowell PD, Begley AE, Hall F, and Dew MA: J Gerontol B Psych Sci Soc Sci. in press).
• In a randomized comparison, the SSRI antidepressant paroxetine appears to be as effective as the older tricyclic antidepressant nortriptyline in the acute treatment of geriatric depression, including patients with melancholic depression (Mulsant BH, Pollock BG, Nebes R, Miller M, Little JT, Stack J, Houck PR, Bensasi S, Mazumdar S, and Reynolds CF: J Clin Psychiatry 60:16-20, 1999). However, paroxetine appears to be better tolerated (Mulsant BH, et al, data presented at APA 2000). Neither seems to be associated with EPS (Mamo DC, Sweet RA, Mulsant BH, Pollock BG, Miller MD, Stack JA, Begley AE, and Reynolds CF: Am J Geriatr Psychiatry 8:226-231, 2000) or excessive weight gain (Weber E, Stack J, Pollock BG, Mulsant BH, Begley A, Mazumdar S, and Reynolds CF: Am J Geriatr Psychiatry 8:245-250, 2000).
• Data from the 1993, 1994, 1996, and 1997 National Ambulatory Medical Care Surveys (NAMCS) indicate that significant differences in rates of treatment for depression during office visits made by African American patients, elderly patients, or patients on Medicaid that occurred in 1993-1994, are no longer evident in 1996-1997, reflecting improved rates of depression treatment in these vulnerable populations (Harman JS, Mulsant BH, Kelleher K, Schulberg HC, Kupfer DJ, and Reynolds CF: Psychiatric Services. under review).
• Given the limited capacity to predict seizure threshold in ECT, empirical titration remains the only accurate method to determine electrical dosage in unilateral ECT (Boylan, Haskett , Mulsant et al, Journal of ECT 16(1):3-18, 2000).

• Elderly patients with major depression and accompanying cognitive impairment, show improvement in executive functions but not in memory, with successful depression treatment. (Butters MA, Becker JT, Nebes RD, Zmuda M, Mulsant BH, Pollock BG, and Reynolds CF: Am J Psychiatry. in press).
• Elderly patients with major depression exhibit a lack of activation of frontal lobe circuits while performing a working memory task. (Butters, Zmuda et al., presented at the 6^th International Conference on Functional Mapping of the Human Brain, June, 2000, San Antonio, TX)
• Platelet activation is increased in depressed  patients with ischemic heart disease. This finding may explain, in part, the association between depression and increased mortality in ischemic heart disease (Pollock BG, Laghrissi-Thode F, and Wagner WR: J Clin Psychopharmacol 20:137-140, 2000).
• Allelic variation in the gene that expresses the serotonin transporter is strongly associated with the speed of antidepressant response to a selective serotonin reuptake inhibitor (SSRI) in those suffering from late-life depression (Pollock BG, Ferrell RE, Mulsant BH, Mazumdar S, Miller MD, Sweet RA, Davis S, Kirshner MA, Houck PR, Stack JA, Reynolds CF, and Kupfer DJ: Neuropsychopharmacology 23:587-590, 2000).
• An antidepressant (citalopram) was found to be as efficacious as antipsychotic medication for non-depressed, Alzheimer's disease patients hospitalized because of severe behavioral problems (Pollock BG, Hirschfeld R, Rush AJ, APA Annual Meeting- New Research, 2000).
• Estrogen replacement therapy in post-menopausal women was found to inhibit one of the important drug metabolizing enzymes (CYP 1A2) (Pollock BG, Wylie M, Stack JA, Sorisio DA, Thompson DS, Kirshner MA, Folan MM, and Condifer KA: J Clin Pharmacol 39:936-940, 1999).
• Elderly spousal caregiving is an independent risk factor for mortality. Strained spousal caregivers were 63 % more likely to die within four years than controls. The mortality effect is in part mediated by depression (Schulz R and Beach S: JAMA 282:2215-2219, 1999 (lead article) Editorial on article: Kiecolt-Glaser, J. (1999). Journal of the American Medical Association, 282, 217; Beach SR & Schulz R: Geriatric Times, 1, 26-28, 2000 (edited reprint of JAMA article).
• In one of the largest and most comprehensive studies to date, we found that high levels of depressive symptomatology are an independent risk factor for mortality in the elderly. (Schulz R, Beach SR, Ives DG, Martire LM, Ariyo AA, & Kop W: Archives of Internal Medicine, 160, 1761-1768. (lead article) Editorial on article: Wulsin, L. R. (2000). Does depression kill? Archives of Internal Medicine, 160, 1731-1732).
• In a review of the literature on family caregiving, we found that females respond more negatively to the demands of caregiving at all stages of the stress-health process. (Yee J & Schulz R: The Gerontologist, 40(2), 147-164).
• We have developed a "cascade to death model" to show how depression contributes to increased mortality in the elderly. (Schulz R, Martire LM, Beach SR & Scheier M: Current Directions in Psychological Science, in press).
• Applying a person-environment fit model, we show that older arthritis patients who receive high levels of assistance from their spousal caregiver have fewer negative reactions to this assistance if it is not highly important to them to be functionally independent. Greater negative reactions to assistance are associated with the care-recipient's increased depressive symptomatology over time (Martire LM, Stephens MAP, Druley JA, & Wojno WC, under review).
• In a study designed to better understand the role of caregiving in adjustment to bereavement we show that, following the death of a spouse, strained caregivers have improved health practices and no further increases in depressive symptomatology whereas non-caregivers and non-strained caregivers have significant increases in depressive symptomatology (Schulz R, Beach SR, Lind B, Martire LM, Zdaniuk B, Hirsch C, Jackson S, Burton L, under review).
• Gender differences in aging effect serotonin-1A binding. Preliminary data support gender differences in the effect of age on the 5-HT1A receptor, with aging reductions in binding observed in men only. Ongoing studies are exploring the role of hormone replacement therapy on central serotonin-1A receptor in elderly women. (Meltzer CC, Drevets WC, Price JC, Mathis CA, Lopresti B, Greer PJ, Villemagne VL, Holt D, Mason N, Houck PR, Reynolds CF, and DeKosky ST: Brain Res. in press).
• We have observed an effect of hormone replacement therapy on serotonin-2A binding. [¹⁸F]altanserin PET imaging of the 5-HT2A receptor in post-menopausal subjects before and after initiation of hormone replacement therapy indicates induced increases in prefrontal binding. (Moses EL, Drevets WC, Smith G, Mathis CA, Kalro BN, Butters MA, Leondires MP, Greer PJ, Lopresti B, Loucks TL, and Berga SL: Biol Psychiatry. in press)
• We have developed a novel method for imaging the serotonin transporter. Human studies with [¹¹C]McN5652 PET have demonstrated that serotonin transporter binding can be measured in humans using a single injection of the active enantiomer [¹¹C](+)-McN5652. (Lopresti BJ, Mathis CA, Price JC, Villemagne VL, Meltzer CC, Holt D, Smith GS, and Moore RY: Molecular and Pharmacological Brain Imaging with Positron Emission Tomography, in press).
• Preliminary studies of 5-HT1A receptor binding using PET and [11C-carbonyl] WAY 100635 has shown significantly increased 5-HT1A receptor binding in Type 2 diabetes relative to controls in the mesial temporal cortex. (Price JC, Kelley DE, Ryan CM, Meltzer CC, Drevets WC, Mathis CA, Mazumdar S, and Reynolds CF: Diabetes. under review).
• Item response theory (IRT) models were used to test the equivalence of the Beck Depression Inventory (BDI) in late-life (age 60 or older, n = 218) and mid-life patients (less than 60, n = 613). For 17 of the 21 items on the BDI, the responses of late-life patients differed significantly from responses of younger patients. Late-life patients tended to report fewer cognitive symptoms (e.g., disappointment in self, self-criticism, guilt, and sense of failure), especially at low to average levels of depression. Conversely, they tended to report more somatic symptoms (e.g., sleep disturbance, somatic preoccupation, weight loss), especially at higher levels of depression (Kim Y, Pilkonis P, Frank E, Thase M, & Reynolds CF, Psychology and Aging, under review).
• Regional measurement of dopaminergic function with 3D PET. 3D PET measurements of CBF before and after d-amphetamine have been performed in baboons. Preliminary results indicate modest pre- and post-amphetamine variations in CBF that are within the test-retest variability expected for CBF PET studies. (supported by Whitaker Foundation, PI: Julie C. Price).
• Extended characterization of the 5-HT2A receptor ligand [¹⁸F]altanserin. Building on our prior studies characterizing the metabolism and compartmental modeling of [¹⁸F]altanserin, we have recently successfully implemented a data analysis method that substitutes the cerebellar time-activity data rather than a metabolite-corrected arterial input function. This simplification permits quantitative binding measures to be obtained without invasive arterial catheterization and costly metabolite analysis. (Price JC, Lopresti B, Mason N, Holt D, Meltzer CC, Smith GS, Gunn R, Huang Y, and Mathis CA: Synapse, under review).
• Relationship between deep white matter hyperintensities and depressive symptoms. Using a new scale developed within the SFBIC, we found that hyperintensities present in the deep white matter, but not in the periventricular white matter were associated with increased depressive symptomatology and especially symptoms of impaired motivation, concentration and decision making (Nebes RD, Vora IJ, Meltzer CC, Fukui MB, Williams RL, Kamboh MI, Saxton J, Reynolds CF, and DeKosky ST: Am J Psychiatry, under review). This is consistent with our hypothesis that structural brain insults may contribute to the development, form and treatment responsiveness characteristics of late-life depression.

1995-1999

1.  Randomized Clinical Trials (RCTs) to establish short- and long-term outcomes

• Elderly patients with recurrent episodes of major depression benefit as much as mid-life patients from combined treatment with antidepressant medication and interpersonal psychotherapy. Thus, in both groups approximately 70% of patients recovered (intent-to-treat sample). The temporal course of response to treatment is somewhat slower in late life, however, and a higher proportion of elderly patients experience relapse during continuation treatment, suggesting that in some elderly patients response may be brittle (Reynolds CF, Frank E, Kupfer DJ, Thase ME, Perel JM, Mazumdar S, Houck PR, American Journal of Psychiatry, 153(10):1288-1292, 1996).

• In a mixed-age sample of 100 patients (mean age: 62) with non-bipolar non-psychotic major depression, patients who had failed one or more adequate medication trials were less likely to respond to subsequent treatment with ECT (61% response rate) than patients who were not medication- resistant (91% response rate). This study challenges the widely-held belief that the likelihood of response to ECT is independent of previous treatment with antidepressant medications and challenges the field to improve outcome in treatment-resistant patients. (Prudic J, Haskett RF, Mulsant BH, Mann JJ, Pettinati HM, Stevens S, Greenberg R, Rifas SL, Sackeim HA. American Journal of Psychiatry, 153: 985-992; 1996).

While both nortriptyline and interpersonal psychotherapy are superior to placebo in preventing recurrence of major depressive episodes in elderly patients, the best three-year outcome was associated with combined treatment using NT + IPT. Recurrence rates over three years of maintenance treatment were as follows: NT + IPT (20%); NT + medication clinic (43%); IPT + placebo (64%); and placebo + medication clinic (90%) (Reynolds CF, Frank E, Perel JM, Imber SD, Cornes C. Miller MD, Mazumdar S, Houck PR, Dew MA, Stack JA, Pollock BG, Kupfer DJ. Journal of the American Medical Association, 281:39-45; 1999).

• Full-dose maintenance nortriptyline (steady-state levels of 80-12 ng/ml) is superior to half-dose maintenance (40-60 ng/ml) in assuring good symptomatic control and preventing subsyndromal flare-ups (Reynolds CF, Perel JM, Frank E, Cornes C, Miller MD, Houck PR, Mazumdar S, Stack JA, Pollock BG, Dew MA, Kupfer DJ. American Journal of Psychiatry, under review).

• Despite almost identical recovery rates in 60-69 year old patients versus those aged 70 and above, we have observed an overall recurrence rate for major depression of 60.5% in the over-70 group during the first year of maintenance treatment, versus 30.4% in subjects aged 60-69. In survival analyses, higher age was significantly related to shorter time to recurrence; and recurrence rates were significantly greater in older subjects randomized to monotherapy with either nortriptyline or Interpersonal Psychotherapy, while rates of recurrence did not vary significantly as a function of age in subjects randomly assigned to combined treatment (NT + IPT). Thus, there appears to be a clinically significant advantage to combined treatment (and a corresponding disadvantage to monotherapy) in the 70+ year old subjects. The over-70 group is the group in greatest need for further evaluation on how to manage them, since two-thirds do not have a sustained and good response with nortriptyline alone or IPT alone (Reynolds CF, Frank E, Dew MA, Houck PR, Miller MD, Mazumdar S, Perel JM, Kupfer DJ. American Journal of Geriatric Psychiatry, in press).

• In a placebo-controlled study of nortriptyline (NT) and interpersonal psychotherapy (IPT) for treating bereavement-related major depression, rates of remission were as follows: 1) NT + IPT: 11/16 (69%); 2) NT + medication clinic: 14/25 (56%); 3) placebo + IPT: 5/17 (29%); and 4) placebo + medication 10/22 (45%). Thus NT is superior to placebo in achieving remission of bereavement depression. These results support the indication for pharmacologic treatment of depressive symptoms in the wake of a serious life stressor (Reynolds CF, Miller MD, Pasternak RE, Frank E, Perel JM, Cornes C, Houck PR, Mazumdar S, Dew MA, & Kupfer DJ. American Journal of Psychiatry, in press).

• In patients with ischemic heart disease and depression, paroxetine was found to be more effective and better-tolerated than nortriptyline in reducing symptoms of major depression (Roose S, Laghrissi-Thode F, Kennedy JS, Nelson JC, Bigger JT, Pollock BG, Gaffney A, Narayan M, Finkel M, McCafferty J, Gergel I. Journal of the American Medical Association, 279:287-291, 1998).

2.  Studies to broaden outcomes and enhance generalizability

• Being an Alzheimer's Disease caregiver is a risk factor for psychiatric morbidity. The probability of psychiatric illness further increases if caregiver has low income, compromised physical health, and has a small support network (Schulz R, O'Brien AT, Bookwala J, Fleissner K. Gerontologist, 35:771-791, 1995).

• Despite a high level of anxiety symptoms, comorbid anxiety disorders appear to be less prevalent among depressed elderly individuals than they are among depressed younger individuals. Since both anxiety disorders and major depression independently have been reported to be risk factors for premature death, individuals with comorbid anxiety and depression may die prematurely and be underrepresented among aging patients. (Mulsant BH, Reynolds CF, Shear MK, Sweet RA, and Miller MD. Anxiety, 2:242-247, 1996).

• The symptoms of complicated grief were found to be distinct from those of bereavement-related depression and appear to be associated with enduring functional impairments. (Prigerson HG, Frank E, Kasl SV, Reynolds CF. American Journal of Psychiatry. 152:22-30 1995).

• We examined in a sample of 130 elderly bereaved whether symptoms of complicated grief at baseline predicted suicidal ideation during a depressive episode. Fifty-seven percent of the patients with high complicated grief scores were found to be ideators versus 24% of the patients with low complicated grief scores. Thus, the condition of having high levels of complicated grief symptoms and depressive symptoms make bereaved individuals vulnerable to suicidal ideation (Szanto K, Prigerson H, Houck P, Ehrenpreis L, Reynolds CF: Suicidal ideation in elderly bereaved: The role of complicated grief. Suicide and Life-Threatening Behavior, 27:195-207, 1997).

• Symptoms of complicated grief among elderly widows and widowers did not decline significantly as time had elapsed since the death, suggesting that grief does not appear to resolve in stages. The practical implications are that those experiencing high levels of complicated grief are likely to persist at high levels, and that interventions may need to facilitate the resolution of grief which does not always appear to resolve neatly in stages (Bierhals AJ, Prigerson HG, Frank E, Reynolds CF, Fasiczka A. Omega, 32: 303-317; 1996).

3.  New approaches to treatment, rehabilitative intervention, and preventive intervention

• Continuation pharmacotherapy and psychotherapy of major depression are associated with further resolution of symptoms and improvements of function (Opdyke KS, Reynolds CF, Begley AE, Buysse DJ, Dew MA, Frank E, Mulsant BH, Shear MK, Mazumdar S, Kupfer DJ. Depression and Anxiety, 4:312-319; 1997).

• In a sample of 107 depressed elderly who were remitted from a depressive episode, we found that patients with a history of suicide attempt had significantly higher levels of hopelessness following remission than nonattempters. Our finding suggests that suicide attempts are associated with persistent high levels of hopelessness following remission of depression, thus treatment specifically designed to lower hopelessness may be effective in reducing suicide risk (Szanto K, Reynolds CF, Conwell Y, Begley AE, Houck P. Journal of the American Geriatric Society, in press).

• Because primary insomnia is a persistent and recurrent disorder, as well as a risk factor for depression, we conducted an open pilot study to determine whether paroxetine is effective in the treatment of patients with primary insomnia. Eleven of 14 patients (73%) improved with treatment and seven no longer met DSM-IV diagnostic criteria for primary insomnia after treatment. These results support the effectiveness of paroxetine in the treatment of chronic primary insomnia, but further evaluation with controlled and longitudinal designs is warranted to determine if treatment of insomnia prevents depression (Nowell PD, Reynolds CF, Buysse DJ, Dew MA, and Kupfer DJ. Journal of Clinical Psychiatry, in press).

• In a study of treatment resistance in geriatric unipolar recurrent depression, defined as failure to recover despite combined pharmacotherapy (nortriptyline) and Interpersonal Psychotherapy, we reported that 18.4% of patients either failed to remit or relapsed during continuation therapy (and therefore failed to recover) despite vigorous treatment (Little JT, Reynolds CF, Dew MA, Frank E, Begley AE, Miller MD, Cornes C, Mazumdar S, Perel JM, and Kupfer DJ. American Journal of Psychiatry, 155(8):1035-1038, 1998).

• Despite high rates of co-existing general medical illnesses in elders with major depression, the burden of such illness neither retards nor precludes the remission of depression in response to combination treatment with antidepressant medication and interpersonal psychotherapy (Miller MD, Paradis CF, Houck PR, Rifai AH, Mazumdar S, Pollock B, Perel JM, Frank E, and Reynolds CF. American Journal of Geriatric Psychiatry, 4(4):281-290, 1996).

• In a study of the onset of antidepressant activity of paroxetine, we reported that 70% of elderly patients with major depressive episodes were well by two weeks when paroxetine therapy was initiated following one night of total sleep deprivation. The combination of sleep deprivation and paroxetine appears to be useful in accelerating remission from unipolar major depression in later life and in identifying which patients are likely to be treatment-resistant (Bump GM, Reynolds CF, Smith G, Pollock BG, Dew MA, Mazumdar S, Geary M, Houck PR, and Kupfer DJ. Depression and Anxiety, 6: 113-118; 1997).

• Elderly depressed patients who require augmentation of primary pharmacotherapy (e.g., with lithium) to achieve clinical response show a high relapse rate (50%) after discontinuation of the adjunctive medication. This observation suggests that if augmentation of primary pharmacotherapy is necessary to achieve wellness, its continuation may also be necessary to preserve wellness (Reynolds CF, Frank E, Perel JM, Mazumdar S, Dew MA, Begley A, Houck PR, Hall M, Mulsant BH, Shear MK, Miller MD, Cornes C, Kupfer DJ, American Journal of Psychiatry, 152(11):1418-1422, 1996).

4.  Broadening outcomes

• The distinction between active and passive suicidal ideation should not be overdrawn in elders with recurrent major depression, since both groups are about equally likely to have past histories of suicide attempts, active and passive ideation vary interchangeably in the course of the depressive episode, and hopelessness is equally persistent in both groups and more pervasive than in non-ideators (Szanto K, Reynolds CF, et al., American Journal of Geriatric Psychiatry, 4(3): 197-207; 1997).

• For both recovered and non-recovered elderly patients with recurrent major depression, quality of life profiles (as measured by the General Life Functioning [GLF] Scale) improved during combined treatment with nortriptyline and interpersonal psychotherapy. Improvement was greater in recovered than non-recovered patients, after controlling for changes in Hamilton depression ratings, in subscales measuring coping and well-being (Mazumdar S, Reynolds CF, Houck PR, Frank E, Dew MA, Kupfer DJ. Psychiatry Research, 63: 183-190;1996).

• Very little is known about what put demented patients at risk for developing behavioral complications of their dementia. In this study, a history of major depression preceding the onset of dementia was found to increase (triple) the risk for the development of a major depressive syndrome in patients with Alzheimer's disease (Zubenko GS, Rifai AH, Mulsant BH, Sweet RA, Pasternak RE. American Journal of Geriatric Psychiatry, 4: 85-90;1996).

• In a community-based, untreated and relatively large sample, symptoms of traumatic grief form a factor which is distinct from the symptoms of bereavement-related depression and anxiety (Prigerson HG, Bierhals AJ, Kasl SV, Reynolds CF, Shear MK, Newsom JT, Jacobs S. The American Journal of Psychiatry, 153(11):1484-1486; 1996).

• The MHCRC/LLMD, in collaboration with the MHCRC for Affective Disorders (MH30915; David J. Kupfer, M.D., PI) sponsored a workshop in January, 1997, to establish a consensus on preliminary diagnostic criteria for traumatic grief. The criteria set requires the experience of intense symptoms of separation distress (e.g., yearning and searching) and includes bereavement-specific aspects of traumatic distress (e.g., feeling that a part of oneself has died, a shattered world view, and feelings of futility about the future). Preliminary ROC analyses suggested all but one of the analyzed items (experiencing symptoms of the deceased's last illness) worked well to identify bereaved individuals who met criteria for traumatic grief (Prigerson HG, Shear MK, Jacobs SC, Reynolds CF, Maciejewski PK, Davidson J, Rosencheck R, Pilkonis PA, Wortman CB, Williams JBW, Widiger TA, Frank E, Kupfer DJ, Zisook S. British Journal of Psychiatry, in press).

• The presence of traumatic grief symptomatology six months after spousal loss predicted negative health outcomes such as the incidence of cancer, heart trouble, high blood pressure, suicidal ideation, and changes in eating habits at 13 and/or 25 months (Prigerson HG, Bierhals AJ, Kasl SV, Reynolds CF III, Shear MK, Day N, Newsom JT, Jacobs S. American Journal of Psychiatry, 154(5): 616-623, 1997).

• Level of caregiver burden is significantly associated with depression and traumatic grief (Beery LC, Prigerson HG, Bierhals AJ, Santucci LM, Newsom JT, Maciejewski, P, Rapp S, Fasiczka A, Reynolds CF. Omega, 35(3): 261-279, 1997).

5. Clinical Pharmacokinetic and Pharmacodynamic Studies

7.  Psychosocial and biological correlates of treatment response variability

• To test the hypothesis that combined total sleep deprivation and antidepressant treatment would accelerate the clinical and glucose metabolic response to antidepressant treatment, six geriatric depressed patients and six age matched controls underwent serial PET studies at baseline, post-TSD, post-recovery sleep and two weeks post-paroxetine treatment (patients only). The depressed patients demonstrated persistent reductions in both Hamilton Depression Scale Scores and in glucose metabolism in the anterior cingulate gyrus and middle frontal gyrus after sleep deprivation, recovery sleep and after two weeks of antidepressant treatment. In contrast, the normal controls showed increased metabolism after sleep deprivation. This is the first in vivo neurobiologic evidence in geriatric patients that sleep deprivation produces persistent alterations in brain metabolism and that this non-pharmacologic intervention may represent a strategy to accelerate treatment response (Smith G, Reynolds CF, Pollock B, Berbyshire S, Nofzinger EA, Dew MA, Milko D, Meltzer C, Kupfer DJ. Am J Psychiatry, in press).

• In a study of the effects of lifetime age at onset of unipolar depressive illness on rates of remission, relapse, recovery, and recurrence, we reported that age of onset less than 60 versus 60 or later did not affect absolute rates of remission and recovery (during open combined treatment with nortriptyline and IPT), or rates of relapse and recurrence. However, subjects with lifetime onset of depressive illness before age 60 took on average 5-6 weeks longer to achieve remission, possibly a reflection of the greater number of prior lifetime episodes (chronicity) (Reynolds CF, Dew MA, Frank E, Begley A, Miller MD, Cornes C, Mazumdar S, Perel JM, Kupfer DJ, American Journal of Psychiatry, 155(6):795-799, 1998).

• In elders with recurrent unipolar depression, improvement in sleep quality and enhancement of rapid eye movement activity during REM sleep in response to nortriptyline predict lower likelihood of recurrent major depression during maintenance therapy with nortriptyline (Buysse DJ, Reynolds CF, Hoch CC, Houck PR, Kupfer DJ, Mazumdar S, Frank E. Neuropsychopharmacology, 14(4): 243-252; 1996).

• Among bereaved elders, mastery events, global social support, and "appraised" or "belonging" social support, in particular, significantly reduced the severity or likelihood of depression but appeared to have no significant effect on dysthymia. These protective psychosocial factors may, therefore, hold promise for interventions aimed at preventing or reducing bereavement-related depression in late life (Prigerson HG, Frank E, Reynolds CF, George CJ, Kupfer DJ. American Journal of Geriatric Psychiatry Vol 1: 296-309, 1993).

• Lifestyle regularity (i.e., temporal stability of daily social rhythms) protects against depressive symptomatology among spousally bereaved elders, provided a sufficient level of activity is maintained (Prigerson HG, Monk TH, Reynolds CF, Kupfer DJ. Depression, 3(6): 297-302, 1996).

• PET studies using the selective ligand [18F] altanserin have demonstrated a highly significant decline in specific serotonin type 2A receptor binding with age. Compared to young controls, aged 18-31, a group of healthy elderly subjects between the ages of 61 and 76 showed a nearly 60% loss of serotonin 2A receptor binding across many cortical regions. This effect persisted after correcting the PET data for partial volume effects due to age-related cerebral volume loss (Meltzer CC, Smith G, Price JC, Reynolds CF, Mathis CA, Greer PJ, Lopresti B, Mintun MA, Pollock B, Ben-Eliezer D, Cantwell M, Kaye W, DeKosky ST. Brain Research, in press).

• In a study of the temporal profiles of the course of recovery in elderly patients with recurrent unipolar major depression, 30.5% showed rapid sustained response to combined treatment with nortriptyline and Interpersonal Psychotherapy (i.e., were well by four weeks), 22.1% showed gradual sustained response (well by 8-10 weeks), 23.2% showed partial or mixed response, and 24.2% showed little or no evidence of response. Higher levels of acute and chronic stressors, poorer social supports, younger age at first depressive episode, endogenous depression, higher current anxiety, older current age, and poorer subjective and objective sleep quality predicted poorer response profiles (Dew MA, Reynolds CF, Houck PR, Hall MH, Buysse DJ, Frank E, and Kupfer DJ. Archives of General Psychiatry 54: 1016-1024, 1997).

• Severity of objective pretreatment chronic medical burden does not predict likelihood of response or rate of temporal response (Miller MD, Paradis CF, Houck PR, Rifai AH, Mazumdar S, Pollock B, Perel JM, Frank E, Reynolds CF, American Journal of Geriatric Psychiatry, 4(4):281-290; 1996).

• Platelet factor 4 (PF4) is an important marker of platelet activation and prethrombotic states. Recently, we have found significant elevations of PF4 in depressed patients with heart disease (Laghrissi-Thode F, Wagner W, Pollock BG, Johnson P, Finkel M. Biological Psychiatry, 42:290-295; 1997). This suggests at least one possible mediator of the mortality risk associated with depression in patients who have suffered an ischemic event.

• Comparing magnetic resonance imaging scans (MRIs) of older persons with schizophrenia, depression, and no psychiatric illness, deep white matter hyperintensities were greater in the schizophrenic group in the right posterior region, consistent with previous reports associating right parietal-temporal-occipital lesions with psychosis (Keshavan MS, Mulsant BH, Sweet RA, Pasternak RE, Zubenko GS, Krishnan RK. Psychiatry Research, 60: 117-123; 1996; Mulsant BH, Keshavan MS, Pollock BG. Medicine & Hygiene, 53:1567-1569; 1995).

• Self-assessed health predicts response to treatment for depression. Individuals with major depression who rated their health as fair or poor at the beginning of treatment were less likely to show improvement than individuals who rated their health as good or excellent, even after controlling for objective medical status of the patient (Miller MD, Schulz R, Paradis C, Houck PR, Mazumdar S, Frank E, Dew MA, Reynolds CF. American Journal of Psychiatry, 153(10): 1350-1352; 1996).

• Elderly patients with a brittle response to treatment are characterized by a higher burden of anxiety symptoms throughout treatment, greater subjective sleep impairment, and higher levels of rapid eye movement (REM) sleep before treatment (Reynolds CF, Frank E, Kupfer DJ, Thase ME, Perel JM, Mazumdar S, Houck PR, American Journal of Psychiatry,153(10):1288-1292; 1996). Such patients also take longer to respond to treatment initially (Mulsant BH, Reynolds CF, Shear MK, Sweet RA, Miller MD, Anxiety, 2:242-247, 1996). Hence, major depression in later life, complicated by clinically significant symptoms of anxiety, takes longer to respond to treatment, often requires augmentation pharmacotherapy, and presents a high risk for relapse.

• Successful maintenance pharmacotherapy with nortriptyline is associated with higher levels of REM activity generation and an increase in first NREM period delta activity than is seen with maintenance placebo (Reynolds CF, Buysse DJ, Brunner DJ, Dew MA, Hoch CC, Hall M, Begley AE, Houck PR, Mazumdar S, Perel JM, and Kupfer DJ. Biological Psychiatry 42: 560-567, 1997).

• Additional life events occurring after an initial provoking agent significantly alter the risk of illness onset in patients with recurrent depression. Additional severely threatening events decrease the time to onset, but positive events do not appear to delay onset (Frank E, Tu XM, Anderson B, Reynolds CF, Karp JF, Mayo A, Ritenour AM, Kupfer DJ. Psychological Medicine, 26: 613-626; 1996).

• In elders with recurrent unipolar depression, EEG sleep abnormalities are more pronounced earlier in episode than later in episode and in association with being male, being older, and having impaired social support (Dew MA, Reynolds CF, Buysse DJ, Houck PR, Hoch CC, Monk TH, Kupfer DJ. Archives of General Psychiatry, 53:148-156; 1996).

• The relationship between cognition and cerebral integrity in late-life depression (LLD): These analyses focus on the relationship between measures of cerebral integrity as measured by structural MRI and performance on selected cognitive measures--the Folstein Mini-Mental Status Exam (MMSE) and the Mattis Dementia Rating Scale (MDRS), performed at baseline, 12- and 52-weeks on 18 LLD subjects. Analyses revealed very few significant correlations between measures of cerebral atrophy (both sulcal and ventricular size measures) and test performance. However, even with this relatively small sample size (N=26), the number of white matter hyperintensities is significantly correlated with three of the six cognitive measurements, and the correlations between two of the remaining measurements are approaching significance (p<.05). Furthermore, T-Tests comparing MRI measures for those subjects classified as cognitively normal (MDRS 125, N=10) and cognitively impaired (MDRS 124, N=8) at baseline were performed. The white matter hyperintensity rating was the only MRI measure that was significantly different (p=.021) between the two groups. These preliminary results are consistent with the recently published report showing a relationship between white matter hyperintensities and cognitive functioning in LLD (Hickie et al., 1995). As our sample size increases we will further explicate this relationship through analyzing specific neuropsychological domains (Butters MA, Zmuda M, Becker JT, Nebes R, Pollock BG, Mulsant BH, Reynolds CF. Presented at the American Psychological Association Annual Convention, San Francisco, CA, August, 1998).

8. Treatment Adherence and Compliance

• Elderly depressed patients who initially rate their general health as fair to poor are less likely to recover from depression in a standardized treatment protocol. Self-ratings of health improve with resolution of depression. Self-rating of health may be related to a patient's receptivity to treatment or willingness to follow a treatment regimen (i.e., compliance) (Miller MD, Schulz R, Paradis C, Houck PR, Mazumdar S, Frank E, Dew MA, Reynolds CF, American Journal of Psychiatry, 153(10):1350-1352; 1996).

• Maximizing compliance with antidepressant treatment is critical to good patient care and to the success of the proposed studies. We have recently reported low rates of missed medication doses in our studies of maintenance treatment: zero non-compliance occurred in 43% of subjects during acute and continuation therapy, one or less missed doses/month in 50.3%, 1-2 missed doses/month in 5.6%, and two or more missed doses/month in 1.1% (Miller MD, Foglia JP, Pollock BG, Begley A, Reynolds CF. Essential Psychopharmacology, in press).

9. Critical evaluation of clinical care without a base of research support

11. Side effects

• Parkinsonian symptoms due to neuroleptic drug treatment are associated with the potential for serious adverse events in the elderly, such as urinary incontinence and falls. The risk for developing this side effect is reduced in elderly patients treated for delusional depression when compared to patients treated for delusions or hallucinations associated with Alzheimer's disease. The reasons for this difference in side effect rates may be related to the underlying neurochemistry of the two disorders, as measures of the activity of dopamine in the nervous system differ between these diagnostic groups (Sweet RA, Mulsant BH, Pollock BG, Rosen J, Altieri LP. American Journal of Geriatric Psychiatry, 4: 311-319, 1996; Sweet RA, Pollock BG, Mulsant BH, Rosen J, Branch RA. Psychopharmacology Bulletin, 31: 651-657; 1995).

• Tardive dyskinesia is a syndrome of involuntary movements occurring during neuroleptic treatment and frequently persisting after treatment has ended. In contrast to younger patients, the elderly are at an increased risk to develop tardive dyskinesia within months of total lifetime neuroleptic treatment and this risk escalates rapidly, in direct proportion to the duration of neuroleptic treatment. After accounting for the effect of lifetime duration of neuroleptic treatment, other factors, such as increasing age, gender, race, or evidence of cognitive impairment, do not contribute further to the risk for tardive dyskinesia (Sweet RA, Mulsant BH, Gupta B, Rifai AH, Pasternak RE, McEachran A, Zubenko GS. Archives of General Psychiatry, 52:478-486, 1995).

• In a placebo-controlled evaluation of nortriptyline side effects during maintenance treatment, patients maintained on nortriptyline had an excess of dry mouth and constipation, but not of weight gain (Reynolds CF, Frank E, Perel JM, Miller MD, Paradis CF, Stack JA, Pollock BG, Rifai AH, Cornes C, George CJ, Mazumdar S, Kupfer DJ. American Journal of Geriatric Psychiatry, 3: 170-175, 1995).

• Radiotracer development has focus on implementing the synthesis and developing tracer kinetic modeling methods for radiotracers for aspects of the serotonin system. Several serotonin (5-HT2A) receptor radiotracers were synthesized and evaluated in non-human primates. The synthesis of the radiotracers [18F]altanserin and [11C]MDL 100907 were implemented and the radiotracers were evaluated for subsequent use in human subjects (Mathis CA, Mahmood K, Huang Y, Simpson NR, Gerdes JM, Price JC. Med Chem Res 1996; 6:1-10.). While both radiotracers demonstrated that radiometabolites crossed the blood brain barrier, [18F]altanserin was chosen for use in human subjects and the profile of radiometabolites was extensively evaluated (Price, J., Lopresti, B., Huang, Y., Holt, D., Smith, G., Mathis, C. In: Quantitative Brain Imaging with Positron Emission Tomography, Carson RE, et al., eds, Academic Press , pp. 427-434, 1998; Lopresti, B., Holt, D., Mason, N., Huang, Y., Ruszkiewicz, J., Perevuznik, J., Price, J., Smith, G., Mathis, C. In R. Carson (ed.) Quantitative Brain Imaging with Positron Emission Tomography, Academic Press, pp. 293-298, 1998). The test-retest variability of [18F]altanserin binding was consistent with that of other radiotracers and radiotracer binding was highly correlated with the in vitro binding density of the 5-HT2A receptor (Smith G, Price J, Lopresti B., Huang Y., Simpson N., Holt, D., Mason, N.S., Sweet, R., Meltzer C.C., Nichols, T., Sashin, D., Mathis C. Synapse, 30(4), 380-392, 1998). These observations supported the application of [18F]altanserin to the study of psychiatric patients.

• An age related decline in 5-HT2A receptor binding was demonstrated using [18F]altanserin. The decrease in binding persisted after the correction of the data for the effects of cerebral atrophy (Meltzer, C., Smith, G. Price, J., Reynolds, C., Mathis, C., Greer, P., Lopresti, B., Mintun, M., Pollock, B., Ben-Eliezer, D., Cantwell, M., Kaye, W., DeKosky, S. Brain Research , in press). 5-HT2A receptor binding was decreased in AD patients compared to age-matched controls, but no differences in binding were observed in late life depressed patients. The lack of a difference in the post-synaptic component of the serotonin system has led to the application of radiotracers for pre-synaptic aspects of the serotonin system.

• A radiotracer for the serotonin transporter site ([11C]McN5652) has been synthesized and extensively evaluated (Price JC, Lopresti BJ, Huang Y, Simpson NR, Mahmood K, Mathis CA. Neuroimage, 5(4)3:A19, 1997). As this site represents the initial target of action for the selective serotonin reuptake inhibitors, the occupancy of this site by SSRIs will be studied in late life depressed patients as a potential source of treatment response variability. Studies have begun in late life depressed patients to evaluate changes in serotonin transporter density (Meltzer CC, K07 sponsored activity).

• A radiotracer for the 5-HT1A has been developed and evaluated (Price JC, Mathis CA, Simpson NR, Mahmood K, Mintun MA. In: Quantification of Brain Function using PET, Jones T, et al., eds, Academic Press, pp 257-261, 1996). Initial studies have begun in patients with midlife depression (Drevets W, Price J, Kupfer D, Holt D, Proper S, Lopresti B, Mathis C. Soc Neurosci Abstr, in press) and with late life depression (Meltzer CC, K07 sponsored activity) to evaluate changes in 5-HT1A receptor binding and to test the hypothesis that desensitization of the 5-HT1A autoreceptor is involved in the clinical response to antidepressant medications.

• In ongoing studies in the applied methodology of atrophy correction of PET data, computer simulations of PET images created from MRI data in young, elderly, and Alzheimer's disease subjects were used to demonstrate the need for MR-based partial volume correction in PET studies of aging and neurodegenerative disease. Further, the sensitivity of MR-based approaches to atrophy correction to introduced errors in image registration and segmentation, and the impact of resolution effects were quantified (Meltzer CC, Kinahan P, Nichols TE, Greer PJ, Comtat C, Cantwell M, Lin MN, Price J., J Nucl Med, submitted).

• Previous PET studies to measure dopamine activity in vivo in human subjects have used dopamine (D2) receptor radiotracers and pharmacologic challenges with psychostimulant agents. The detection of striatal binding parameter changes were investigated using the reversible very high affinity and lower affinity benzamides, [18F]fallypride and [11C]raclopride, respectively. Following d-amphetamine pretreatment and bolus radioligand injection, [18F]fallypride yielded smaller binding parameter changes than those that were observed for [11C]raclopride, while larger and more variable frontal measures were obtained using [18F]fallypride (Price JC, Mason S, Lopresti B, Holt D, Simpson NR, Drevets W, Smith GS, Mathis CA. In: Quantitative functional brain imaging with positron emission tomography (Carson RE, et al., eds), San Diego, Academic Press, 1998, pp. 441-448).

[11C]Raclopride PET studies to measure dopamine activity in vivo in baboons pre- and post-amphetamine challenge have indicated that reductions in specific binding may be detected in areas other than the D2 rich striatum. Extrastriatal measures (e.g., frontal and temporal cortices) may be possible using [11C]raclopride and 3D PET, especially for studies where large striatal changes are expected. In addition, these studies demonstrated that the mean (n=3 or 4 baboons) striatal post-amphetamine binding reduction was dose dependent; the relationship between the amphetamine dose (i.v. injection) and the corresponding change in the [11C]raclopride specific binding measure was well described by a linear relationship over the dose range studied (0.3 - 1.0 mg/kg),(Price JC, Mathis CA, Lopresti B, Holt D, Mason NS, Drevets W. Neuroimage 1998; 7(4)3:A12).

• We have developed a method for assessing the functional neuroanatomy of human sleep using PET FDG (Nofzinger EA, Mintun MA, Price J, Meltzer CC, Townsend D, Buysse DJ, Reynolds CF, Dachille M, Matzzie J, Kupfer DJ, Moore RY. Brain Research Protocols, 2:191-198, 1998).

Biostatistical Contributions

• We have developed a new method to estimate survival time distributions in clinical psychiatric research when the survival time is defined as the time from the start of the observation period to the time at the completion of successive occurrences of a response for a predetermined number of times. The method is based on transition (Markov) models (Mazumdar S, Liu K, Sang A, Houck PR and Reynolds CF. Communications in Statistic, in press).

• A program package using SAS (1) and S-PLUS (2) is presented for performing random regression residual analysis. The PROCEDURE MIXED from SAS is used for statistical inference. Both elementary-level and individual-level residuals are used. The S-PLUS programs provide: 1) a transformation to orthogonalize the elementary-level correlated residuals for standard regression residual analyses; and 2) several statistics and plots for checking model assumptions, assessing model fitting, and detecting outlying individuals. An illustrative example is provided (Mazumdar S, Begley A, Houck P, Yang Y, Reynolds C, Kupfer D, Computer Methods and Programs in Biomedicine, in press).

• Drop-out is a common phenomenon in clinical trials of drug treatments involving longitudinal assessments for a fixed duration of follow-up. The purpose of this study is to acquaint both clinicians and statisticians with recent statistical methodological advances in handling drop-outs and their usage for intent-to-treat analysis. A sensitivity analysis of month 12 outcomes consisting of monthly Hamilton depression scores to investigate the efficacy of a drug therapy is discussed. The sensitivity analysis includes endpoint analysis, last observation carried forward analysis, repeated measures models, and imputation models. Imputation models are based on multiple imputations of missing responses. Issues related to bias and efficiencies of the estimates are discussed (Mazumdar S, Liu KS, Houck PR, Reynolds CF. Journal of Psychiatry Research, in press).

Recruitment and Assessment