Abstract
Community-based surveys conducted worldwide1–5 have indicated that mood and anxiety disorders are prevalent in the general population and are associated with a wide range of disabilities and functional impairments. In light of these findings, these psychiatric disorders are now recognized as a leading contributor to the global disease burden. 6 Large-scale epidemiological studies conducted in various countries worldwide have yielded lifetime prevalence rates of mood and anxiety disorders ranging between 0.3–25.0% and 4.8–31.0%, respectively. 7 In Australia, data from the 1997 National Survey of Mental Health and Wellbeing (NSMHW) demonstrated that approximately one in five adults had experienced a psychiatric disorder in the previous 12 months, 4 while more recent data indicate that 14.4%, 6.2% and 5.1% have had a 12-month anxiety, mood or substance-use disorder, respectively. 8
Subthreshold psychological symptomatology is even more prevalent and is known to be associated with more service utilization and social morbidity than threshold disorders. 9 , 10 Furthermore, longitudinal data indicate that individuals with subsyndromal depressive symptoms are more than four times more likely to develop de novo major depression within one year. 11 In Australia, 12.9% of adults were identified with current subsyndromal depression, based on the mood module of the Primary Care Evaluation of Mental Disorders (PRIME-MD). 10 As such, mood and anxiety disorders, as well as symptoms, impose huge costs on individuals and the public health burden. In Australia alone, over $3 billion per annum (or 9.6% of the total health care expenditure) is spent on associated mental health system costs. 12
In this study, we aimed to estimate the lifetime and current prevalence, age-of-onset, and comorbidity of mood and anxiety disorders in a representative sample of the adult female population in Australia, utilizing a gold-standard clinical interview (Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, non-patient edition; SCID-I/NP). Furthermore, this study presents information regarding psychological symptomatology, measured with a self-report instrument.
METHODS
Participants
The Geelong Osteoporosis Study (GOS) is a large epidemiological study involving an age-stratified, random, population-based sample of women, initially developed to investigate predictors and consequences of osteoporosis. This study has been recently expanded to examine other non-communicable diseases in the population, with a particular focus on the common mental disorders. Participants were recruited from the Commonwealth of Australia Electoral Rolls for the Barwon Statistical Division (BSD), a geographically well-defined area surrounding the regional city of Geelong, located in southeastern Australia. The BSD reflects the Australian population in terms of age distribution, country of birth, education, marital status, employment and income profiles, allowing data from this sample to be extrapolated to national levels. 13
Originally, 1494 randomly selected women (aged 20–94 years) were recruited between 1994 and 1997, with a participation of 77.1%. Further details are provided elsewhere. 13 Of this group, 881 of the 1494 original women returned for a 10-year follow-up appointment. Reasons for non-participation included death (n = 226), migration from the region (n = 82), inability to give informed consent (n = 12), pregnancy (n = 1) or inability to be contacted (n = 100). Of the remaining 192 who declined, 35.9% cited personal reasons (disinterested, invasion of privacy, fear of hospitals), 39.1% old age, 14.1% illness, 6.3% time constraints, 4.2% distance, and in 0.5% a reason was unable to be established. The response among eligible women was 82.1%.
Coinciding with the 10-year follow-up period, between 2004 and 2008, an additional sample with a minimum of 100 women in each 5-year age stratum between ages 20 and 29 years were randomly selected from the BSD and invited to participate in the GOS. This allowed the full adult age range to be investigated. Of the 347 eligible women, 246 consented to participate, with a response of 70.9%. Reasons for non-participation included inability to be contacted (n = 179), no longer a resident of the study region (n = 140), pregnancy (n = 11) or aged over 29 years (n = 73). Of the remaining 101 who declined participation, 34.7% cited personal reasons, 53.5% time restraints, 5.0% illness, 2.0% language barrier, 1.0% too far to travel and 4.0% repeatedly failed to keep appointments.
From a potential pool of 1127 women, participants for whom psychiatric data were not available (n = 32) were excluded from the analyses, resulting in a sample of 1095 women aged 20–93 years eligible for inclusion. The study was approved by the Barwon Health Human Research Ethics Committee and written informed consent was obtained from all participants.
Assessments
Psychiatric history and current status of each participant was determined using the SCID-I/NP. 14 This is the gold-standard assessment tool comprising a validated, semi-structured clinical interview for the major axis psychiatric disorders in DSM-IV-TR. In this study, the SCID-I/NP was used to assess the presence of past or current mood disorders including major depressive disorder (MDD), minor depression (MinD), dysthymia (current only), mood disorder due to a general medical condition (GMC), substance induced mood disorder, and bipolar disorder (I, II and not otherwise specified). Past and current anxiety disorders including panic disorder, agoraphobia, social phobia, specific phobia, obsessive-compulsive disorder (OCD), generalized anxiety disorder (current only), anxiety disorders due to a GMC, substance induced anxiety disorder and anxiety disorders not otherwise specified (NOS) were also assessed. Eating and substance use disorders were not ascertained in this study. Further information regarding age-of-onset was also gathered. Trained personnel with qualifications in psychology conducted all psychiatric interviews.
Psychological symptomatology was assessed using the General Health Questionnaire (GHQ-12). 15 This is a well-established screening instrument designed to detect non-psychotic psychiatric disorders in community and medical settings using a 12-item self-report questionnaire based on respondent's assessment of their present state. Binary scoring was used for each item, yielding a total score ranging from 0 to 12. A cut-off point of 3/4, being the most widely used convention, was used to define caseness of mental health dysfunction. 15 , 16
Data regarding demographic, lifestyle, medical and other parameters were also collected. Height and weight were measured to the nearest 0.1 cm and 0.1 kg, respectively. Socio-economic status was determined using Socio-Economic Index For Areas (SEIFA) index scores based on the 2006 Australian Bureau of Statistics (ABS) Census data. Using ABS software, individuals were assigned a SEIFA index score based on their residential address. It was decided a priori to use the SEIFA Index of Relative Socio-economic Advantage and Disadvantage (IRSAD), which accounts for high and low income, and the type of occupation from unskilled employment to professional positions. A low score as measured by IRSAD identifies the most disadvantaged (quintile 1), and a high score identifies the most advantaged (quintile 5). Education (highest level completed) was self-reported, as was country of birth and menopause status. Participants were asked to bring in a list of medications or containers to assist with accurate recording of details and medications were then classified according to a pharmaceutical database (MIMS Australia Pty Ltd, 2003).
Statistical analyses
The lifetime and current prevalence of psychiatric disorders was determined from the study population and standardized to the 2006 census data from the Australian Bureau of Statistics population figures for Australia. 17 Mean age-of-onset was also determined. Statistical analyses were performed using Minitab (Version 15; Minitab, State College, PA, USA).
RESULTS
Characteristics of the women included in this study are shown in Table 1. Approximately one in three women (34.8%) reported a lifetime history of a mood and/or anxiety disorder and one in seven women (14.4%) reported a current mood and/or anxiety disorder at the time of interview. The overall prevalence of psychological symptomatology, as measured by the GHQ-12, was 21.2%. Half of all women who suffered from any mood and/or anxiety disorder had developed the disorder by 24 years of age (IQR 16.0–38.0). Of those with a lifetime history of any mood and/or anxiety disorder, 9.8% met criteria for at least one other lifetime disorder and 3.2% had two or more current disorders.
Characteristics of study sample. Values are given as median (interquartile range), mean (standard deviation) or n (%)
Mood disorders
Prevalence data were standardized to the Australian population. After standardization, the lifetime prevalence of any mood disorder was 30.0%. The most prevalent mood disorder was MDD, with 23.4% of respondents reporting a lifetime history. Other mood disorders were less common; approximately 3.0%, 2.5%, 0.6% and 0.3% met criteria for a lifetime history of MinD, bipolar disorder, mood disorder due to a GMC, and substance induced mood disorder respectively (Figure 1). The median age-of-onset for any mood disorder was 27.0 years (IQR 19.0–40.0) (Figure 2).
Lifetime and current prevalence of mood disorders standardized to Australian population (2006) (error bars represent 95% CI).
Age-of-onset distribution for any mood disorder.
Overall, 8.9% of the population were identified as suffering from a current mood disorder. MDD was the most common current disorder, with 4.7% meeting criteria at the time of interview. Bipolar disorder, MinD and dysthymia yielded similar rates, with approximately 1.4%, 1.2% and 1.8% currently affected, respectively. Mood disorder due to a GMC and substance induced mood disorder were the least common mood disorders (approximately 0.1% and 0.3%, respectively) (Figure 1).
Anxiety disorders
The lifetime prevalence of any anxiety disorder was 13.5%. Panic disorder was the most common disorder, with 5.5% meeting lifetime criteria, while specific and social phobia were the next most prevalent (3.5% and 2.2%, respectively). The other anxiety disorders were less common; approximately 1.6%, 1.4% and 0.8% met criteria for a lifetime history of anxiety disorder NOS, OCD and agoraphobia, respectively. The least prevalent disorders were substance induced anxiety disorder and anxiety disorder due to a GMC (0.3% and 0.2%, respectively) (Figure 3). The median age-of-onset for any anxiety disorder was 18.5 years (IQR 13.3–35.8) (Figure 4).
Lifetime and current prevalence of anxiety disorders standardized to Australian population (2006) (error bars represent 95% CI).
Age-of-onset distribution for any anxiety disorder.
The standardized prevalence for current anxiety disorders was 8.0%. Specific phobia was the most prevalent, with 3.0% meeting current criteria. The next most common disorders were panic disorder, anxiety disorder NOS and generalized anxiety disorder (GAD) with current prevalence of approximately 1.4%, 1.2% and 1.1%, respectively. The prevalence of the remaining anxiety disorders was low (Figure 3).
DISCUSSION
These findings are consistent with those of other epidemiological studies conducted in Australia and other countries in demonstrating that mood and anxiety disorders are highly prevalent in the general population. Approximately one in three women reported a lifetime history of any mood and/or anxiety disorder, with mood disorders being the most prevalent. Approximately one in seven women reported a current mood and/or anxiety disorder at the time of interview.
Based on the cross national results from the World Health Organization's World Mental Health (WMH) Survey Initiative, between 0.3–25.0% and 4.8–31.0% of the study populations met criteria for a lifetime history of mood and anxiety disorders, respectively, with six of the 17 countries included also finding mood disorders to be the most prevalent. 7 Compared to our results, previous epidemiological studies have reported lower prevalence rates for lifetime mood and anxiety disorders. For example, results from the National Comorbidity Survey Replication study (NCS-R), conducted between 2001 and 2003, reported the lifetime prevalence of any mood disorder to be 20.8%, and 28.8% for any anxiety disorder in a sample of 9282 community-dwelling adults. 2 The reported rates in the European Study of the Epidemiology of Mental Disorders/Mental Health Disability (ESEMeD), a representative cross-sectional community sample of 21 425 adults, were even lower, with 14.0% and 13.6% of the sample meeting lifetime criteria for any mood and anxiety disorder respectively. 3 In New Zealand, a nation wide survey of 12 992 participants was carried out between 2003 and 2004. 5 In that study, anxiety disorders were most prevalent, with 24.9% meeting criteria for a lifetime history, followed by mood disorders (20.2%). As with the Australian study (NSMHW), all of the above have used the Composite International Diagnostic Interview (CIDI) as the assessment tool. It has been proposed, however, that this instrument may underestimate the true prevalence of depression, 18 suggesting one possible explanation as to why our prevalence rates are higher than those previously reported.
The prevalence rates for lifetime mood disorders in our study are largely attributable to MDD diagnoses. Lifetime rates of MDD have previously been reported to range between 0.9% (Taiwan) and 29.6% (Canada). 19 As such, the lifetime rates of MDD reported in our study are relatively high compared to the previous literature. The SCID-I/NP is similar to other instruments used in recent epidemiological studies in that it is based on DSM-IV criteria. However, differences in wording and depth of probing during the interviews could have had important effects on diagnosis rates. It is possible that a proportion of individuals meeting criteria for a past history of single-episode MDD following an adverse life event could have been better explained by a diagnosis of adjustment disorder at the time of the event; the SCID has a module to differentiate current but not lifetime adjustment disorder from MDD. Another obvious disparity between our figures and those of previous studies is the low prevalence observed for specific phobia. Others have reported lifetime figures of between 7.7% (ESEMeD) and 12.5% (NCS-R). The threshold for meeting criteria for a specific phobia was rigorously adhered to, and the criterion regarding impairment in daily functioning had to be evident, possibly precluding diagnoses of milder cases. Sample size also has to be taken into consideration when interpreting the prevalence of the less common disorders, and our estimates may have lacked precision. However, prevalence of these disorders was similar to other previously conducted surveys.
Comorbidity has been shown to be an index of a more severe course and poorer outcome of mental illness. 20 In this study, approximately, 10% of women with a lifetime history of any mood and/or anxiety disorder met criteria for at least one other lifetime disorder, while nearly 3% of women had two or more current disorders. These rates are considerably reduced in comparison to other reports. The NCS 1 and NCS-R 2 reported lifetime comorbidity rates of 56% and 27.7%, respectively. The NSMHW 21 and the New Zealand Mental Health Survey 22 reported 12-month comorbidity rates of 39% and 37%, respectively. Extraneous variables may have contributed to the low rates of comorbidity reported in this study. It is likely that the exclusion of the substance use and eating disorder modules of the SCID-IV/NP, which are known to be highly comorbid with both mood and anxiety disorders, may have reduced the comorbidity rate in our study. Moreover, given that comorbidity is associated with increased severity and chronicity and significantly greater impairment, the more severe cases may not have been represented in our community-based sample.
Our findings are consistent with those reported in previous studies, in finding that anxiety disorders had an earlier age-of-onset than mood disorders. Cross national results from the World Health Organization's WMH Survey, reported the age-of-onset distribution for mood disorders to be between 29 and 43 years, compared to anxiety disorders, where the age-of-onset distribution were different for particular disorders. 7 For example, phobias demonstrated a very early onset (7–14 years), in comparison to GAD, panic disorder and posttraumatic stress disorder (24–50 years). 7 In our study, age-of-onset for any mood disorder declined after 28 years. However, for anxiety disorders the age-of-onset declined after approximately 19 years, although a bimodal distribution was evident with a second peak observed at approximately 45 years. A possible explanation for this second peak may relate to the experience of psychological, as well as physiological, symptoms during menopause, such as negative mood, anxiety and irritability, possibly prompted by changes in hormone levels. 23
Inaccurate recall of previous experiences is a common bias when measuring lifetime prevalence of depression. 24 In a study investigating recall bias and depression, 70% of people who were hospitalized for a MDD episode recalled being depressed, but only half could recall sufficient detail to satisfy the diagnostic criteria when interviewed 25 years later. 25 Furthermore, the accuracy of age-of-onset recall is often biased, with age at interview, in conjunction with time between onset and recall, being influential factors. However, the use of a gold-standard semi-structured clinical interview, wherein participants were encouraged to actively recall previous illnesses in great detail, may have reduced the impact of such biases. Another limitation is that the psychiatric profile of GOS non-respondents may have differed from responders, thus potentially biasing our results. As such, it is not unrealistic to suggest that our reported rates, particularly those for comorbidity, are underestimated.
In conclusion, these findings provide epidemiological data utilizing the SCID-IV/NP in a representative sample of the adult female population in Australia. Given that mood and anxiety disorders are common, it is of importance to have a good knowledge of their prevalence, age-of-onset, and related specifiers such as comorbidity. Community-based studies investigating the prevalence of mood and anxiety disorders in a representative sample yields a more accurate reflection of the true burden of illness. This study provides important information that can be utilized from a clinical and preventative standpoint – complementing growing attention to gender-sensitive mental health care 26 – but also from an economic and public health perspective, in terms of service planning and anticipating disability.
Footnotes
Acknowledgements
The study was funded by the National Health and Medical Research Council of Australia (251638) and supported by an unrestricted educational grant from Eli Lilly. Postgraduate scholarships were provided by the University of Melbourne, Faculty of Medicine, Dentistry and Health Sciences and the Australian Rotary Health Research Fund. We thank Sharon Brennan for obtaining the socio-economic data for the study. The funding providers played no role in the design or conduct of the study; collection, management, analysis, and interpretation of the data; or in preparation, review, or approval of the manuscript.
Lana Williams, Julie Pasco, Felice Jacka and Seetal Dodd have received research support from an unrestricted educational grant from Eli Lilly. Margaret Henry, Mark Kotowicz and Geoffrey Nicholson have no conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript.
Michael Berk has received grant or research support from the Stanley Medical Research Foundation, Medical Benefits Fund (MBF), National Health and Medical Research Council (NHMRC), Beyond Blue, Geelong Medical Research Foundation, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Organon, Novartis, Mayne Pharma and Servier; and is a speaker or consultant for Astra Zeneca, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Janssen Cilag, Lundbeck, Pfizer, Sanofi Synthelabo, Servier, Solvay and Wyeth.