Prevalence of and Risk Factors for Depressive Symptoms and Antidepressant Use in Young Australian Women: A Cross-Sectional Study

Abstract

Background:

Despite the high prevalence of depression among adult women, the proportion of reproductive-aged women with moderate or severe depressive symptoms is uncertain, as is the proportion taking antidepressant medication. We report the prevalence of depressive symptoms in young Australian women, risk factors for depressive symptoms, and psychoactive drug use.

Methods:

An online survey was completed by population-based sample of 6,986 Australian women, aged 18–39 years, recruited from November 2016 to July 2017. Depressive symptoms were assessed by the Beck Depression Inventory-II, and psychotropic medication use was self-reported.

Results:

The prevalences of moderate and severe depressive symptoms were 15.0% (95% confidence interval [CI] 14.1%–15.8%) and 14.8% (95% CI 14.0%–15.7%), respectively. Housing insecurity was associated with over a twofold likelihood of moderate to severe depressive symptoms, whereas being parous or at least 25 years of age was protective. Use of any psychotropic medication was reported by 16.3% (95% CI 15.4%–17.2%). A previous cancer diagnosis was the strongest risk factor for current antidepressant use, whereas compared with being of European ancestry, being Asian or of another ancestry was associated with a lower likelihood of antidepressant use.

Conclusion:

The prevalence of moderate to severe depressive symptoms among young Australian women is alarming. Prevention strategies targeting the sociodemographic circumstances underpinning the identified risk factors are urgently needed.

Introduction

Depression is ranked by the World Health Organization (WHO) as the single leading cause of disability.¹ Women are disproportionately affected, with global estimates that depression affects 4.1% of women compared with 2.7% of men.² Depression causes functional disability across work, social, and domestic life and is a major antecedent to suicide.³ Despite the profound detrimental effects of depressive symptoms, data for the extent of the problem among young women are scant. The 1993 National Comorbidity Survey in the United States reported that the prevalence of depression ranged from 11% to 16% in women aged 15–44.⁴

The prevalence of depression in premenopausal Australian women is uncertain with estimates ranging from 3% to 23%.^5
–7 There has been extensive research into risk factors for antenatal and postnatal depression. However, the factors that specifically underpin depression in young women not associated with pregnancy and childbirth are not known. We have previously identified independent risk factors for depression in women at midlife to include being unpartnered or unemployed, obesity, and housing insecurity.⁸

Depression has been associated with worse outcomes for assisted reproduction.⁹ But whether women undergoing assisted reproduction are more likely to be depressed than other women is not known. In addition, the use of prescription psychotropic medication in countries, such as Australia, is escalating.^10,11 From 2000 to 2011 the dispensing of psychotropic medications in Australia increased by 58%.¹² Australians are among the most frequent users of antidepressants,¹³ particularly Australian women^14,15 with approximately one quarter of women at midlife taking an antidepressant medication.¹⁶ However, it is not known what proportion of younger Australian women are taking antidepressants nor which type of antidepressants younger women are likely to be taking.

It is of the utmost importance to identify the magnitude of the prevalence of depressive symptoms during an especially challenging time in a woman's life: the years spent juggling pregnancy and childbirth, child rearing, relationships, study, and workforce participation. Public health strategies and clinical treatments alike need to be based on sound epidemiological evidence in order for interventions to be targeted and effective in reducing the burden of depression in this vulnerable age group. Our goal was to estimate the prevalence of, and risk factors for, moderate to severe depressive symptoms and the use of prescribed antidepressant medicines in a representative sample of Australian women aged 18 to 39 years.

Methods

Data source and sample

The Grollo-Ruzzene Foundation Younger Women's Health Study (GR Study) was an online, questionnaire-based cross-sectional study designed to investigate the health and well-being of Australian women aged 18 to 39 years. A representative, community-based sample of 6,986 women was recruited from Victoria, Queensland and New South Wales, by e-mail invitation.¹⁷

Two national databases, Roy Morgan Research (RMR) and Survey Sampling International (SSI), were used for recruitment. The RMR database is based on Australian electoral areas and updated continuously. The SSI is an international, digital research organization with opt-in recruitment via banners, invitations, messages, and partnerships. Digital fingerprinting is used to identify duplicate responses. E-mail accounts, two-factor authentication, and pattern recognition software are used to identify fraudulent responses, thereby ensuring quality control of recruitment.

Women were only excluded if they could not read and understand English proficiently enough to complete the online survey. E-mail invitations included a link to the study explanatory statement. Following this, an “agree” option could be selected to progress to the questionnaire. The survey was piloted and modified according to community feedback regarding the presentation of some questions and response options.¹⁷ In total, 94,546 invitations were sent, 8,193 women accessed the questionnaire (8.7%), and 7,053 (7.5%) questionnaires were completed between November 11, 2016 and July 21, 2017.¹⁷ Sixty-six duplicate responses were removed due to recruitment by both RMR and SSI, and one participant was outside the age range. Ultimately, 6,986 participants provided completed questionnaires. All participants provided informed consent before linkage with the study questionnaire, and no payment was made for questionnaire completion.

Based on Australian Census data our sample is representative of Australian women of the respondents' age in terms of age distribution, state of residence, relationship status, employment, and occupation.¹⁷ Study participants were more likely to have been Australian born, as Australian Census data include international visitors, and more likely to have completed high school because the census data for education include 15–17-year olds in the under 40 years group.¹⁷

The study was approved by the Monash University Human Research Ethics Committee [Approval number: CF16/2322-2016001166 (7703)].

Measures

Study questionnaire

The questionnaire collected information about sociodemographics (ancestry, level of education, relationship status and household composition, employment, household financial security, and whether they were a carer for another person with special needs, henceforth described as a “carer”), body mass index (BMI) (calculated on the basis of self-reported height and weight), lifestyle (alcohol, smoking, and exercise), reproductive (parity and current use of assisted reproductive technology), and general medical history (specifically ever having been diagnosed with cancer).

Assessment of depressive symptoms

Depressive symptoms were ascertained by the Beck Depression Inventory-II (BDI-II),¹⁸ which shows high discrimination between depressed and nondepressed people.^18
–20 As respondents could not progress without providing a response, there are no missing data for the BDI-II. The BDI-II refers to symptoms over the prior two weeks. The BDI-II consists of 21 questions, each answer being scored on a scale value of 0 to 3 such that the total score may range from 0 to 63. Higher total scores indicate more severe depressive symptoms. The standardized scoring for symptoms is as follows: none-minimal (BDI-II score 0–13), mild (14–19), moderate (20–28), and severe (29–63).^18,21

Psychoactive medication use

Participants identified psychoactive medication used in the prior 4 weeks from a complete list of the proprietary names of all prescription psychoactive medications available in Australia. The included antidepressants were selective serotonin reuptake inhibitors (SSRIs), serotonin-norepinephrine reuptake inhibitors (SNRIs), tricyclics, tetracyclics, and agomelatine; and antipsychotics, benzodiazepines, anticonvulsants, and sleep specific medications.

Sample size and data analysis

The prevalence of depressive symptoms was a predefined study outcome. A sample size of 7,000 women provides sufficient power for a prevalence estimate for depressive symptoms of 12% to have a 95% confidence interval (CI) from 11.3% to 12.7%.

For age, the youngest age group was compared with older women in 5-year age groups. For BMI, women in the normal range were compared with underweight, overweight, and obese women. For ancestry, women who classified themselves as being of European ancestry were compared with women who self-classified as having Aboriginal or Torres Strait Islander, Asian, or other ancestry. In terms of education, women were classified as either having completed high school or not. Women were classified as studying or working or both as one category compared with those women who were neither studying nor working. For relationship status, the reference group was women who were married or in a de facto relationship compared with women who had either a boyfriend/girlfriend or described themselves as single.

For living arrangements, women were classified as either living with at least one other adult or not. Women who did not live with at least one other adult could have either lived alone or been a single parent with children. Women were described as either parous or nulliparous. They had also been asked if they were currently using assisted reproductive technology. Women were classified as either having secure housing or not, depending on their answer to a question about whether they were confident that they could afford their current housing arrangements. Alcohol use was classified as heavy (≥4 standard drinks per occasion, at least once a week) versus none/occasional. Smoking was classified as current versus past/never. Exercise was classified as at least once per week versus less often.

For the analysis of psychotropic medications, we excluded participants with epilepsy, migraine, or another neurological condition as the included medications may have been prescribed for these conditions, and their inclusion could result in overestimation of the prevalence of treatment of depressive symptoms. As some participants reported multiple medications, including more than one in the same class, women reporting three or more psychotropic medications were considered separately, due to the likelihood that they were reporting lifetime use, rather than use in the previous 4 weeks.

Chi-square analysis was used to test for differences in categorical variables, between groups of women. The score of the BDI-II was dichotomized into “none-mild” and “moderate-severe” depressive symptoms based on previous research that found this cutoff appropriate.²¹ Univariate logistic regression was used to assess the association between depressive symptoms (moderate–severe symptoms on the BDI-II) and other variables. Variables were tested for collinearity, and state of residence was excluded from the multivariable model, as it was associated with place of residence. Potential interactions explored for both models included age and parity, relationship status and parity, relationship status and employment, and household composition and employment. Multivariable logistic regression analysis, including all variables which were significant at a univariate level, was used to examine factors associated with the presence of depressive symptoms and to the use of antidepressants. All analyses were performed using the Stata statistical software (version 15.0; StataCorp LP, College Station, TX).

Results

Our sample of 6,986 participants was predominantly of European ancestry (73.1%) and living in metropolitan areas (66.6%) (Table 1). The majority of women completed high school (94.7%), and 80.5% of women reported either working or studying or both. Nearly one quarter were obese (24.2%), 14.4% were current smokers, and 11.4% were classified as heavy drinkers. Housing insecurity was reported by 17.0%, and 3.6% gave a past history of cancer.

Table 1.

Sociodemographic, Reproductive, and Lifestyle Characteristics of Young Australian Women (n = 6,986)

Characteristics	n (%)
Age in years
18–<25	2,062 (29.5)
25–<30	1,595 (22.8)
30–<35	1,695 (24.3)
35–<40	1,634 (23.4)
BMI (kg/m²)
<18.5 (underweight)	429 (6.2)
18.5–<25 (normal weight)	3,270 (47.1)
25–<30 (overweight)	1,563 (22.5)
≥30 (obese)	1,688 (24.3)
Ancestry
European	5,109 (73.1)
Aboriginal/Torres Strait Islander	127 (1.8)
Asian	1,074 (15.4)
Other	676 (9.7)
Place of residence
Metropolitan	4,652 (66.6)
Rural	2,334 (33.4)
Completed high school
No	372 (5.3)
Yes	6,614 (94.7)
Relationship status^a
Married/de facto	3,561 (51.1)
Girl or boyfriend/single	3,411 (48.9)
Working and/or studying
Yes	5,626 (80.5)
No	1,360 (19.5)
Parity
0	4,333 (62.0)
≥1	2,653 (38.0)
Living with at least one other adult
Yes	6,076 (87.0%)
No	910 (13.0)
Housing financial security^b
Secure	5,772 (83.0)
Insecure	1,179 (17.0)
Current smoker
No	5,983 (85.6)
Yes	1,003 (14.4)
Heavy drinking^c
No	6,194 (88.7)
Yes	792 (11.3)
Exercise
Once a week minimum	5,165 (73.9)
<Once a week	1,821 (26.1)
Carer for special needs person
Yes	698 (10.0)
Currently using assisted reproductive technology
Yes	118 (1.7)
Diagnosed with any type of cancer
Yes	254 (3.6)
Antidepressant medication use^d
Yes	1,131 (13.2)

n = 6,972 (lower due to missing data).

n = 6,951 (lower due to missing data).

≥4 drinks on one occasion at least weekly.

Includes: SSRIs; SNRIs; tricyclics, tetracyclics, and agomelatine.

BMI, body mass index; SNRIs, serotonin-norepinephrine reuptake inhibitors; SSRIs, selective serotonin reuptake inhibitors.

Moderate depressive symptoms (median BDI-II score [interquartile range] 23 [5]) were present in 15.0% (95% CI 14.1%–15.8%) of women and severe symptoms (36 [11]) in 14.8% (95% CI 14.0%–15.7%) in the previous 2 weeks. Mild depressive symptoms (16 [3]) were reported by 13.9% (95% CI 13.1%–14.7%) of participants, while no symptoms (5 [7]) were reported by 56.3% (95% CI 55.9%–57.5%).

In the multivariable-adjusted analyses, the only interaction term that achieved statistical significance at the 5% level was between relationship status and employment status in the model for depressive symptoms. Women were more likely to have moderate to severe depressive symptoms if they were obese, had not completed high school, were a carer, had housing insecurity, were a current smoker, reported heavy alcohol consumption, exercised less often than once a week, were single or had a boyfriend/girlfriend, plus were neither working nor studying (interaction term), had a previous cancer diagnosis, or reported current antidepressant medication use (Table 2). The largest odds ratio was for housing insecurity (2.3) and psychotropic medication use (3.1). Factors that emerged as protective against moderate to severe depressive symptoms were being parous and, compared with being aged 18 to 25 years, being at least 25 years of age (p < 0.001 for all age groups). The smallest odds ratio (0.4) was for the age group 35–<40 years, compared with women aged <25 years.

Table 2.

Prevalence Estimates of Depressive Symptoms and Logistic Regression Analysis Results for Variables Associated with the Likelihood of Depressive Symptoms

Characteristics	Moderate–severe depressive symptoms % (95% CI)	Unadjusted OR (95% CI)	p-Value	Adjusted ^a OR (95% CI)	p-Value
Age (years)
18–<25	37.0 (34.9–39.0)	1.00		1.00
25–<30	29.5 (27.3–31.8)	0.7 (0.6–0.8)	<0.001	0.6 (0.5–0.7)	<0.001
30–<35	26.7 (24.6–28.8)	0.6 (0.5–0.7)	<0.001	0.5 (0.4–0.6)	<0.001
35–<40	24.1 (22.1–26.2)	0.5 (0.5–0.6)	<0.001	0.4 (0.3–0.6)	<0.001
BMI (kg/m²)
<18.5 (underweight)	35.7 (31.1–40.2)	1.5 (1.2–1.8)	<0.001	1.2 (0.9–1.5)	0.088
18.5–<25 (normal weight)	27.3 (25.8–28.9)	1.00		1.00
25–<30 (overweight)	26.6 (24.4–28.8)	1.0 (0.8–1.1)	0.579	1.0 (0.8–1.1)	0.717
≥30 (obese)	35.8 (33.5–38.1)	1.5 (1.3–1.7)	<0.001	1.4 (1.2–1.6)	<0.001
Ancestry
European	29.8 (28.6–31.1)	1.00		1.00
Aboriginal/Torres Strait Islander	44.9 (36.5–53.6)	1.9 (1.3–2.7)	<0.001	1.4 (0.9–2.1)	0.079
Asian	25.8 (23.3–28.5)	0.8 (0.7–0.9)	0.008	1.1 (0.9–1.2)	0.510
Other	32.7 (29.3–36.3)	1.1 (0.9–1.3)	0.128	1.1 (0.9–1.3)	0.332
Place of residence
Metropolitan	29.7 (28.3–31.0)	1.00
Rural	30.0 (28.1–31.8)	1.0 (0.9–1.1)	0.807	—	—
Completed high school
Yes	28.9 (27.8–30.0)	1.00		1.00
No	45.7 (40.7–50.8)	2.1 (1.7–2.7)	<0.001	1.5 (1.2–1.9)	0.002
Relationship status
Married/de facto	24.1 (22.7–25.6)	1.00		1.00
Girl-boyfriend/single	35.6 (34.0–37.2)	1.6 (1.5–1.8)	<0.001	1.1 (0.9–1.3)	0.099
Working/studying or both
Yes	28.8 (27.6–30.0)	1.00		1.00
No	33.7 (31.2–36.2)	1.1 (0.9–1.3)	0.211	1.1 (0.9–1.3)	0.276
Interaction between relationship status and working and/or studying
Married/de facto and working and/or studying	23.6 (22.0–25.2)	1.00		1.00
Girl/boyfriend/single and neither working nor studying	49.0 (44.4–53.6)	1.7 (1.3–2.2)	<0.001	1.4 (1.1–1.9)	0.019
Parity
0	32.6 (31.2–34.0)	1.00		1.00
≥1	25.1 (23.5–26.8)	0.7 (0.6–0.8)	<0.001	0.8 (0.7–0.9)	0.005
Living with at least one other adult
Yes	29.1 (27.9–30.2)	1.00		1.00
No	34.4 (31.4–37.6)	1.3 (1.1–1.5)	0.001	1.0 (0.8–1.2)	0.887
Housing financial security
Secure	26.1 (25.0–27.3)	1.00		1.00
Insecure	47.4 (44.6–50.3)	2.5 (2.2–2.9)	<0.001	2.3 (2.0–2.6)	<0.001
Current smoker
No	27.2 (26.1–28.3)	1.00		1.00
Yes	45.3 (42.2–48.4)	2.2 (1.9–2.5)	<0.001	1.8 (1.5–2.1)	<0.001
Heavy drinking^b
No	28.7 (27.4–29.7)	1.00		1.00
Yes	39.1 (35.8–42.6)	1.6 (1.4–1.9)	<0.001	1.4 (1.1–1.6)	<0.001
Physical exercise
Once a week minimum	28.0 (26.8–29.2)	1.00		1.00
<Once a week	34.7 (32.6–36.9)	1.4 (1.2–1.5)	<0.001	1.3 (1.1–1.5)	<0.001
Carer for special needs person
No	28.6 (27.5–29.7)	1.00		1.00
Yes	40.4 (36.8–44.1)	1.7 (1.4–2.0)	<0.001	1.5 (1.3–1.8)	<0.001
Currently using assisted reproductive technology
No	29.3 (28.2–30.4)	1.00
Yes	33.1 (25.1–42.1)	1.2 (0.8–1.8)	0.374	—	—
Diagnosed with any type of cancer
No	29.0 (27.9–30.1)	1.00		1.00
Yes	49.6 (43.5–55.7)	2.4 (1.9–3.1)	<0.001	1.9 (1.4–2.5)	<0.001
Antidepressant medication use^c
No	25.0 (23.9–26.2)	1.00		1.00
Yes	54.2 (21.3–57.1)	3.5 (3.1–4.0)	<0.001	3.1 (2.7–3.5)	<0.001

Variables not significant in the univariate analysis were not included in the multivariable model, and so no data are shown for those variables in the multivariable model. The odds ratios for some variables which were included in the multivariable model did not reach statistical significance in the final model, but were included in the table for completeness

≥4 drinks on one occasion at least weekly

Includes: SSRIs; SNRIs; tricyclics, tetracyclics, and agomelatine.

CI, confidence interval; OR, odds ratio.

Of the 6,226 participants who did not report a diagnosis of epilepsy, migraine, or other neurological condition, 16.3% (95% CI 15.4%–17.2%) were taking one of the listed psychotropic medications (Table 3). The most commonly reported therapies among women reporting fewer than three medications were antidepressants, including SNRIs (48.4%) and SSRIs (33.5%), with 13.2% taking a benzodiazepine and 7.6% an antipsychotic. With respect to antidepressant use specifically, 26.6% of the women with moderate to severe depressive symptoms were taking an antidepressant. Conversely, of the women taking an antidepressant, 50.7% had moderate to severe depressive symptoms. Most SNRI and SSRI use was monotherapy, whereas the majority of benzodiazepine and antipsychotic use was as combination therapy. The most commonly reported combinations were an SSRI with a benzodiazepine or antipsychotic (3.5% and 2.5%, respectively).

Table 3.

Psychotropic Medication Use Among Young Australian Women Aged 18–39 Years (n = 6,226^a)

Medication use	n (%)
Nil	5,214 (83.7)
Any	1,012 (16.3)
No. of Medications (% all medication users, n = 1,012)
1 medication	787 (77.8)
2 medications	159 (15.7)
≥3 medications	66 (6.5)
Most common medications (% medication users, n = 946), alone or in combination^b
SSRI
Alone	232 (24.5)
+ benzodiazepine	33 (3.5)
+ antipsychotic	24 (2.5)
+ antiepileptic	3 (0.3)
+ other^c	16 (1.7)
+ SNRI^d	4 (0.4)
+ SSRI^d	5 (0.5)
SNRI
Alone	412 (43.5)
+ benzodiazepine	12 (1.3)
+ antipsychotic	17 (1.8)
+ antiepileptic	3 (0.3)
+ other^c	12 (1.3)
+ SNRI^d	2 (0.2)
Antipsychotic
Alone	18 (1.9)
+ benzodiazepine	5 (0.5)
+ antipsychotic	4 (0.4)
+ antiepileptic	2 (0.2)
+ other^c	3 (0.3)
Benzodiazepine
Alone	61 (6.4)
+ amitriptyline	7 (0.7)
+ mirtazapine	3 (0.3)
+ other^c	2 (0.2)
+ benzodiazepine	2 (0.2)
Antiepileptic alone	10 (1.1)
Other (alone) or with one other medication	54 (5.7)

Women with epilepsy, migraine, or other neurological condition excluded.

Does not include combinations of ≥3 medications.

Other = sleep specific (zolpidem and temazepam), agomelatine, tricyclics, and tetracyclics.

Reported combination therapy.

Multivariable-adjusted analyses revealed that antidepressant use was positively associated with being overweight or obese, single or in a girlfriend/boyfriend relationship compared with being married/de facto, not completing high school, being a current smoker, exercising less than once a week, being a carer, or having been previously diagnosed with cancer (Table 4). The largest odds ratio (2.2) was for a previous diagnosis of cancer. Factors that emerged as protective, in terms of antidepressant use, were being Asian (smallest odds ratio of 0.2) or of “other” ancestry compared with being of European ancestry and being parous.

Table 4.

Unadjusted and Adjusted Logistic Regression Analysis for Predictors of Antidepressant Use

Characteristics	Unadjusted OR (95% CI)	p-Value	Adjusted ^a OR (95% CI)	p-Value
Age (years)
18–<25	1.00
25–<30	1.1 (0.9–1.3)	0.350	—	—
30–<35	1.0 (0.8–1.2)	0.937	—	—
35–<40	1.1 (0.9–1.3)	0.301	—	—
BMI (kg/m²)
<18.5 (underweight)	1.1 (0.8–1.5)	0.568	1.2 (0.8–1.6)	0.422
18.5–<25 (normal weight)	1.00		1.00
25–<30 (overweight)	1.4 (1.2–1.7)	<0.001	1.4 (1.1–1.6)	0.003
≥30 (obese)	2.3 (2.0–2.8)	<0.001	2.0 (1.7–2.4)	<0.001
Ancestry
European	1.00		1.00
Aboriginal/Torres Strait Islander	1.1 (0.6–1.8)	0.820	0.8 (0.5–1.4)	0.542
Asian	0.2 (0.1–0.3)	<0.001	0.2 (0.2–0.3)	<0.001
Other	0.5 (0.4–0.7)	<0.001	0.5 (0.4–0.7)	<0.001
Place of residence
Metropolitan	1.00		1.00
Rural	1.3 (1.1–1.5)	<0.001	1.0 (0.9–1.2)	0.753
Completed high school
Yes	1.00		1.00
No	2.1 (1.6–2.7)	<0.001	1.6 (1.2–2.1)	0.001
Relationship status
Married/de facto	1.00		1.00
Girl-boyfriend/single	1.5 (1.3–1.7)	<0.001	1.4 (1.1–1.7)	0.001
Working/studying or both
Yes	1.00
No	1.1 (0.9–1.4)	0.120	—	—
Parity
0	1.00		1.00
≥1	0.8 (0.7–0.9)	0.036	0.7 (0.6–0.9)	0.002
Living with at least one other adult
Yes	1.00		1.00
No	1.6 (1.3–1.9)	0.001	1.1 (0.9–1.4)	0.314
Housing financial security
Secure	1.00		1.00
Insecure	1.4 (1.2–1.7)	<0.001	1.2 (0.9–1.5)	0.054
Current smoker
No	1.00		1.00
Yes	1.8 (1.5–2.2)	<0.001	1.4 (1.2–1.8)	<0.001
Heavy drinking^b
No	1.00		1.00
Yes	1.4 (1.1–1.7)	0.002	1.2 (0.9–1.5)	0.056
Physical exercise
Once a week minimum	1.00		1.00
<Once a week	1.4 (1.2–1.6)	<0.001	1.4 (1.2–1.6)	<0.001
Carer for special needs person
No	1.00		1.00
Yes	1.5 (1.2–1.9)	0.009	1.4 (1.1–1.8)	0.008
Currently using assisted reproductive technology
No	1.0
Yes	1.2 (0.7–2.0)	0.575	—	—
Diagnosed with any type of cancer
No	1.00		1.00
Yes	2.3 (1.7–3.3)	<0.001	2.2 (1.5–3.1)	<0.001

≥4 drinks on one occasion at least weekly.

Discussion

In this representative sample of young Australian women, one in three women screened positive for moderate to severe depressive symptoms and one in six women, without a neurological condition, was taking a psychotropic medication. In contrast to the global pattern within this age group,¹ depressive symptoms declined with age. That antidepressant use was less prevalent among parous women and women with non-European ancestry are novel findings in the Australian context.

There are few studies with which to compare our findings, particularly with respect to sociodemographic risk factors for depressive symptoms. Two Australian studies reported depressive symptoms in women across the age ranges of 18 to 27 years to be in the order of 20%–30%.^22,23 Analyses of data from face-to-face interviews of Australian women aged 18 to 44 years, conducted in 1997, resulted in a prevalence of depression of ∼10%⁷ and of a major depressive episode in the preceding 30 days of ∼4%.⁶ Consistent with the decline in the proportion of women with depressive symptoms between our youngest and oldest subgroups, Henderson et al. also reported a decline in depression with age in their 1997 study.⁷ This differs from the global trend for the prevalence of depressive symptoms to increase across the reproductive years,¹ indicating regional variations in factors contributing to depressive symptoms and depression.

A novel finding in our study was that being parous was independently associated with a substantially lower risk of depressive symptoms, despite the parous group, including single mothers and women in the early postpartum period, who are at higher risk. The trend is for Australian women to delay childbirth, with the average age of first-time mothers in Australia now 30.7 years,²⁴ compared with 20–22 years in low income countries.²⁵ We hypothesize that the positive effect of parity seen in our study is a function of Australian women having control of their fertility and electing when they wish to conceive.

In relation to having moderate–severe depressive symptoms, being in the category of neither working nor studying was a particular risk factor for women who were unpartnered. The direction of the association cannot be established in our study. It may be that being socially isolated and not employed/studying makes depressive symptoms more likely or that if young women are depressed, they are more likely to lose their social connections and work/study opportunities.

There was commonality among several risk factors for both depressive symptoms and antidepressant use, namely obesity, not completing high school, smoking, being sedentary, being a carer, and having a previous cancer diagnosis. Although none of these findings is surprising, they highlight the sociodemographic and other factors that make young women vulnerable to depression and antidepressant use. The burden of care giving has been extensively documented, although few studies have focused on the impact on younger care providers. We did not find being a care giver to be an independent risk factor for depression in midlife Australian women.⁸ Similarly, a Canadian study reported that younger care providers were substantially more likely to have reduced their exercise, to eat a less healthy diet, and to have increased their alcohol consumption due to care giving than care givers aged 65 years or more.²⁶ In that study, the adverse health impact by caregiving was found to be greater for women than for men. Our findings reinforce that more needs to be done to support young carers in the community.

Although the number of women in our study identifying as being an Aboriginal or Torres Strait Islander was small, that our point estimate indicated 40% of these women had moderate to severe depressive symptoms is shocking. The prevalence of depression among Aboriginal and Torres Strait Islander women is uncertain,²⁷ and our findings highlight the need for the mental welfare of young Indigenous Australian women to be a national health priority.

A concerning proportion of our study participants reported housing insecurity, and almost half of these women had moderate to severe depressive symptoms. Housing insecurity, a known risk factor for depression, was associated with a more than twofold risk of depression.²⁸ Despite this, they were not more likely to be taking an antidepressant. Whether this reflects less utilization of health care services due to access or cost merits investigation. This is a group that needs to be targeted for intervention and support. We found no association between depressive symptoms and residing in urban or rural areas, which contrasts with findings of a European study which found that the prevalence of depressive symptoms was higher in urban areas of all five European countries (United Kingdom, Ireland, Finland, Norway, and Spain) compared with rural settings.²⁹ However, the authors did not report prevalence data stratified by age. We also found no association between undergoing assisted reproduction and depression.

There has been substantial interest in the increase in depressive symptoms in association with the menopause transition.³⁰ However, the proportion of women with moderate to severe depressive symptoms in our study sample is double of what we have reported for Australian women aged 40 to 64 years, assessed by exactly the same criteria.⁸ The women in the latter midlife study were equally likely to be current smokers and heavy drinkers as the young women in the present study, but more likely to be obese. The risk factors we did not assess, such as family history and prior physical, emotional, or sexual abuse, would be operative in both age groups. The noteworthy differences were the much greater proportions of midlife women with protective factors against depressive symptoms. Specifically, a greater proportion of the women at midlife were married or in a de facto relationship (71%) and parous (85%), and fewer (10%) reported housing insecurity.³¹

A high usage of antidepressant therapy in Australia has been attributed to the pharmaceutical marketing of SSRIs and SNRIs in primary care, combined with promotion of awareness of depression among primary care practitioners.^11,32 Only one quarter of the women in our study with moderate to severe depressive symptoms were taking an antidepressant, and many of these may have been taking their antidepressant for conditions such as anxiety, sleep problems, or pain.³³ Even though the BDI questionnaire does not equate to a clinical assessment, our findings suggest that a concerning number of our study participants have depression which is underdiagnosed and untreated. The persistence of moderate to severe symptoms in women taking antidepressants suggests that for many, treatment remains inadequate, including the use of nonpharmacological interventions. The use of benzodiazepines alone or in combination with other psychoactive drugs (13.2%) raises concern about dependence and misuse.³⁴

Women being Asian or “other” ancestry were less likely to report using antidepressants. This may be due to ethno-cultural differences in the expression of depression, with a greater tendency for depression to be manifest as somatic symptoms in Asian cultures as opposed to affective symptoms in Western cultures.³⁵

Lack of personal interviews could be considered a study limitation. However, this was not feasible across the vast geographic area of recruitment required to ensure representativeness. Furthermore, a face-to-face interview may have deterred participation due to both scheduling and confrontation, and women may have been less honest in their responses to highly personal questions. We decided against including questions about physical, emotional, or sexual abuse as the survey was already lengthy and emotionally demanding, such that the inclusion of questions about abuse may have diminished the number of women willing to complete the questionnaire. We also had limited information on comorbidities that may have influenced the likelihood of depression. Being cross-sectional we cannot comment on symptom duration or trends in depressive symptoms or medication use over time and can only report associations and not assume cause and effect. Although our study sample is representative of Australian women of the same age, our findings are not generalizable to countries with different cultural and religious compositions and socioeconomic circumstances. However, our findings highlight the need for similar studies to be undertaken in other settings.

Conclusion

This study reveals that substantial proportions of young Australian women suffer moderate to severe depressive symptoms and are using antidepressant medication. These findings indicate a serious public health concern that may not be unique to Australia. Investment in strategies targeting the identified risk factors is urgently needed, as are resources for adequate clinical care of affected women.

Footnotes

Author Contributions

Study design: S.R.D., R.J.B.; data collection: S.R.D., R.J.B., M.A.S.; data analysis: R.M.I., R.J.B., M.A.S., L.M.; data interpretation: all authors; writing: S.R.D., R.J.B.; article review: all authors.

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Author Disclosure Statement

Dr. Davis reports having received support unrelated to the study content, including: honoraria from Besins Health care and Pfizer Australia, consultancy to Mayne Pharmaceuticals, Lawley Pharmaceuticals, and Que Oncology and has received institutional grant funding for Que Oncology research. No other potential conflicts of interest relevant to this article are reported.

Funding Information

The study was supported by the Grollo-Ruzzene Foundation. Dr. Davis is an NHMRC Senior Principal Research Fellow (Grant No. 1135843); Ms. Marsh is an Australian Rotary Health Scholar.

The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. All authors vouch for the adherence of the study to the protocol, completeness and accuracy of the data, and integrity of analysis and reporting. The senior author affirms that the article is an honest, accurate, and transparent account of the study being reported; no important aspects of the study have been omitted; and there are no discrepancies from the study as planned.

References

World Health Organization. Depression and other common mental disorders: Global health estimates. Switzerland: Geneva, 2017.

Ritchie

, Roser

. “Mental Health”. 2019. Available at: ourworldindata.org Accessed November 7, 2019 .

World Health Organization. Depression. 2018. Available at: https://www.who.int/news-room/fact-sheets/detail/depression Accessed November 7, 2019 .

Kessler

, McGonagle

, Swartz

, Blazer

, Nelson

. Sex and depression in the National Comorbidity Survey. I: Lifetime prevalence, chronicity and recurrence. J Affect Disord, 1993; 29:85–96.

Sia

, Williams

, Pasco

, Jacka

, Brennan-Olsen

, Veerman

. The population mean mood predicts the prevalence of depression in an Australian Context. Aust N Z J Psychiatry, 2018; 52:461–472.

Wilhelm

, Mitchell

, Slade

, Brownhill

, Andrews

. Prevalence and correlates of DSM-IV major depression in an Australian national survey. J Affect Disord, 2003; 75:155–162.

Henderson

, Andrews

, Hall

. Australia's mental health: An overview of the general population survey. Aust N Z J Psychiatry, 2000; 34:197–205.

Worsley

, Bell

, Gartoulla

, Robinson

, Davis

. Moderate-severe vasomotor symptoms are associated with moderate-severe depressive symptoms. J Womens Health, 2017; 26:712–718.

Purewal

, Chapman

SCE

, van den Akker

OBA

. Depression and state anxiety scores during assisted reproductive treatment are associated with outcome: A meta-analysis. Reprod Biomed Online, 2018; 36:646–657.

10.

Smith

, Sketris

, Cooke

, Gardner

, Kisely

, Tett

. A comparison of antidepressant use in Nova Scotia, Canada and Australia. Pharmacoepidemiol Drug Saf, 2008; 17:697–706.

11.

McManus

, Mant

, Mitchell

, Montgomery

, Marley

, Auland

. Recent trends in the use of anti-depressant drugs in Australia. Med J Aust, 2000; 173:458–461.

12.

Stephenson

, Karanges

, McGregor

. Trends in the utilisation of psychotropic medications in Australia from 2000 to 2011. Aust N Z J Psychiatry, 2013; 47:74–87.

13.

Davey

, Chanen

. The unfulfilled promise of the antidepressant medications. Med J Aust, 2016; 204:348–350.

14.

Hollingworth

, Burgess

, Whiteford

. Affective and anxiety disorders: Prevalence, treatment and antidepressant medication use. Aust N Z J Psychiatry, 2010; 44:513–519.

15.

Schofield

, Khan

. Predictors of prescribed medication use for depression, anxiety, stress, and sleep problems in mid-aged Australian women. Soc Psychiatry Psychiatr Epidemiol, 2014; 49:1835–1847.

16.

Paige

, Korda

, Kemp

, Rodgers

, Banks

. Characteristics of antidepressant medication users in a cohort of mid-age and older Australians. Aust N Z J Psychiatry, 2015; 49:275–290.

17.

Skiba

, Bell

, Islam

, Davis

. Challenges in recruitment to an epidemiological study of young Australian women: The Grollo-Ruzzene Foundation Young Women's Health Study. Aust N Z J Public Health, 2019; 43:131–136.

18.

Beck

, Steer

, Brown

. Manual for the Beck Depression Inventory-II. San Antonio, TX: The Psychological Corporation, Harcourt, Brace and Co., 1996.

19.

Richter

, Werner

, Heerlein

, Kraus

, Sauer

. On the validity of the Beck Depression Inventory. A review. Psychopathology, 1998; 31:160–168.

20.

Titov

, Dear

, McMillan

, Anderson

, Zou

, Sunderland

. Psychometric comparison of the PHQ-9 and BDI-II for measuring response during treatment of depression. Cogn Behav Ther, 2011; 40:126–136.

21.

Oliver

, Simmons

. Depression as measured by the DSM-III and the Beck Depression Inventory in an unselected adult-population. J Consult Clin Psychol, 1984; 52:892–898.

22.

Donald

, Dower

, Lucke

, Raphael

. Prevalence of adverse life events, depression and suicidal thoughts and behaviour among a community sample of young people aged 15–24 years. Aust N Z J Public Health, 2001; 25:426–432.

23.

France

, Lee

, Powers

. Correlates of depressive symptoms in a representative sample of young Australian women. Aust Psychol, 2004; 39:228–237.

24.

Births

Australia

, 2010. Canberra, ACT, Australia: Australian Bureau of Statistics, 2011.

25.

Bongaarts

, Mensch

, Blanc

. Trends in the age at reproductive transitions in the developing world: The role of education. Popul Stud (Camb), 2017; 71:139–154.

26.

Ysseldyk

, Kuran

, Powell

, Villeneuve

. Self-reported health impacts of caregiving by age and income among participants of the Canadian 2012 General Social Survey. Health Promot Chronic Dis Prev Can, 2019; 39:169–177.

27.

Black

, Ranmuthugala

, Kondalsamy-Chennakesavan

, Toombs

, Nicholson

, Kisely

. A systematic review: Identifying the prevalence rates of psychiatric disorder in Australia's Indigenous populations. Aust N Z J Psychiatry, 2015; 49:412–429.

28.

Bentley

, Baker

, Aitken

. The ‘double precarity’ of employment insecurity and unaffordable housing and its impact on mental health. Soc Sci Med, 2019; 225:9–16.

29.

Dowrick

, Casey

, Dalgard

, et al. Outcomes of Depression International Network (ODIN). Background, methods and field trials. ODIN Group. Br J Psychiatry, 1998; 172:359–363.

30.

Freeman

, Sammel

, Boorman

, Zhang

. Longitudinal pattern of depressive symptoms around natural menopause. JAMA Psychiatry, 2014; 71:36–43.

31.

Gartoulla

, Bell

, Worsley

, Davis

. Moderate-severely bothersome vasomotor symptoms are associated with lowered psychological general wellbeing in women at midlife. Maturitas, 2015; 81:487–492.

32.

Mant

, Rendle

, Hall

, et al. Making new choices about antidepressants in Australia: The long view 1975–2002. Med J Aust, 2004; 181.

33.

Patten

, Esposito

, Carter

. Reasons for antidepressant prescriptions in Canada. Pharmacoepidemiol Drug Saf, 2007; 16:746–752.

34.

Lader

, Kyriacou

. Withdrawing benzodiazepines in patients with anxiety disorders. Curr Psychiatry Rep, 2016; 18:8.

35.

Kalibatseva

, Leong

. Depression among Asian Americans: Review and recommendations. Depress Res Treat, 2011; 2011:320902.