Approximately 14.8 million US adults (6.7%) experience major depression in a given year (1), and women are 1.7 times as likely to experience depression as men (2). The 12-month prevalence of major depressive disorder, 1 form of clinical depression, for nonpregnant women aged 18 to 50 ranges from 8% to 16% (2,3) and may be highest among low-income women, for whom prevalence of depression (defined as minor depression/dysthymia or major depression) is estimated at 29% (4). Yet, only half of women with depression have ever received a clinical diagnosis of the disorder (4), which is the first step toward treatment and recovery. Major depression is the leading cause of disability in the United States for adults aged 15 to 44 (5) and is associated with diabetes (6), stroke, and coronary heart disease (7). Depression affects a woman's preconception health (8); depression during pregnancy may affect pregnancy outcomes, such as preterm delivery and low birth weight (9-11) and may adversely affect the child's intellectual development, behavior, and mental health (12). In 1 study among women enrolled in Kaiser Permanente Northwest, approximately half of women with clinically diagnosed perinatal depression experienced depression in the 39 weeks before pregnancy (13), highlighting the need for identification and treatment of depression among nonpregnant women of reproductive age.

Nonpsychiatric clinicians have difficulty recognizing depression in their patients (14,15). One meta-analysis found that nonpsychiatric clinicians accurately diagnosed depression in only 36% of depressed patients; another meta-analysis found a rate of 47% (14,15). However, it remains unclear whether routine screening for depression among nonpregnant women of reproductive age, defined as aged 15 to 44 years, improves identification, increases treatment rates, and leads to better overall and preconception health. To consolidate current knowledge in this area, we conducted a systematic review of the literature to address 5 key questions specific to this group of women: 1) What are the current national clinical practice recommendations and guidelines for depression screening; 2) What are the prevalence and predictors of screening; 3) How well do screening tools detect depression; 4) Does screening lead to diagnosis, treatment, and improved outcomes; and 5) What are the most effective treatment methods?

We identified recommendations and guidelines from 5 national groups to answer question 1, and we identified 16 studies that met our final review criteria for questions 2 through 5. For question 2, we found 4 individual studies; for question 3, 1 meta-analysis; for question 4, 2 meta-analyses and 1 systematic review; and for question 5, 5 meta-analyses and 3 post-hoc analyses of pooled data.

Question 1. What are the current national clinical practice recommendations and guidelines for depression screening?

We identified recommendations and guidelines from 5 national groups for screening adults for depression (16-20); however, none provided specific guidance for nonpregnant women of reproductive age (Table 2). The US Preventive Services Task Force (USPSTF) recommends that providers screen adults for depression only when staff-assisted depression care supports are in place to ensure accurate diagnosis, effective treatment, and follow-up (17). The USPSTF defines "staff-assisted care supports" as clinical staff who assist the primary care clinician by providing some direct depression care or coordination, case management, or mental health treatment. The American Academy of Family Physicians follows USPSTF guidelines for screening all adult patients for depression, using the clinician's choice of screening method (19). The US Department of Veterans Affairs (VA) recommends annual depression screening for adult patients in primary care but notes the USPSTF recommendation of screening only when staff-assisted supports are in place (18). The American College of Obstetricians and Gynecologists (ACOG) does not advocate for or against screening for depression during well-woman care, but states that when clinicians identify a woman with depression, they must provide follow-up care if they do not refer her for care elsewhere (16). The American Academy of Pediatrics (AAP) recommends screening mothers for depression at the 1-, 2-, 4- and 6-month well-child visits and beyond the postpartum period (20). The American Psychiatric Association has no published guidelines specific to depression screening, but in its guidelines on treatment for major depressive disorder, it acknowledges that primary care physicians, obstetricians, and physicians of other disciplines may screen for depression and initiate treatment for patients (21). The USPSTF, VA, ACOG, and AAP mention the 2-item Patient Health Questionnaire (PHQ-2), as an example of a short, standardized screener for depression.

Question 2. What are the prevalence and predictors of screening?

We did not find any studies that assessed prevalence and predictors among nonpregnant women of reproductive age only. Four individual studies reported on prevalence rates of screening for depression among adult women of any age in primary care settings (22,23) or by providers serving women of reproductive age in obstetric practices (24,25). Desai et al evaluated the VA experience in implementing universal screening, using a validated instrument, for depression (22). This study included 21,000 people receiving care at VA facilities across the United States in 2002. More than 84% of women of any age and almost 80% of women and men aged 45 or younger were screened for depression. Among patients of all ages, no difference was found in screening rates between women and men, but patients were less likely to be screened if they were younger, unmarried, had greater service disability, or had medical comorbidity.

Tudiver et al examined screening rates for 615 adult women aged 21 to 89 years accessing primary care in 2003 at 19 rural health clinics in the United States (23). Rates of screening were low; 2.4% of visits documented formal screening (ie, through use of a validated instrument), and 33.2% of visits documented informal screening (ie, depression questions noted without mention of a screening tool).

Two studies examined the prevalence of screening for depression among obstetrician-gynecologists (OB/GYNs) (24,25). LaRocco-Cockburn et al surveyed practicing OB/GYNs in Washington State in 2001 (25). Of the 282 (56%) who returned the survey, 44% reported often or always screening for depression in patients, 41% reported screening sometimes, and 15% reported never screening for depression regardless of signs or symptoms. In this study, 81% of OB/GYNs used their own questions about mood or mental health, 32% used a validated screening tool (specific tool not reported), 16% used a validated patient self-report paper-and-pencil test, and 7% used a structured clinical interview.

Dietrich et al conducted a cross-sectional survey among 437 randomly selected US OB/GYNs (response rate, 58.3%) who had completed residency training in the previous 5 years and currently provided care (24). Approximately 40% of recent graduates and 50% of residents reported that depression was included on their practice encounter form. Only 9% to 12% reported routinely asking about depression or using a screening questionnaire to identify major or minor depression. The most common reasons for recognition of depression by OB/GYNs were that the patient appeared distressed (38%), presented with a symptom (34%), or introduced the topic directly (26%).

Question 3. How well do screening tools detect depression?

We did not find any studies that assessed performance of depression screening tools among nonpregnant women of reproductive age only. Several short (1-2 questions) and longer (up to 30 questions) validated tools are available to screen for depression in primary care populations (26,27). The PHQ-2, which asks, “Over the past 2 weeks, how often have you been bothered by a) little interest or pleasure in doing things and b) feeling down, depressed, or hopeless,” may perform as well as longer tools (28) or could be used as an initial screen, with a longer screen administered to patients with affirmative answers to both questions (18). Depression-specific instruments may help clinicians recognize depression more easily than instruments that measure multiple mental health conditions (29). With any screening tool, a diagnostic interview is needed to confirm the presence of depression.

We found 1 meta-analysis of the performance of instruments among adult men and women in primary care settings (30). It examined performance of case-finding instruments used for routine screening in primary care populations of men and women of all ages (30). This article evaluated only 21 of the 38 studies identified (because of study limitations); the 21 studies examined 16 validated case-finding instruments. For detecting major depression in a primary care population using routine screening with a validated case-finding tool, the median sensitivity was 85% (range, 50%-97%), the median specificity was 74% (range, 51%-98%), and no statistically significant differences between instruments were found. The study concluded that several case-finding instruments are feasible to use in primary care settings and that the instruments perform sufficiently to facilitate identification of depression.

Question 4. Does screening lead to diagnosis, treatment, and improved outcomes?

We did not find any systematic reviews or individual studies that addressed this question among nonpregnant women of reproductive age only. We found 2 meta-analyses (29,31) and 1 systematic review (32) that examined this question among the general population in primary care and hospital settings. Two studies (31,32) were conducted to inform and update USPSTF guidelines on screening for depression in adults. Pignone et al reviewed randomized trials published between January 1994 and August 2001 and included 14 studies in primary care settings that examined the effect of screening patients for depression on identification, treatment, and health outcomes (31). The included trials examined a range of screening intervention strategies, including feedback of screening scores, feedback and general education of providers, feedback and treatment advice, and integrated recognition and management approaches with coordinated follow-up of diagnosis and treatment. Screening resulted in a 2- to 3-fold increase in clinicians' recognition of depression. However, in a comparison of absolute differences in proportions treated, the effect of screening on rates of treatment was mixed; 4 studies found positive effects and 5 studies found no effect. Increases in rates of treatment generally resulted in increases in prescriptions for antidepressants rather than referrals to mental health professionals. Three out of 7 studies included in the meta-analysis found significant improvement in depression between groups screened for depression and groups not screened. The meta-analysis found that patient and provider characteristics, use of particular outcome measures, follow-up time, or trial quality did not explain the mixed findings; however, insufficient power may explain the results of some negative trials. A meta-analysis of the 7 studies showed that screening with or without further intervention was associated with a 13% (95% confidence interval [CI], 5%-21%) reduction in risk of remaining depressed. Additionally, variations in interventions limited their interpretation of findings.

O'Connor et al (32) published an update to the study by Pignone et al. The update included studies published from January 1998 to December 2007 on randomized controlled trials (RCTs) conducted in primary care settings among the general adult population. The authors found 2 good-quality and 2 fair-quality RCTs not included in the study by Pignone et al. The updated review supported the original findings, that primary care depression screening may be effective when the treating physician works with other staff who provide part of the depression care, such as assessment and monitoring, or when extra efforts are made to enroll patients in mental health specialty care.

The study by Gilbody et al (29) examined the effect of screening using a standardized depression screening or outcome assessment instrument alone, without substantial organizational enhancements (such as clinician education, nurse case management, and integration between primary and secondary care), on recognition of depression and improvement in outcomes among nonpsychiatric patients in primary care and hospital settings. The study identified 16 RCTs comparing usual care with routine screening administered by research staff and feedback of results to clinicians. The RCTs did not report effects separately among women. Eleven of the 16 examined the effect of screening on the clinician's recognition of depression. Seven of the 16 were conducted among the general population and found screening was not associated with increased likelihood of recognition of depression (relative risk [RR], 1.03, 95% CI, 0.85-1.24). In 4 trials conducted among high-risk populations, screening increased the likelihood of recognition of depression by 67%. Ten of the 16 studies examined the effect of depression screening on management of depression. Screening marginally increased the likelihood of the patient receiving any intervention for depression (RR, 1.30, 95% CI, 0.97-1.76) with no difference between high-risk and general primary care populations.

In pooled data from 5 studies that examined the effect of screening on depression outcomes, the meta-analysis found no effect (standardized mean difference, −0.02; 95% CI, −0.25 to 0.20) (29). The authors conclude that use of screening instruments alone in unselected populations within primary care and hospital settings, without organizational enhancements, does not improve rates of depression treatment or outcomes. They state that routine screening among high-risk populations may be more effective.

Question 5. What are the most effective treatment methods

We found 5 meta-analyses and no individual studies that met the inclusion criteria for nonpharmacologic treatments and 3 post-hoc analyses of pooled data that examined the effectiveness and safety of SSRIs and SNRIs in women.

Three meta-analyses explored the association between exercise and depression and evaluated differences by sex (33-35) (Table 3). All found that exercise reduced mean depression scores (effect size [ES], 0.53-0.80) with no difference by sex. North et al (35) reviewed 80 studies of any design, 16 of which included women only. Exercise was defined as aerobic exercise and muscular strength-building. The overall ES was −0.53. Craft et al (33) included 30 studies of any design, 4 that included women only and included aerobic and resistance exercise. They found a greater reduction in depression symptoms among subjects with more severe depression and those undergoing longer interventions.

In the third meta-analysis, Rethorst et al (34) reviewed 58 published RCTs, 7 of which included women only, and evaluated moderate to vigorous exercise as a treatment for depression. Controls either received no treatment or were on a wait list for treatment. Exercise was found to be beneficial in studies of participants clinically diagnosed with depression (17 studies) and among participants not clinically diagnosed with depression (40 studies), although the ES was greater among clinically diagnosed samples (ES, −1.03 for clinically diagnosed and ES, −0.59 for not clinically diagnosed).

We found 2 meta-analyses (36,37) that evaluated different types of psychotherapy (cognitive therapy [CT], individual versus group psychotherapy, and short-term psychodynamic psychotherapy) by sex (Table 3). Overall, the meta-analyses consistently found psychotherapy more effective than no treatment. Robinson et al (37) analyzed 58 studies published during 1976 through 1986, 29 of which focused on depression outcome measures. The combined sample was 80% female with an average age of 39.4 years. Psychotherapy resulted in lower mean depression scores compared with no treatment (ES = 0.93). No differences were found by sex. Gloaguen et al (36) conducted a meta-analysis of 48 RCTs to evaluate the effectiveness of CT for treatment of major depression or dysthymic disorder. They found that CT was more effective than antidepressants, and there were no differences by sex.

Three articles examined the effectiveness and safety of SSRIs and SNRIs in women, and all were post-hoc analyses of pooled data (Table 3). Two studies examined the effect of SSRIs on level of depression (38,39). Khan et al (39) found rates of response (65% vs 40%, P < .001) and remission (45% vs 14%, P < .001) greater for women taking SSRIs than placebo. Entsuah et al (38) found higher rates of depression absence (31% vs 20%) and remission (34% vs 24%) among women taking SSRIs compared with those taking placebo (P < .05). Among participants aged 40 years or younger, differences in rates of remission and absence of depression between those taking SSRIs and those taking placebo did not reach statistical significance (38). Three studies examined the effects of different SNRIs on depression among women (38-40) and found greater response and remission rates among women taking SNRIs compared with those taking placebo (38-40).

Our systematic review of the literature identified several gaps in knowledge of screening, detection, and treatment of depression among nonpregnant women of reproductive age. Current USPSTF screening guidelines do not recommend universal screening for depression among adults unless staff supports are in place to diagnose, treat, and follow up patients (17). It is unclear what percentage of clinical practices serving nonpregnant women of reproductive age fit these criteria. The percentage of women who are screened and the percentage of providers who screen nonpregnant women for depression are largely unknown. The limited data available suggest low screening rates, although prevalence of depression is high. Several validated screening tools exist and perform equally well in primary care settings. However, we found no studies examining the performance of the screening tools specifically among nonpregnant women of reproductive age.

Additionally, no studies were found that examined the effect of depression screening on diagnosis, treatment, and outcomes specifically among nonpregnant women of reproductive age. Studies suggest that screening in high-risk populations may be effective in clinician recognition of depression and patient receipt of treatment (29). Women of reproductive age may be considered high risk, especially low-income women attending public family planning and OB/GYN clinics, where, in 3 studies from different parts of the United States, 19% to 48% of women screened positive for moderate to severe or clinically relevant depression (41-43). Engaging low-income women in treatment can improve depressive symptoms; however, it takes considerable time and resources, and engagement is difficult to achieve, even when treatment is free and child care and transportation are provided (44).

Various treatment options exist for depressed women, including exercise, psychotherapy, and pharmacotherapy. In England, exercise is a first-line therapy (45), but not in the United States. All treatments reviewed performed better than placebo; however, treatment response rates were low. Further research is needed to identify whether combining treatments can further improve effectiveness.

In summary, although studies have documented high rates of depression among women of reproductive age and significant detrimental health effects for them and their families, how to engage women in effective treatment and how to create systems of affordable and acceptable care remain future challenges. Recent changes in health insurance through national health care reform may provide insurance coverage for more people with mental illnesses. Monitoring access and use among women of reproductive age will help evaluate whether this policy improves mental health in the United States.