|

Received: 15 February 2017    Accepted: 6 April 2017

DOI: 10.1111/bdi.12499

BRIEF REPORT

Is a delay in the diagnosis of bipolar disorder inevitable?

Kristina Fritz1,2,3  | Alex M T Russell4 | Christine Allwang5 | Sandy Kuiper1,2,3 | 

Lisa Lampe1,2,3 | Gin S Malhi1,2,3

1
Academic Department of
Psychiatry, Northern Sydney Local Health
District, St Leonards, NSW, Australia
2
Sydney Medical School Northern, University
of Sydney, Sydney, NSW, Australia
3 BD, the research on this delay in diagnosis is sparse. Therefore, the present study ex-
CADE Clinic, Royal North Shore
Hospital, Northern Sydney Local Health
District, St Leonards, NSW, Australia
4
School of Health, Medical and Applied
Sciences, CQUniversity, Sydney, NSW,
Australia
5
Klinikum Rechts der Isar, München, Germany
Correspondence
Gin S Malhi, Department of Academic
Psychiatry, Royal North Shore Hospital, St
Leonard’s, NSW, Sydney.
Email: Gin.malhi@sydney.edu.au
Objective: A diagnosis of bipolar disorder (BD) is often preceded by an initial diagnosis
of depression, creating a delay in the accurate diagnosis and treatment of BD. Although
previous research has focused on predictors of a diagnosis change from depression to
amined the time taken to make a BD diagnosis following an initial diagnosis of major
depressive disorder in order to further understand the patient characteristics and psy-
chological factors that may explain this delay.
Method: A total of 382 patients underwent a clinical evaluation by a psychiatrist and
completed a series of questionnaires.
Results: Ninety patients were initially diagnosed with depression with a later diagnosis
of BD, with a mean delay in diagnostic conversion of 8.74 years. These patients who
were later diagnosed with BD were, on average, diagnosed with depression at a
younger age, experienced more manic symptoms, and had a more open personality
style and better coping skills. Cox regressions showed that depressed patients with
diagnoses that eventually converted to BD had been diagnosed with depression ear-
lier and that this was related to a longer delay to conversion and greater likelihood of
dysfunctional attitudes.
Conclusion: The findings from the present study suggested that an earlier diagnosis of
depression is related to experiencing a longer delay in conversion to BD. The clinical
implications of this are briefly discussed, with a view to reducing the seemingly inevi-
table delay in the diagnosis of BD.

KEYWORDS
bipolar disorder, delay in diagnosis, depression, diagnosis

1 | INTRODUCTION treatment is often not prescribed for almost a decade. Therefore, it is

The natural history of bipolar disorder (BD) is that it presents with de-
pression, so initial episodes are invariably diagnosed as major depres-
sive disorder (MDD). However, once manic symptoms are evident, the
diagnosis “converts” to BD. This means that there is often a delay in
the accurate diagnosis of BD, and any significant delay has an impact
on the commencement of appropriate treatment. In some instances,
the treatment of MDD with antidepressants may contribute to the de-
velopment of symptoms of BD.
Previous research has shown that the mean time from a diagnosis
of MDD to BD is just under 10 years,1–4 which means that optimal
imperative to have insight into not only what predicts the conversion,
but also what affects the delay from an MDD to a BD diagnosis.
Antidepressant treatment resistance is one of the most com-
mon predictors of diagnostic conversion from MDD to BD,5–8 with
Bukh and colleagues5 reporting a more than twofold increase in
the rate of diagnostic conversion to BD after failure to respond to
two initial antidepressant treatments. Another common risk factor
for diagnostic modification from MDD to BD is the early onset of
depression.9–12 Dudek and colleagues6 reported that patients with
an onset of depression before the age of 30 years had conversion
rates to BD more than double those of patients with a later onset.

396  |  © 2017 John Wiley & Sons A/S. wileyonlinelibrary.com/journal/bdi Bipolar Disorders. 2017;19:396–400.
Published by John Wiley & Sons Ltd
FRITZ et al.
Lastly, conversion to BD has been found to be related to a family
history of affective disorders but this has not been a consistent
finding.5,9,11–13
These studies suggest that patients who have been diagnosed with
MDD are more likely to be subsequently diagnosed with BD if they
had an early onset of depression, are unresponsive to antidepressant
treatment, and have a family history of affective disorders. Although
these studies assist in identifying a future BD diagnosis after an MDD
diagnosis, they only touch on illness characteristics and have over-
looked other possible influences, such as psychological factors and
patient characteristics.
Beyond predicting the diagnostic conversion from MDD to BD, a
relatively unexplored area is the period of delay (i.e., the time interval
between the diagnosis of MDD and that of BD). To date, the only vari-
able that has been examined is the age of onset of depression. Dudek
et al.6 found a negative correlation between the age at illness onset
and the time to diagnostic conversion – in other words, the younger
the age at onset, the greater the delay.
Therefore, the present study examined the time taken to make a
BD diagnosis following an initial diagnosis of MDD, in order to further
understand the patient characteristics and psychological factors that
may explain this delay.
2 | METHOD
A total of 382 adults with mood disorders (other DSM-­IV Axis I/II dis-
orders were excluded), who were referred to the Clinical Assessment
Diagnostic Evaluation (CADE) Clinic (www.cadeclinic.com), were
examined. This clinic is an outpatient service that is based within a
university teaching hospital in Sydney, Australia. It provides psy-
chological and psychiatric assessment of patients for the purpose of
clarifying diagnosis and offering recommendations for the treatment
of mood disorders. On arrival at the clinic, and after completing con-
sent forms, participants underwent a self-­report structured diagnostic
assessment and clinical evaluation by a psychiatrist. Diagnostic and
management questions were discussed in a multidisciplinary team
meeting, which included two to three psychiatrists and two to three
clinical and research psychologists. Prior to their assessment, partici-
pants completed a series of questionnaires, including the State–Trait
Anxiety Inventory (STAI),14 Rosenberg Self-­Esteem Scale,15 Mood
Disorders Questionnaire (MDQ),16 Anxiety Sensitivity Index-­3 (ASI-­
3),17 COPE Inventory,18 Dysfunctional Attitudes Scale (DAS),19 and
NEO Personality Inventory [NEO-­FFI].20 Additionally, patients were
asked to report the age at which they were first diagnosed with MDD,
BD, or both.
2.1 | Statistics
Analyses were carried out in SPSS 22 for Windows (IBM Corp.,
Armonk, NY, USA). Participants who converted to BD were compared
with those who did not, using an independent samples t test for con-
tinuous variables and chi-­square tests for categorical measures.
|
      397
Survival was examined using Cox proportional hazards regression,
with survival time reflecting the outcome of time to BD diagnosis fol-
lowing a diagnosis of MDD. Participants were censored on loss to fol-
low-­up, dropout, or end of study. Bivariate correlations are typically
analyzed before multivariate regression models; however, due to the
censored nature of some of the data, these were not considered to be
appropriate. Instead, individual Cox regressions were conducted for
each possible independent variable, to determine which independent
variables were related to survival time. All significant independent
variables were then entered into a multivariate model, which took into
account any overlap between independent variables. Hazard ratios, as
well as 95% confidence intervals (CIs), are reported for all significant
predictors.
Hazard ratios greater than 1 are interpreted as indicating that
higher scores in that independent variable are associated with a
shorter survival time or, more specifically, a shorter period of time until
the event. With respect to the present study, this means a shorter
conversion time to a BD diagnosis after the initial MDD diagnosis.
Conversely, hazard ratios less than 1 indicate a longer survival time.
All reasonable measures were taken to ensure that all participants
completed the full assessment but not all participants were able to
complete it in its entirety. This was because some patients found the
questionnaires too stressful and/or had difficulty in concentrating.
Therefore, listwise deletion was used for missing data.
3 | RESULTS
3.1 | Preliminary analyses
The age at initial diagnosis of MDD ranged from 10 to 78 years, with a
mean of 26.46 (standard deviation [SD]=11.53) years. Of the 382 par-
ticipants, a diagnostic conversion from MDD to BD had been made in
90 (26%) participants. The mean delay in diagnostic conversion from
MDD to BD was 8.74 years (SD=9.40), with a range of less than 1 year
to 45 years. Figure 1 shows that a diagnostic conversion to BD is most
likely to occur within the first decade after an initial diagnosis of MDD.
Table 1 shows the current pharmacological treatment for participants,
based on their diagnosis.
3.2 | MDD vs BD
Those who eventually developed BD were diagnosed with depression
at a younger age. Further, patients who developed BD experienced
more manic symptoms (according to the MDQ), were more sensitive
to anxiety and related symptoms (according to ASI-3), had a more
open personality style (according to the NEO-­O), and better coping
skills (according to the COPE Inventory) compared with patients who
were not diagnosed with BD following MDD (Table 2).
3.3 | Delay in diagnosis of BD
There was a negative association between the age at initial MDD diag-
nosis and the delay in BD diagnosis, as shown in Figure 2. A younger
 |
398       FRITZ et al.

50
Gender
Male
Female
Male

Delay in diagnostic conversion (years)

40 Female

30

20

10

10 20 30 40 50 60
F I G U R E   1   Schematic of delay in diagnostic assignment of bipolar
Age at initial diagnosis of depression
disorder (BD). The delay in BD diagnosis is divided into groups (blocks
of years as indicated; lowest bar [green]=<1 year, followed by 10-­year
F I G U R E   2   Scatterplot depicting the relationship between the
increments: 1–10 [purple], 11–20 [blue] and 21+ [red]). The width of
age at initial diagnosis and the delay in diagnostic conversion from
each bar (along the x-­axis) represents the number of patients in that
major depressive disorder to bipolar disorder (in years), separated by
group [Colour figure can be viewed at wileyonlinelibrary.com]
gender. The relationship is more pronounced for males than females
[Colour figure can be viewed at wileyonlinelibrary.com]

T A B L E   1   Current pharmacological treatment based on diagnosis

T A B L E   3   Simple Cox proportional hazards ratio estimates for the
Medication class MDD BD association of patient characteristics with the diagnosis of bipolar
disorder
Antidepressant 75% 67%
Mood stabilizer 11% 48% Variable Hazard ratio 95% CI P
Antipsychotic agent 21% 44% Gender 0.699 0.541–0.903 .006

BD, bipolar disorder; MDD, major depressive disorder. Age diagnosed 1.028 1.017–1.040 <.001
with MDD
age at initial diagnosis was associated with a longer delay in BD diag- Trait anxiety 1.015 1.003–1.026 .01
nosis. This relationship was especially pronounced for males (r=−.31) Self-­esteem 0.976 0.955–0.998 .04
compared with females (r=−.13). Dysfunctional 1.004 1.001–1.006 .006
attitudes

3.4 | Cox proportional hazards regressions CI, confidence interval; MDD, major depressive disorder.

The results from the individual Cox regressions are displayed in the age at initial MDD diagnosis, the time to BD conversion decreased
Table 3. by 2.8%. For every one-­unit increase in the STAI score, the time to
The conversion time from an MDD to a BD diagnosis was 42.8% BD conversion decreased by 1.3%, while the time to BD conversion
shorter for females compared with males. For every 1-­year increase in increased by 2.4% for every one-­unit increase in self-­esteem. Further,

T A B L E   2   Mean differences between

MDD MDD-BD Statistic P
major depressive disorder (MDD) and
Gender (% female) 58% 64% delayed bipolar disorder (BD) groups on
Age diagnosed with mood 27.38 (12.29) 23.63 (9.88) t (193)=2.88 .004 psychological variables
disorder
Number of manic 6.25 (4.11) 10.48 (2.98) t (152)=8.61 <.001
symptoms (MDQ)a
Anxiety Sensitivity Index 26.16 (14.85) 31.85 (16.65) t (222)=2.44 .015
NEO Openness 29.12 (6.43) 31.53 (6.78) t (290)=2.79 .006
Coping skills 128.39 (17.24) 139.64 (21.21) t (174)=3.65 <.001
a
Maximum score of 13.
MDQ, Mood Disorders Questionnaire; NEO, NEO-FFI Openness Subscale.
FRITZ et al.
for every one-­unit increase in DAS scores, the time to BD conversion
decreased by 0.4%. We note that these variables are on different scales
and thus the percentages are not necessarily directly comparable.
The significant univariate predictors were entered into a multi-
variate model, which accounts for any overlap between predictors
(Table 4). The age at initial diagnosis remained significant (hazard
ratio=1.04, 95% CI: 1.03–1.05; P<.001) when controlling for the other
variables in the model, indicating that among those of the same gen-
der and with the same levels of trait anxiety and self-­esteem, every
additional year of age at first diagnosis of MDD predicted a 3.8% in-
creased chance of a conversion from an MDD to BD diagnosis, and a
shorter time between the diagnosis of MDD and of BD. Further, dys-
functional attitudes also remained significant (hazard ratio=1.00, 95%
CI: 1.00–1.01; P<.05) when controlling for the other variables, signify-
ing that every one-­unit increase in dysfunctional attitudes predicted a
.4% increased chance of a conversion from an MDD to a BD diagnosis,
and a shorter conversion time.
4 | DISCUSSION
Almost a quarter of our patients (23.5%) who were initially diagnosed
with MDD had their diagnosis modified to BD, supporting previous
work that indicates that a high percentage of patients who are initially
diagnosed with MDD are eventually rediagnosed (converted) with
BD. Although some authors argue that almost half the patients with
BD (40%) are not diagnosed correctly at the time of initial presenta-
tion,6,16,21 our study found a lower percentage, which was probably
a result of the nature of the sample, which comprised an outpatient
sample that presented mainly for tertiary advice, indicating perhaps a
greater degree of complexity in pathogenesis and treatment response.
The present study suggested that patients whose diagnosis was
recalibrated to BD differed from those who remained with an MDD di-
agnosis in the age at initial mood disorder diagnosis, number of manic
symptoms experienced, anxiety sensitivity, coping skills, and personality
openness. Patients whose diagnosis was converted to BD were diag-
nosed with depression at a younger age, consistent with the key results
of the Zurich study9 and the Polish TRES-­DEP study.22 Further, the
Polish DEP-­BI study found that a depressive episode prior to the age of
25 years nearly triples the likelihood of a subsequent diagnosis of BD.23
T A B L E   4   Multivariate Cox proportional hazards ratio estimates
for the association of patient characteristics with the diagnosis of
bipolar disorder (N=279)
Variable Hazard ratio 95% CI P
Gender 0.901 0.667–1.218 .498
Age diagnosed 1.038 1.025–1.051 <.001
with MDD
Trait anxiety 1.003 0.986–1.022 .710
Self-­esteem 0.982 0.949–1.017 .305
Dysfunctional 1.004 1.000–1.008 .027
attitudes
CI, confidence interval; MDD, major depressive disorder.
|
      399
Although an early onset and diagnosis of depression is likely to be
driven by biology, it does not equate to a faster conversion to BD, sug-
gesting that other factors probably come into play. The Polish TRES-­
DEP study found that higher scores on the Hypomania Checklist-­32
and the MDQ were associated with an earlier onset of depressive ep-
isodes and a diagnosis of BD.22 The latter was corroborated by the
findings of our study.
The patients in the present study experienced more manic symp-
toms, which may have increased the likelihood of detection and the
subsequent conversion to a diagnosis of BD. These patients also suf-
fered from more anxiety symptoms, which is likely to have complicated
their presentation further, making it more difficult to confirm the cor-
rect diagnosis. However, anxiety is also likely to prompt an individual
to seek help, potentially increasing the patient’s chance of an earlier
diagnosis. The results showed that these patients also have a more
open personality style and better coping skills, protective factors that
may potentially lead them to be more likely to seek help,24–26 be more
accepting of their illness, and have a better chance of coping – two
variables that have not been explored previously. Therefore, patients
whose diagnosis has been revised to BD are not only diagnosed with
depression at an earlier age and experience more manic symptoms and
anxiety sensitivity, but also possess resilience factors, such as an open
personality style and better coping skills – two psychological variables
that are beneficial in patients with mental illness, and potentially assist
in delaying the onset, and thereby the diagnosis, of BD.
The main aim of the present study was to examine how patient
characteristics and psychological factors affect the delay in the rec-
ognition and diagnosis of BD in patients who have previously been
diagnosed with MDD. It showed that, in patients whose diagnosis is
eventually modified to BD, an earlier diagnosis of MDD is related to a
longer conversion delay than an initial MDD diagnosis later in life. This
could be the result of numerous factors, such as a delay in subsequent
psychiatric care following the initial diagnosis and treatment plan, a
change in psychiatrist, or a possible decrease in symptom severity
and, thus, no follow-­up care. Therefore, it is possible that the delay
for each patient is longer than necessary. As previous research sug-
gests, it takes approximately 3–4 incorrect clinical assessments prior
to the establishment of the diagnosis of BD,3,16 and this may all be
delayed because younger patients are less likely to seek help when
“feeling high or energized” with (hypo)manic symptoms or there is in-
sufficient awareness of the pathological character of these symptoms.
The longer diagnosis time for younger MDD-­diagnosed patients may
also be a result of the fear associated with the stigma of mental illness,
a concern more pronounced for males,27 thus leading to a stronger
association between an earlier age at onset and the delay in conver-
sion in males. However, this longer delay in younger patients may also
be simply because they are younger and have more years ahead of
them. Further, Figure 1 clearly illustrates the variability in diagnostic
conversion and the reality of the long delay in diagnosis for many pa-
tients. Almost one-­third of the patients in the present study were not
diagnosed with BD until more than 10 years after their initial diagnosis
of MDD. Early recognition and initiation of effective treatment for BD
is likely to reduce disability and enhance outcomes.
 |
400       FRITZ et al.
4.1 | Limitations
The current study provides another piece of evidence on the retro-
spective examination of the conversion from MDD to BD diagnoses
but the imperative remains to conduct prospective research. Further,
as a natural limitation in this line of work, missing data and patients
who choose to discontinue prior to the completion of the study may
affect the full understanding of these patients whose diagnosis con-
verts to BD. Finally, antidepressants are likely to obfuscate the natural
course of the illness, and, as the majority of the patients included here
and in similar studies are, or have been, treated with antidepressants,
further research should examine how pharmacological treatment may
influence this conversion.
5 |  CONCLUSIONS
The delay in the diagnosis of BD may be caused by: (i) an inherent in-
ability to define BD (taxonomic delay) prior to the occurrence of a manic
episode, even if the patient has had numerous depressive episodes;28
or (ii) an identification delay (detection delay), in which the patient ex-
periences hypomanic symptoms but does not seek help, or the clinician
does not detect symptoms. The present study, by focusing on associa-
tions with the delay in diagnosis, revealed potential clinical markers for
the early detection of BD in patients who have not yet experienced or
heralded a manic episode. Research such as this may assist in diminish-
ing the seemingly inevitable delay in the diagnosis of BD.
REFERENCES
1. Baethge C, Tondo L, Bratti IM, et al. Prophylaxis latency and outcome
in bipolar disorders. Can J Psychiatry. 2003;48:449‐457.
2. Hirschfeld RM, Lewis L, Vornik LA. Perceptions and impact of bipolar
disorder: how far have we really come? Results of the national depres-
sive and manic-depressive association 2000 survey of individuals with
bipolar disorder. J Clin Psychiatry. 2003;64:161‐174.
3. Ghaemi SN, Ko JY. “Cade’s disease” and beyond: misdiagnosis, antide-
pressant use, and a proposed definition for bipolar spectrum disorder.
Can J Psychiatry. 2002;47:125‐134.
4. Ghaemi SN, Sachs GS, Chiou AM, Pandurangi AK, Goodwin FK. Is bi-
polar disorder still underdiagnosed? Are antidepressants overutilized?
J Affect Disord. 1999;52:135‐144.
5. Bukh J, Andersen P, Kessing L. Rates and predictors of remission,
recurrence and conversion to bipolar disorder after the first lifetime
episode of depression–a prospective 5-year follow-up study. Psychol
Med. 2016;46:1151‐1161.
6. Dudek D, Siwek M, Zielińska D, Jaeschke R, Rybakowski J. Diagnostic
conversions from major depressive disorder into bipolar disorder in
an outpatient setting: results of a retrospective chart review. J Affect
Disord. 2013;144:112‐115.
7. Li C-T, Bai Y-M, Huang Y-L, et al. Association between antidepressant
resistance in unipolar depression and subsequent bipolar disorder:
cohort study. Br J Psychiatry. 2012;200:45‐51.
8. Sharma V, Khan M, Smith A. A closer look at treatment resistant depres-
sion: is it due to a bipolar diathesis? J Affect Disord. 2005;84:251‐257.
9. Angst J, Sellaro R, Stassen HH, Gamma A. Diagnostic conversion from
depression to bipolar disorders: results of a long-term prospective
study of hospital admissions. J Affect Disord. 2005;84:149‐157.
10. Suominen K, Mantere O, Valtonen H, et al. Early age at onset of bi-
polar disorder is associated with more severe clinical features but de-
layed treatment seeking. Bipolar Disord. 2007;9:698‐705.
11. Coryell W, Endicott J, Maser JD, Mueller T, Lavori P, Keller M. The
likelihood of recurrence in bipolar affective disorder: the importance
of episode recency. J Affect Disord. 1995;33:201‐206.
12. Tondo L, Visioli C, Preti A, Baldessarini RJ. Bipolar disorders following
initial depression: modeling predictive clinical factors. J Affect Disord.
2014;167:44‐49.
13. Goldberg JF, Harrow M, Whiteside JE. Risk for bipolar illness in pa-
tients initially hospitalized for unipolar depression. Am J Psychiatry.
2001;158:1265‐1270.
14. Spielberger CD. Manual for the State-Trait Anxiety Inventory (STAI). Palo
Alto: Consulting Psychologists Press; 1983.
15. Rosenberg M. Society and the Adolescent Self-Image. Vol. 11. Princeton,
NJ: Princeton University Press; 1965.
16. Hirschfeld R, Calabrese J, Weissman M, et  al. Screening for bipolar
disorder in the community. J Clin Psychiatry. 2003;64:53‐59.
17. Taylor S, Zvolensky MJ, Cox BJ, et  al. Robust dimensions of anx-
iety sensitivity: development and initial validation of the Anxiety
Sensitivity Index-3. Psychol Assess. 2007;19:176.
18. Carver CS, Scheier MF, Weintraub JK. Assessing coping strategies: a
theoretically based approach. J Pers Soc Psychol. 1989;56:267.
19. Weissman AN, Beck AT. Development and Validation of the Dysfunctional
Attitude Scale: A Preliminary Investigation. Paper Presented at the
Meeting of the Annual Meeting of the American Educational Research
Association, Toronto, ON, Canada; 1978.
20. Costa PT, MacCrae RR. Revised NEO personality inventory (NEO
PI-R) and NEO five-factor inventory (NEO FFI): Professional manual:
Psychological Assessment Resources; 1992.
21. Akiskal HS, Akiskal KK, Lancrenon S, et  al. Validating the bipolar
spectrum in the French National EPIDEP Study: overview of the phe-
nomenology and relative prevalence of its clinical prototypes. J Affect
Disord. 2006;96:197‐205.
22. Rybakowski J, Dudek D, Pawlowski T, Lojko D, Siwek M, Kiejna A. Use
of the Hypomania Checklist-32 and the Mood Disorder Questionnaire
for detecting bipolarity in 1,051 patients with major depressive disor-
der. Eur Psychiatry. 2012;27:577‐581.
23. Rybakowski JK, Suwalska A, Lojko D, Rymaszewska J, Kiejna A.
Types of depression more frequent in bipolar than in unipolar af-
fective illness: results of the Polish DEP-BI study. Psychopathology.
2007;40:153‐158.
24. Tracey TJ, Sherry P, Bauer GP, Robins TH, Todaro L, Briggs S. Help
seeking as a function of student characteristics and program descrip-
tion: a logit-loglinear analysis. J Couns Psychol. 1984;31:54.
25. Komiya N, Good GE, Sherrod NB. Emotional openness as a predictor
of college students’ attitudes toward seeking psychological help. J
Couns Psychol. 2000;47:138.
26. Kakhnovets R. Relationships among personality, expectations about
counseling, and help-seeking attitudes. J Couns Dev. 2011;89:11.
27. Chandra A, Minkovitz CS. Stigma starts early: gender differences
in teen willingness to use mental health services. J Adolesc Health;
2006;38:754. e1‐754. e8.
28. Malhi GS, Bassett D, Boyce P, et al. Royal Australian and New Zealand
college of psychiatrists clinical practice guidelines for mood disorders.
Aust N Z J Psychiatry. 2015;49:1087‐1206.
How to cite this article: Fritz K, Russell AMT, Allwang C,
Kuiper S, Lampe L, Malhi GS. Is a delay in the diagnosis of
bipolar disorder inevitable?. Bipolar Disord. 2017;19:
396–400. https://doi.org/10.1111/bdi.12499