The Fuzzy World of Subsyndromal Depression: And Epidemiologic Challenge

1
The Fuzzy World of Subsyndromal Depression And
Epidemiologic Challenge

• Dan G. Blazer MD, PHD
• JP Gibbons Professor of Psychiatry and Behavioral
  Sciences

2
The Modern Epidemiologic Assumptions

• Human pathophysiology should be studied as
  discrete entities - diseases.
• The phenotypic expressions of these discrete
  entities represent underlying discrete
  pathophysiological processes.
• These pathophysiological processes result from
  the interaction of the genetic make-up of the
  individual with specific environmental challenges
  or support to the individual.

3
The Modern Epidemiologic Assumptions

• The study of the causes of disease has shifted
  away from the environment as a whole to specific
  factors within the environment (e.g. biological
  organisms) and to the behaviors of individuals
  (e.g. smoking).
• All variables are thus best measured at the
  individual level for it is the individual who is
  truly important in the causation of disease
• Diez - Roux, 1998

4
The Modern Epidemiologic Assumptions (cont.)

• Phenylketoneuria (PKU) represents the classic
  example of this genetic/environmental
  interaction.
• Treatment of disease consists of specific
  environmental interventions, such as the change
  of a subjects diet or the prescription of a
  specific drug.
• Some have labeled this view as methodological
  individualism.

5
The Modern Epidemiologic Conclusion

• A combination of symptoms, signs, clinical
  course, family history, biological markers and
  response to treatment (?) will enable the
  epidemiologist to develop the criteria for
  identifying a case of the discrete entity (the
  disease)
• Methods will be established which will become the
  gold standard for identifying the case.
• Screening methods will emerge which can be tested
  for sensitivity and specificity

6
The Modern Epidemiologic Conclusion

• Risk should be individualized. Risk is
  individually determined rather than socially
  determined. (e.g. stressful life events)
• Lifestyle and behaviors are matters of free
  individual choice.
• Therefore facts about society and social
  phenomenon are to be explained solely in terms of
  facts about the individual.
• Duncan et al, 1996 Lukes, 1970 Diez-Roux, 1998

7
The Case for Subsyndromal or Minor Depression
8
The Case for Subsyndromal Depression - Clinical
Experience

• Persons are receiving treatment for depression
  which does not meet criteria for major depression
  in primary care
• Primary care physicians see much more in the way
  of subthreshold conditions, whereas specialty
  clinicians see the more severe end of the
  spectrum. This leads to varying views regarding
  the prevalence of depression across the life
  cycle.
• Pincus et al, 1999

9
The Case for Subsyndromal Depression - Prevalence
Studies
Mean CES-D Scores (modified) by Age in the Duke
EPESE sample in 1986-87Blazer et al, 1991
10
The Case for Subsyndromal Depression - Prevalence
Studies

• Many depressive symptoms are not captured by DSM
  in community based epidemiologic surveys
• Minimal or no symptoms - 75
• Dysphoric symptoms - 19
• Symptomatic (minor) Depression - 4
• Mixed depression/anxiety - 1.2
• Dysthymia - 2.1
• Major Depression - 0.7
• Blazer et al, 1987

11
The Case for Subsyndromal Depression Outcome
Studies
Wave I
Minor Depression with Mood Disturbance
Minor Depression without Mood Disturbance
Major Depression
Wave II
Asymptomatic 35.4
37.2 65.1 Major Depression 23.7
10.3 1.8 Dysthymia 2.6
2.4 2.0 Minor Depression without mood
disturbance 17.6 16.0
5.6 Minor depression with mood disturbance
20.8 34.2 25.5
Broadhead et al., 1993
12
The Case for Subsyndromal Depression Case
Identification

• Of five pure types in grade of membership
  analysis (GOM), one approximated major depression
  and older persons loaded on this pure type.
• Symptoms which loaded included depressed mood,
  decreased appetite, psychomotor retardation
  anxiety and memory loss.
• There was a smooth distribution of subjects who
  loaded upon this pure type. Blazer et al, 1988

13
The Case for Subsyndromal Depression Case
Identification

• Among persons studied in the ECA survey, more
  than 50 of cases of first onset major depression
  in the community were associated with prior
  depressive symptoms Horwarth, 1992
• Many persons only experience partial recovery
  from major depression.
• Keller et al, 1981 Angst and Merikangas, 1997

14
The Case for Subsyndromal Depression - Family
History

• In a study of 1420 subjects with subsyndromal
  depression compared to hypertensives and major
  depression, family history of 41 in subsyndromal
  group compared to 59 in major depressive group
  and higher than among hypertensives for both
  depressive groups.
• Shelbourne, 1994

15
The Case for Subsyndromal DepressionRisk Factor
Profiles

• Subsyndromal depression and major depression
  associated with functional impairment, financial
  impairment, bed days, high levels of functional
  strain and limitations in job functioning.
• Conclusion - subsyndromal depression is a
  clinically significant variant of unipolar major
  depression
• Judd et al, 1996

16
The Case for Subsyndromal DepressionRisk Factor
Profiles

• In a community based survey of older adults, the
  prevalence of CES-D was 9.1 and the prevalence
  of subthreshold depression was 9.9. In ordinal
  logistic regression, both were associated with
  impaired physical functioning, disability days,
  poorer self-rated health, use of psychotropic
  medications, perceived low social support, female
  gender and being unmarried.
• Hybels et al, 2001

17
The Case for Subsyndromal Depression - Treatment
Studies

• Pharmacologic therapy is effective for treating
  minor depression
• Paroxetine was superior to placebo in treating
  415 primary care patients experiencing minor
  depression and dysthymia in a clinical trial
  (HSCL-D-20 MOS Short-Form 36 HDRS).
• Williams et al, JAMA, 2000

18
The Emergence of Subsyndromal Depression
19
DSM-IV Criteria for Minor Depressive Disorder
(Appendix)

• Depressed moon or loss of interest/pleasure.
• Other symptoms may include sleep disturbance,
  weight loss, agitation or retardation, fatigue,
  feelings of worthlessness, decreased ability to
  concentrate
• At least two weeks duration
• Cause clinically significant distress

20
The Frequency of Minor Depression in Late Life in
the Community

• 4 - 8 using the DIS - some functional impairment
  (Blazer et al, 1987)
• 14.6 using the DIS - two or more depressive
  symptoms (Judd et al, 1994)
• 11 using the CES-D (Kennedy 1990)
• 12.9 using the CES-D (Beekman et al, 1995)
• 8.3 using the GMS/AGECAT (Copeland et al, 1987)

21
Prevalence () of Minor Depression by Age and
Gender (Beekman et al, 1995)
22
Prevalence Studies in Inpatient Settings

• Koenig et al, 1988 - 11.5 of hospitalized
  elderly diagnosed with major depression. 23 had
  clinically significant depressive symptoms.
• ORiordan et al, 1989 - 23 of patients admitted
  to an acute medical geriatric assessment unit
  screened positive for depression, 10.8 had
  comorbid depression/dementia and 13.5 were
  judged to need antidepressant medication.

23
Prevalence in Outpatient Settings

• 20.2 using RDC (Oxmam et al, 1990)

24
Prevalence Studies in Long-Term Care

• Parmelee et al, 1989 - 12.4 of subjects met
  criteria for MDE. 30.5 reported less severe but
  clinically significant depressive symptoms.
• Ames, 1990 - 24 of residents in homes for the
  elderly screened positive for depression. 12 had
  evidence of a mood disorder and 8 had comorbid
  depression/dementia. At one year, 25 had died
  and 28 had recovered.

25
Risk Factors for Major Depression, Minor
Depression, and Dysthymia in Late Life
Major Depression
Minor Depression
Dysthymia
Not married Female gender Younger age Low
SES Cognitive impairment Comorbid
anxiety Internal locus of control Loneliness Funct
ional impairment Beekman et al., in press
Perceived poor health Functional
limitations Loneliness Internal locus on
control Not/no longer married History of major
depression Cognitive impairment Functional
impairment Stressful life events (Beekman et al.,
in press Blazer et al., 1991
No gender difference Stressful life
events Comorbid disorders less common Devenand
et al., 1994
26
Proportion of Elderly Community Sample followed
for 10 Years Taking Different Categories of
Antidepressant Medications
Blazer et al, 2000
27
Conclusion

• Subsyndromal or minor Depression has been reified
  by clinicians as an entity.
• Therefore

28
The Resulting Epidemiologic Questions

• How can we better develop criteria for a case of
  subsyndromal depression?
• What is the frequency and distribution of
  subsyndromal depression?
• What are the correlates ( individual risk
  factors) of subsyndromal depression?
• What is the treatment of subsyndromal depression?
• Caveat - Subsyndromal depression may be a variant
  of unipolar depressive disorder

29
Has the research agenda therefore been set in
stone for subsyndromal or minor depression?
30
The Case Against Subsyndromal Depression as an
Entity (a thing)
31
The Case Against Subsyndromal or Minor Depression

• the authors ...want to apply their medical
  interpretations and their pharmacological
  treatment across the board, beyond the so-called
  clinically depressed ...to those who are unhappy
  without apparent reason, the theory being that
  these conditions i.e. minor depressions
  negatively affect quality of life and are
  associated with increased risk of comorbid
  medical illness and clinical depression.on the
  other hand, a depressive reaction to life

32
The Case Against Subsyndromal or Minor Depression
(cont.)

• experience is one thing, and vulnerability to a
  diagnosable disease called depression is
  anotherconsider depression as a personality
  train, a tendency to experience feelings which
  varies in strength from person to person. The
  disposition is not pathological but normally
  distributed, stable personality trait that
  neither increases nor declines with age.
• Stanley Jacobson, Atlantic Monthly , April, 1995,
  pp 46-51 (in response to a consensus statement
  regarding minor depression in the elderly)

33
What is a Case of Subsyndromal Depression?
34
Research Diagnostic Criteria for Minor Depression

• An Episode with relatively persistent depressed
  mood.
• Two or more criteria symptoms, such as poor
  appetite or sleep difficulty
• Duration of at least one week
• May be superimposed on another disorder such as
  alcoholism
• Must result in impairment and/or use of health
  services

35
ICD-10 Proposed Criteria for Mild Depression

• Lowering of mood, reduction of energy and
  decreased activity
• Self-esteem reduced and ideas of guilt and
  worthlessness.
• Biological symptoms mild or absent
• Causes distress and interference with normal
  activity
• Duration of at least two weeks

36
Examples of Other Operational Definitions Used
in Research Studies

• Two or more current depressive symptoms lasting
  for at least two weeks excluding major
  depression. (Judd et al, 1994, Kessler et al,
  1997)
• A score of gt15 on the CES-D but not meeting
  criteria for major depression. (Beekman et al,
  1997)
• Scores of 12 -15 on the CES-D (Hybels, et al,
  2000)

37
Snaith Criteria (1987)

• Snaith proposes a biogenic from of mild
  depression. Anhedonia is the central and
  reliable symptm of hypomelancholia (or mild
  biogenic depression

38
Differences Between Community Based and Clinic
Based Cases (unpublished data)

• 19 subjects who met CIDI criteria for major
  depression were assessed by clinical examination.
  80 were determined to meet criteria following
  the clinical examination.
• These 19 subjects were further evaluated for
  dysfunction and health service use. None
  reported work days missed during the episode nor
  other significant physical or social impairment.
  All had recovered from the episode within one
  month. None sought professional consultation for
  the episode.
• Blazer, Kessler and Swartz (unpublished data)

39
What is a Case of Subsyndromal Depression?

• Except for the fact that the symptoms are less
  severe than major depression yet can be
  disabling, we dont know the answer to this
  question.
• We can operationalize criteria, yet no one set of
  operational criteria appears to trump the others.

40
What is the Frequency and Distribution of
Subsyndromal Depression?
41
The Epidemiologic Quagmire of Subsyndromal
Depression

• Community prevalence of 2.2 (Skodol et al, 1994)
• mD without mood disturbance of 23.4 (Broadhead
  et al, 1990)
• Depressive symptom community prevalence of 23.1
  (Johnson et al, 1992)
• Episodic mD community prevalence of 52.6 of
  elderly patients (Oxman et al, 1990)

42
What is the Frequency and Distribution of
Subsyndromal Depression?

• If we cannot agree upon a definition of a case,
  we cannot determine the frequency and
  distribution of subsyndromal depression.

43
What are the Risk Factors for Subsyndromal
Depression?

• All the risk factors for major depression and
  more.

44
What is the Outcome of Subsyndromal Depression?
45
The Outcome of Subsyndromal Depression

• In a longitudinal study over 15 years of young
  adults, few subjects with depression meet the
  criteria for only one depressive subtype.
• One third of the subjects eventually develop a
  major depressive disorder (MDD).
• One-half of persons with MDD meet criteria for
  subsyndromal depression at follow-up. (Angst and
  Merikangas, 1997)
• Most cases do not evolve into a clearly defined
  entity

46
Odds of Mortality in Females in Controlled
Analyses
Hybels et al, in preparation
47
What are We Treating with What?

• the current antidepressants SSRIs are at
  present all but misbranded as antidepressants.
  They are effective for a wide range of neurotic
  conditions. Klines term, psychic energizer
  seems much more appropriate (David Healey The
  Antidepressant Era, 1997)
• Are we treating symptoms not fully explained
  with tonics and energizers or symptoms of a
  specific disorder with a specific, targeted
  therapy?

48
Are we asking the wrong questions? Are we
looking in the wrong place?
49
A Brief History of the Diagnosis of Depression
50
A Brief History of the Diagnosis of Depression

• Melancholia and underactive madness (from
  Hippocrates, the two sides of the maniac, the
  wholly mad person)
• Religious melancholia (1650 - 1800) - sickness of
  the soul (the entire soul)
• Lypemania (Esquirol, 1838, a partial insanity
  dominated by sadness, a specific disorder)
• Manic Depressive psychoses (from Kraepelin, 1899,
  one of the two forms of mental illness)

51
A Brief History of the Diagnosis of Depression

• Depression and the depressive neuroses as
  distinct from melancholia, was introduced by
  Adolf Meyer (early 1900s), a depression of mental
  energies. Neurosis derived from the late 18th
  century to refer to a presumed disorder of the
  nerves. Meyer distinguished a constitutional
  depression (pessimistic temperament), simple
  melancholic (much like our major depression) and
  other forms characterized by neurasthenic malaise
  and hypochondriacal complaints.

52
A Brief History of the Diagnosis of Depression

• Depressive psychoneuroses distinguishes
  melancholia from mourning (Freud, 1917, the
  neurotic variant of a normal adaptation to a
  stressful event, a psychoneurisis - the rigid
  distinction between personalities or
  constitutions and diseases was not drawn)
• Endogenous (autonomous) and reactive depression
  distinguished (Mobius, 1893 Gillespie, 1929)

53
A Brief History of the Diagnosis of Depression

• Major affective disorders (involutional
  melancholia and manic-depressive illness)
  distinguished from depressive neuroses in DSM II
  (1968)
• Major Depression (Feighner, 1972 DSM -III, 1980)
  -
• The ECA Epidemiologic gap and Depression NOS
  (Myers et al, 1984)
• Minor Depression (Broadhead et al, 1990)

54
A Brief History of the Diagnosis of Depression

• The evolution of the diagnosis of depression,
  especially over the past 30 years, has
  contributed in part to the emergence of the
  diagnosis of minor or subsyndromal depression.

55
A Brief History of Unexplained Psychiatric and
Medical Symptoms
56
A Brief History of Unexplained Psychiatric and
Medical Symptoms

• War syndromes (Hyams, 1998)
• Da Costras irritable heart syndrome (Civil War)
  - shortness of breath, palpitations, chest pain,
  fatigability, headache, diarrhea, dizziness and
  disturbed sleep
• The Effort Syndrome (World War I) - fatigue,
  headache, dizziness, confusion, concentration
  problems, forgetfulness, nightmares

57
A Brief History of Unexplained Psychiatric and
Medical Symptoms

• War Syndromes
• Battle Fatigue (World War II) - fatigue,
  palpitations, diarrhea, headache, impaired
  concentration, forgetfulness, and disturbed
  sleep.
• Gulf War Syndrome (Persian Gulf War) - sleep
  disturbances, impaired concentration,
  forgetfulness, irritability, muscle and joint
  pain, and depression

58
A Brief History of Unexplained Psychiatric and
Medical Symptoms

• Other syndromes
• Neurasthenia (1870s to 1880s) - anxiety, chronic
  disposition to irritability, fatigue (especially
  mental fatigue), lethargy, exhaustion
• Hysteroid dysphoria - histrionic patients with
  chronic dispositions to depression. Impaired
  anticipatory pleasure, what appears to be
  character pathology is secondary to a biological
  disturbance

59
A Brief History of Unexplained Psychiatric and
Medical Symptoms

• Demoralization (Frank, 1973 Dohrenwend, 1980) -
  poor self-esteem, helplessness-hopelessness,
  dread, sadness, anxiety, confused thinking,
  psychophysiologic symptoms, perceived poor
  physical health

60
Common Symptoms Across Multiple Syndromes

• Depression
• Anxiety (agitation)
• Sleep disturbance
• Psychophysiological complaints (or medically
  unexplained physical symptoms)
• Problems with concentration
• Fatigue
• DSM-IV Criteria for Minor Depressive Disorder

61
The Past and Present History of Subsyndromal
Depression

• What we currently diagnose as minor or
  subsyndromal depression probably was captured in
  a number of diagnostic categories in the past,
  most of which were considered a response to
  general environmental stressors or overwhelming
  specific stressors, such as war.
• The fact that we label these symptoms minor or
  subsyndromal depression shapes both our view of
  the origin of the symptoms and their treatment.

62
If we have identified a non-specific symptom
complex, what can we learn about the etiology?

• Epidemiology has become excessively concerned
  with individual risks and inadequately engaged
  with the social production of disease. Smith,
  2001
• This sounds very much like a message from
  psychiatric epidemiologys past - Stirling County

63
Multilevel or Contextual Analysis

• Lives of individuals are affected not only by
  their personal characteristics but also by
  characteristics of the social groups to which
  they belong.
• The proposal has been made that, in developing
  causal models, we should include group- or macro-
  level variables along with individual-level
  variables in public health research. Dies-Roux,
  1998

64
Multilevel or Contextual Analysis

• Variables that reflect characteristics of groups
  can be either
• Derived or aggregate variables (also contextual),
  that is, summarized characteristics of
  individuals in groups such as average income in a
  neighborhood.
• Integral variables are characteristics of the
  group not derived from characteristics of its
  members, such as availability of health care
• Dies-Roux, 1998

65
Conclusions and Implications

• Psychiatric epidemiology should for the time
  being abandon its assumption that there is a
  specific disease subsyndromal depression (or a
  variant of major depression) and take an honest,
  empirical view of our data regarding
  subsyndromal symptoms
• We dont have to name everything!

66
Conclusions and Implications

• A focus upon specific symptoms (such as sleep
  disturbance) or small clusters of symptoms (such
  as the melancholic symptoms of depression) with
  the use of cluster and factor analytic studies
  should assist psychiatric epidemiology to focus
  down upon manageable (though perhaps not all
  inclusive) syndromes for future studies.

67
Conclusions and Implications

• Psychiatric epidemiology should take full
  advantage of the rich data sets available, such
  as the ECA and NCS, to further study more focused
  groups of symptoms.
• Psychiatric epidemiology might do well to revisit
  novel groupings of symptoms, such as Dohrenwends
  demoralization but again we must take care not
  to reify such groupings prematurely.

68
Conclusions and Implications

• Psychiatric epidemiology should consider bringing
  context back into its studies, perhaps through
  multilevel analysis, again taking advantage of
  existing data sets for preliminary analyses.
• Psychiatric epidemiology should be more sanguine
  regarding the nature of human nature and avoid
  the modern day myth that happiness is the natural
  state of our species
• We dont have to explain everything!

69
Conclusions and Implications

• The need to diagnose and treat specific
  disorders, which dominates clinical medicine (and
  psychiatry) currently, should not unduly
  influence our explorations of emotional suffering
  in the community.
• Nevertheless, we must never take lightly the
  reality of emotional suffering among the
  depressed.