Beyond Symptoms: Race and Gender Predict Anxiety Disorder Diagnosis

Abstract

Research shows an unequal distribution of anxiety disorder symptoms and diagnoses across social groups. Bridging stress process theory and the sociology of diagnosis and drawing on the National Longitudinal Study of Adolescent to Adult Health, we examine inequity in the prevalence of anxiety symptoms versus diagnosis across social groups (the “symptom-to-diagnoses gap”). Bivariate findings suggest that while several disadvantaged groups are more likely to experience symptoms of anxiety, they are not more likely to receive a diagnosis. Multivariate results indicate that after controlling for anxiety symptoms: (1) Being female still predicts an anxiety disorder diagnosis, and (2) Native American, white, and Hispanic/Latino respondents are more likely than black respondents to receive an anxiety disorder diagnosis. We conclude by reflecting on the implications of race and gender bias in diagnosis and the health trajectories for persons with undiagnosed anxiety disorders.

Keywords

anxiety race gender inequality diagnosis

Anxiety is a common and often debilitating disorder. Nearly one–third of the US population has met the diagnostic criteria for an anxiety disorder in their lifetime, and about 18 percent of the population has met the criteria in the previous 12 months (Kessler, Berglund, et al. 2005; Kessler, Chiu, et al. 2005; see also Kessler et al. 2012). The onset of anxiety is early, impacting people from very young ages, with the median age of symptom onset at 11 years old (Kessler, Berglund, et al. 2005) and the lifetime prevalence of “severe” anxiety disorders of 8.9 percent among 13- to 18-year-olds (Merikangas et al. 2010). These statistics suggest that a substantial portion of the population experiences severe anxiety and from early ages.

Anxiety disorders have significant personal and social consequences. For instance, anxiety disorders are associated with an inability to carry out daily tasks (e.g., grocery shopping and housework) and decreased work productivity (see Hoffman, Dukes, and Wittchen 2008; Wittchen 2002). Likewise, anxiety disorders are associated with high rates of unemployment, lower levels of educational achievement, higher rates of marital strife, and impaired social skills (Ansseau et al. 2008; Davidoff et al. 2012; Shearer 2007). People with anxiety disorders also are more likely to experience depression and attempt suicide (Bauer et al. 2005; Nutt et al. 2002; Sareen et al. 2005; Shearer 2007; Simon et al. 2004). A meta–analysis of 23 studies showed that individuals with an anxiety disorder have significantly lower quality of life, specifically in the domains of mental health and social functioning (Olatunji, Cisler, and Tolin 2007). Research also shows that before diagnosis, individuals with anxiety disorders utilize the medical system in expensive and ineffective ways—such as emergency room care for panic attacks and cardiology specialists for racing heart or heart palpitations (Deacon, Lickel, and Abramowitz 2008).

Anxiety disorders pose meaningful personal, social, and economic challenges that are not experienced equally or equitably in the population. Women are twice as likely as men to meet the criteria for an anxiety disorder (Kessler, Berglund, et al. 2005), and research shows complex relationships between race and anxiety symptoms (Kessler, Chiu, et al. 2005). In addition to the unequal distribution of symptoms, sociologists of diagnosis emphasize the potential for the unequal distribution of diagnosis (Jutel 2009a, 2009b; Jutel and Conrad 2014; Jutel and Nettleton 2011; Nettleton 2006). Diagnosis is an important social and medical tool that is not equally accessible across social groups, which represents a critical form of social inequality. In this article, we study the unequal distribution of anxiety symptoms and diagnosis across racial groups and genders. We explore potential under- and overdiagnosis of anxiety disorder across social groups.

Theoretical Framework

Stress Process

The burden of anxiety is not evenly distributed across the population. Stress process theory is one sociological framework—supported by substantial research—that helps explain why some groups are more vulnerable to disorders, such as anxiety, than others. We know that various social categories of people are more likely to meet the criteria for an anxiety disorder and that some groups face greater barriers to diagnosis and subsequent access to care (Abbo et al. 2013; Everett, Onge, and Mollborn 2016; Hines-Martin et al. 2003; McLean et al. 2011; Ozer et al. 2017). Stress process theory offers a framework for understanding this uneven distribution by assessing which social factors contribute to health or disease. For instance, in a foundational social epidemiological study, Pappas and colleagues (1993) draw on two sets of nationally representative survey data to assess the relationship between socioeconomic status (SES) and mortality. They find an inverse relationship between SES and mortality; in other words, higher levels of income and social status are strongly associated with lower rates of death. Research in this field has linked social factors, such as race, gender, marital status, employment, and geographical location, to a wide variety of health outcomes (Latkin and Curry 2003; Link and Phelan 1995; Ross 2000; Schieman, Whitestone, and Van Gundy 2006; Turner, Wheaton, and Lloyd 1995). This body of research has revealed how social inequality contributes to unequal levels of stress, which give way to an unequal distribution of morbidity and mortality.

Thoits (2010) offers a clear summary of what stress process research has established. First, when stressors (e.g., major life events and traumas) are well measured, they have a clear and strong impact on both mental and physical health. Second, groups’ exposure to various stressful events helps explain inequality in mental and physical health. Third, minority groups experience and suffer the consequences of discrimination stress, which is the stress derived from serious forms of unfair treatment or discrimination. Fourth, there is a transgenerational passing of stressors that reinforces and strengthens health inequalities. Lastly, there is a buffering effect against stressors by factors like social support (Thoits 2010).

Following this line of research, we expect that disadvantaged groups—like women and racial minorities—will be more likely to meet the criteria for anxiety disorders.¹ This certainly holds true for gender. Women are twice as likely as men to meet the criteria for an anxiety disorder (Kessler, Berglund, et al. 2005), and research shows that this gendered difference emerges by age six (Lewinsohn et al. 1998). However, it remains an open question whether the higher rates of symptoms among women fully explain their higher rates of diagnosis. Research on race demonstrates an even more complicated relationship with anxiety. Despite well–documented discrimination stress (Everett et al. 2016; Pascoe and Richman 2009; Taylor and Turner 2002; Williams et al. 1997), non-Hispanic black populations are actually 20 percent less likely to have an anxiety disorder (Kessler, Chiu, et al. 2005). The relationship between race and mental health is further complicated by poverty. As we would expect, research shows that poverty is associated with higher rates of anxiety; research also shows, however, that after adjusting for poverty, nonwhite populations are less likely to report mental health symptoms (Samaan 2010; see also Chow, Jaffee, and Snowden 2003). In this study, we explore those social factors that predict key anxiety symptoms and then the diagnosis of an anxiety disorder over time—with a particular focus on gender and race. Importantly, one of the limitations of the stress process model is that it is typically tested using symptomology rather than diagnosis (Pearlin et al. 1981; see also Haley et al. 2003; Perry, Harp, and Oser 2013; Turner and Lloyd 1999; Turney, Wildeman, and Schnittker 2012). In this study, we take an initial step of locating which social groups show a disjuncture between anxiety symptoms and anxiety disorder diagnosis.

We seek to more precisely assess the uneven distribution of anxiety; to do this, we make the empirical distinction between symptoms (individual mental, emotional, and physical experience) and diagnosis (medical acknowledgement and labeling). There is no reason to believe that the diagnostic process works the same across different social categories even when they experience similar types and intensities of symptoms. For instance, sociological research established that the accurate diagnosis of a condition as assumedly clear as a heart attack has been complicated by the gender of the patient, with heart attacks among women being less likely to be recognized by physicians (McKinlay 1996). A more recent study found that implicit racial bias among physicians (against black patients) predicted lower rates of diagnosis and treatment for black patients with acute coronary syndromes (Green et al. 2007). While these are two specific studies, they illustrate the critical importance of social conditions, contexts, and categories in the diagnostic process. Researchers across disciplines have explored reasons why the diagnostic process is uneven across social groups. Proposed explanations include: limited or culturally biased diagnostic tools, individual provider bias, providers’ lack of technical knowledge or cultural competence, limited patient disclosure, different treatment–seeking behaviors across groups, and barriers to care and to various types of care (see Chapman, Tashkin, and Pye 2001; Croskerry 2003; Graber, Franklin, and Gordon 2005; Whiting et al. 2004).

While stress process theory offers a sociological framework for understanding the unequal distribution of disease, the sociology of diagnosis places emphasis on the importance of diagnosis, which is a potential location to explore a second source of inequality. Receiving a diagnosis is a critical and essential component of accessing care in the modern medical system. Without a diagnosis, the medical system leaves people with distressing symptoms and no explanation or access to medical treatment (Brown 1990, 1995; see also Blaxter 1978). Diagnosis allows individuals to make sense of pain and distress; it provides the opportunity for social inclusion and access to medical care and treatment (Nettleton 2006). Indeed, diagnosis is an important social and medical tool that is not equally accessible across social groups, which represents a critical form of social inequality. It is important to note that stress process theorists, along with other medical sociologists, have long noted methodological and conceptual incongruity between actual symptoms and established diagnostic categories (Klerman 1989; Mirowsky and Ross 1989). While stress process research helps us understand disjuncture in symptoms and diagnosis overall, there is still much work to be done to identify and explain the uneven distribution of symptom and then diagnosis across social groups.

In this study, we explore which social groups are more likely to experience the symptoms of anxiety—which stress process theory helps us understand—and which groups are more likely to receive a diagnosis. We anticipate that certain groups will experience inequality doubly: being at an increased risk of experiencing anxiety symptoms and being less likely to receive a diagnosis—limiting access to appropriate care.

Medicalization

Thus far, we have emphasized the potential underdiagnosis across social groups. Now we turn to the other end of the spectrum. Medicalization research provides a lens into overdiagnosis or often, the social construction of medical conditions. This line of research challenges the expanding reach of the medical field. Medicalization literally means “to make medical” (Conrad 1992). Accordingly, medicalization research addresses how phenomena come under the purview of medicine. One key critical insight offered by medicalization research is that medicine individualizes social problems. Peter Conrad (1975:12) observes that as a society, “We tend to look for causes and solutions to complex social problems in the individual rather that in the social system.” That is, social problems are “medicalized” (or made medical) and in turn treated as patient problems. Conrad (1975:19) continues, “Rather than seeing certain behavior as symptomatic of problems in the social system, the medical perspective focuses on the individual diagnosing and treating the illness, generally ignoring the social situation.” Similarly, Scheper-Hughes (1993) explains that illness obfuscates important social arrangements and power relations; more specifically, illness excuses the powerful from their role in the suffering of the less powerful. She writes, “A sick body implicates no one. Such is the special privilege of sickness as a neutral social role, its exceptive status. In sickness there is (ideally) no blame, no guilt, no responsibility . . . society and its ‘sickening’ social relations are gotten off the hook” (Scheper-Hughes 1993:174).

Medicalization research has also thoroughly investigated the organizational and capitalistic foundations of illness. Researchers have shown how seemingly “natural” illnesses are often the creations of corporations. For instance, researchers have linked the contemporary surge in medical diagnoses to the advent of new medications. In a now classic piece on medicalization, Conrad (1975) linked the “discovery” of hyperkinesis (contemporarily, attention deficit disorder) to the development of a biomedical treatment. Specifically, researchers were surprised to find that amphetamines helped calm a subset of rowdy and disobedient school children, which prompted medical research that identified a set of behavioral problems among children that could be treated with a family of synthetic medications, including Ritalin. Hyperkinesis quickly became among the most prevalent childhood psychiatric diagnoses, which Conrad linked to relentless promotion from the pharmaceutical industry and support from the government. This “discovery” and then expansion of disorder represents an institutionally derived “putting the cart before the horse”; in other words, the “disorder” was identified by its unexpected (and then, profitable) treatment. Medicalization researchers are often critical of this approach to developing diagnostic categories, demonstrating that this process creates illness where there was once none.

Medicalization research also offers another, related critique: that social institutions contribute to the overmedicalization of normal experience (Christopher 2007; Szasz 2007) and deviant behavior (Conrad and Schneider 2010). This strain of research reveals how medicine serves as a “boundary manager” (Rosenberg 2006:416) that is able to redefine “normal” behavior and experience as “abnormal” (Brown 1995:39; Chiong 2001; Jutel 2009b) and suggests that this degree of definitional control bestows unacceptable amounts of power on medical institutions and industries. In this study, we apply this framework to consider the symptom–to–diagnosis gap, which is when a group of people have a higher or lower than expected rate of diagnosis than would be expected based on symptom levels in that group.

Study aims

Taking the sociology of diagnosis and medicalization research into account, we are cautious about making claims about “overdiagnosis” and “underdiagnosis.” We explore more descriptively which social groups have lower or higher rates of diagnosis (1) as compared to each other and (2) considering the level of symptoms they experience.

Research questions

Research Question 1: Do respondents who endorse high levels of anxiety symptoms (physical symptomology) and feeling fearful (named fear) also have higher rates of anxiety disorder diagnosis compared to those who report lower levels of anxiety symptoms and feeling fearful?

Research Question 2: Are respondents in certain social groups more at risk for anxiety symptoms (physical symptomology), feeling fearful (named fear), and anxiety disorder diagnosis?

Research Question 3: In what social groups, if any, do we see a gap between symptoms (physical symptomology and named fear) and anxiety disorder diagnosis?

Research Question 3a: Do some social groups experience higher rates of anxiety symptoms (physical symptomology) and feeling fear (named fear) but a lower likelihood to receive an anxiety disorder diagnosis?

Research Question 3b: Conversely, do some social groups experience lower rates of anxiety symptoms (physical symptomology) and feeling fear (named fear) but a higher likelihood to receive an anxiety disorder diagnosis?

Research Question 4: When adjusting for known predictors of anxiety disorders, what are the most important factors for receiving or not receiving an anxiety disorder diagnosis?

Methods

Researchers assess whether rates of diagnosis are more or less than expected in several ways. One approach is to compare rates of diagnosis in study samples to what one would expect in the population (Connolly et al. 2011). A second approach is to compare individual results from a medical screener or diagnostic tool compared to the same individual’s diagnosis in his or her medical chart (Asch et al. 2003; Sorkin et al. 2011). Yet another approach is to gather physician diagnoses based on patient vignettes with symptom presentations (Leddy et al. 2011). Indeed, researchers have been creative in how they have studied diagnosis of various disorders and among social groups. To our knowledge, however, there is no research on differences between key symptoms and diagnosis over time with a sample that is generalizable to the population. This is one of our contributions to the literature.

Data

This study uses four waves of the restricted sample from the National Longitudinal Study of Adolescent to Adult Health (Add Health). Add Health is a nationally representative sample of high school youth (9th–12th grade) in the 1994–1995 school years. Following initial in–school interviews, in–home interviews were completed with a subset of the youth and repeated for a total of four waves. Data for the fourth wave were collected in 2008 when the participants were between 24 and 32 years old. We use a sample of adolescents to young adults because of the research that shows an early onset of symptoms. After excluding those without data in any of the four waves, we have an analytic sample of 9,421 individuals. Additional information on the Add Health data sampling and data collection can be found at the University of North Carolina Population Center (http://www.cpc.unc.edu/projects/addhealth).

Measures

Anxiety

Anxiety is measured in three ways in this study, as: anxiety disorder diagnosis, physical symptomology, and named fear.

Anxiety disorder diagnosis measures whether a respondent received an anxiety disorder diagnosis within a certain time period. Respondents were asked, retrospectively, about anxiety diagnoses in the Wave 4 interview (“Has a doctor, nurse, or other health care provider ever told you that you have or had: anxiety or panic disorder?”). If the respondent endorsed this question, they were asked at what age they were diagnosed (“How old were you when the doctor, nurse, or other health practitioner diagnosed you with anxiety or panic disorder?”). To approximate time order, new variables were computed to determine if the respondent was diagnosed after Wave 2 and before Wave 3 using the respondent age at each wave variable. In Wave 2, the mean age was 15.95, ranging from 12 to 22, and in Wave 3, the mean age was 21.39, ranging from 18 to 27 years old. Any youth with diagnoses prior to Wave 2 and after Wave 3 were excluded from the analysis to account for time order of symptomology and diagnosis.

Physical symptomology is measured using an additive composite scale of endorsements of feeling key symptoms of anxiety every day or often, including: feeling hot all over, having stomachaches, experiencing cold sweats, or feeling physically weak (collected in Wave 2 during the in–home interview). The variable was then recoded into a binary variable with the highest decile of symptoms coded as 1 and all others as 0.² Coding for the extreme cases parallels diagnostic criteria that emphasize high diagnostic thresholds for diagnosis. A polychoric factor analysis confirmed that these four symptoms were measuring a single underlying factor with an eigenvalue of 1.31. While the original items do not represent DSM criteria for any specific anxiety disorder, they do represent common symptoms of anxiety; additionally, in our analysis, we assess the relationship between this physical symptomology variable and anxiety disorder diagnosis to establish criterion–related validity.

Named fear is measured using a single item measured in Wave 2, “How often have you felt fearful in the past seven days.” The responses were recoded, with a lot or most of the time coded as 1 and all other responses (never or rarely and sometimes) as 0. This was recoded from an ordinal level variable into a dichotomous variable to reflect those who endorsed high levels of fear. Analyses indicate that the two measures of anxiety symptom (physical and named) are significantly related but not measuring the same underlying construct (Pearson’s correlation coefficient = .22). At a conceptual level, we included physical symptomology to capture the more difficult to diagnose somatic presentations of anxiety and named fear to capture those instances when diagnosis should theoretically be more straightforward (with the active identification of fear).

Established predictors of anxiety

We also include variables in this analysis that are known to be important in both symptomology and diagnosis of anxiety disorders (see Abbo et al. 2013; Caughy, O’Campo, and Muntaner 2003; Everett et al. 2016; Fernandes and Osório 2015; Hovens et al. 2010; McLean et al. 2011; Ozer et al. 2017; Samaan 2010). These variables include gender, race, ethnicity, poverty, housing stability, violence exposure, and perceived safety in neighborhood. All respondents were asked these questions at Wave 1 and/or 2 during the in–home interview.

The focus of our analyses is on gender and race. Gender was measured in Wave 1 with potential response categories of male (0) or female (1). Race was measured in accordance with guidance from the Carolina Population Center (http://www.cpc.unc.edu/projects/addhealth/faqs/aboutdata). Race was measured as six racial/ethnic categories, including: Hispanic, black, Asian or Pacific Islander, Native American, other races, and white. The race variable (“What is your race? Select all that apply”) and Hispanic/Latino ethnicity variable (“Are you of Hispanic or Latino origin?”) were combined into a single race variable with only one race designation per respondent. Consistent with the recommended use of these measures, if the respondent answered yes to the Hispanic/Latino ethnicity question, that respondent was given a race designation of Hispanic and excluded from any race category that was selected for the race variable. If the respondent selected black or African American (only) or black or African American and any other race (Asian, Native American, other, or white) but also answered no to the Hispanic or Latino question, they were designated as black or African American in the recoded race variable and excluded from the other marked categories. We repeated this recoding procedure (in order) for all remaining race categories: Asian, Native American, other, and white.

Poverty was measured during the parent interview in Wave 1 with a single question: “Are you receiving public assistance, such as welfare?” Housing stability was measured using a single question: “Think about the house or apartment building in which you lived in January 1990, when you were {AGE IN JANUARY 1990} years old. Do you still live there?,” with potential responses of yes (1) and no (0). Violence exposure was measured using a single item, which asked the youth: “During the past 12 months, how often did the following happen? You saw someone shoot or stab another person.” The response categories were recoded into never (0) and at least once (1). Violence exposure was asked in Waves 1 and 2, and the responses were combined for the analysis. Perceived neighborhood safety was measured in Wave 2 with a single question, “Do you usually feel safe in your neighborhood?,” with responses of yes or no.

Protective factors

Additionally, we included key known protective factors, including social support and general health. Social support was measured in three ways: teacher, family, and peer support (from the Wave 1 interview): “How much do you feel your (teachers, parents, and then friends) care about you?” Responses included not at all (1) through very much (5). General health was measured with a single question asking, “In general, how is your health?,” with response options from poor (1) to excellent (5), which was then recoded into fair or poor = 0 and all other responses = 1.

Control variables

Lastly, other important controls included respondent age and a proxy for access to medical care. Age was calculated by subtracting the respondents’ birth date from the interview date. Access to medical care was measured through responses to a single question, “In the past year have you had a routine physical examination?”

Analytic Approach

Descriptive univariate and bivariate statistics (χ² and correlation) were examined to answer Research Questions 1 through 3. For Research Question 4, weighted logistic regression models were used. The weighted logistic regression models were entered in a step–wise manner to examine the impact of important variables. For all analyses, we used the post–stratification longitudinal weights recommended for models involving individual–level analyses across all four waves of data.

Results

Less than 4 percent of respondents were diagnosed with an anxiety disorder between Waves 2 and 3 of this study (anxiety disorder diagnosis; Table 1). Nearly 10 percent of respondents were categorized as have high levels of anxiety symptomology, and another 3.55 percent endorsed feeling fearful (named fear).

Table 1.

Characteristics of the Sample (N = 9,421).

Variable	Weighted Percentage	Unweighted (N)
Anxiety disorder diagnosis	3.62	303
Physical symptomology	9.90	947
Named fear	3.55	316
Gender: female	49.98	5,142
Race
Hispanic	11.99	1,448
Black	15.50	1,969
Asian	3.62	615
Native American	2.09	177
Other	0.88	73
White	65.92	5,134
Hispanic
Poverty	9.28	720
Stable housing	55.32	5,278
Violence exposure	14.93	1,473
Safe neighborhood	89.53	8,348
Fair or poor health	7.21	691
Age (Wave 1)
11–13	21.54	1,589
14–15	38.44	3,445
16–17	33.33	3,678
18–21	6.62	704
Access	65.77	6,142
	Mean	SD
Social support (scale 1–5)
Teacher	3.53	.99
Parent	4.80	.54
Peer	4.24	.79

Table 2 shows the results of bivariate analyses for Research Question 1. As expected, respondents with high levels of physical symptomology are significantly more likely to receive an anxiety disorder diagnosis (9.1 percent compared to 3.0 percent, respectively). Also consistent with expectations, those who endorsed feeling fearful had higher rates of diagnosis (8.7 percent) compared with those who did not endorse feeling fearful (3.4 percent).

Table 2.

Weighted Bivariate (χ²) Analyses Examining the Relationship between Anxiety Symptoms, Feeling Fearful, and Anxiety Disorder Diagnosis with Other Relevant Factors.

Variables	Physical Symptomology (%)		Named Fear (%)		Anxiety Disorder Diagnosis (%)
Physical symptomology
Top 10			10.8	***	9.1	***
Bottom 90			2.7	***	3.0	***
Named fear
Yes	30.2	***			8.7	***
No	9.1	***			3.4
Gender
Female	12.6	***	4.2	**	5.7	***
Male	7.3	***	2.9	**	1.6	***
Race
Hispanic	11.90	*	4.24	*	3.31	**
Black	11.48	*	4.47	*	1.26	**
Asian	8.39		2.70		1.88
Native American	8.78		6.74		6.50
Other	18.27		7.14		3.17
White	9.17		3.10		4.25
Poverty
Yes	12.4	*	5.0		2.3
No	9.3		3.2		3.8
Housing stability
Yes	9.1	*	3.6		3.9
No	11.0		3.4		3.3
Violence exposure
Yes	17.5	***	5.3	**	3.1
No	8.5		3.2		3.7
Safe neighborhood
Yes	9.32	***	2.95	***	3.46
No	14.83		8.65		4.95
General health
Healthy	9.1	***	3.3	***	3.6
Fair or poor	20.3	***	6.6	***	4.3
Access to physician
Yes	9.5		3.4		3.4
No	10.8		3.8		4.1

p < .05. **p < .01. ***p < .001.

We turn now to the relationships between social categories and physical symptomology, named fear, and anxiety disorder diagnosis (see Table 2), which address Research Questions 2 and 3. We find that female respondents are consistently more likely to report physical symptomology, named fear, and to have an anxiety disorder diagnosis than their male counterparts. There is no observable, descriptive symptom–to–diagnosis gap by gender.

Table 2 reveals significant differences in physical symptomology, named fear, and anxiety diagnosis across racial groups. Respondents identifying as Hispanic, black, and “other races” have the highest rates of physical symptomology; Native Americans and “other races” have the highest rates of named fear; but Native American and whites respondents have the highest rates of diagnosis. High rates of anxiety diagnosis among Native Americans is expected given the very high rates of physical symptomology; likewise, low rates of anxiety diagnosis among Asians is expected given the low levels of physical symptoms and named fear reported. For all other racial categories, however, we see a symptoms–to–diagnosis gap by race. For instance, black respondents report higher rates of physical symptomology (compared to other race groups), yet this group has the lowest rate of diagnosis. Conversely, white respondents report relatively low rates of symptoms of anxiety (compared to other race groups), yet they are diagnosed with anxiety disorders at a high rate. And, respondents in the other race category have the highest rates of both physical symptoms and named fear, and yet this group does not show high rates of diagnosis.

Youth with violence exposure, reporting not feeling safe in their neighborhood, and poor/fair general health are significantly more likely to have both physical symptomology and named fear but are not more likely to have an anxiety disorder diagnosis compared to those with no violence exposure or higher levels of social support and better health. Here we see the opposite pattern to race; at–risk respondents report higher levels of physical symptomology and named fear, and yet they are not reporting higher rates of anxiety disorder diagnosis. Similarly, youth living in poverty and those with unstable housing have significantly more physical symptomology (but not named fear) but are less likely to receive an anxiety disorder diagnosis compared to those youth not living in poverty and with stable housing.

Table 3 shows the results of logistic regression models predicting an anxiety disorder diagnosis, addressing Research Questions 3 and 4. Model 4 shows the final models predicting diagnosis. Model 4 shows the importance of gender (female), named fear, physical symptomology, race, living in a safe neighborhood, as well as social support. Female respondents are at higher risk for a diagnosis even after controlling for symptomology and all other variables. Native American respondents are at higher risk for a diagnosis even after controlling for symptomology and all other variables (odds ratio [OR] = 6.36), as are white respondents (OR = 3.91) and Hispanic respondents (OR = 2.63) compared to black respondents.

Table 3.

Weighted Logistic Regression Results (Odds Ratios) Predicting Anxiety Disorder Diagnosis on Anxiety Symptoms, Feeling Fearful, and Other Social Factors.

	Model 1		Model 2		Model 3		Model 4
Variables	Odds Ratio	95% Confidence Intervals	Odds Ratio	95% Confidence Intervals	Odds Ratio	95% Confidence Intervals	Odds Ratio	95% Confidence Intervals
Female^a	3.69***	2.6–5.2	3.44***	2.4–4.9	3.41***	2.4–4.9	3.35***	2.4–5.1
Physical symptomology			2.80***	1.9–4.1	2.62***	1.9–4.1	2.46***	1.7–3.7
Named fear					1.89*	1.0–3.5	1.69	.8–3.5
Race^b
Hispanic							2.63*	1.1–6.0
Asian							1.76	.4–8.7
Native American							6.36***	2.5–16.4
Other							1.01	.2-6.3
White							3.91***	1.9–7.9
Poverty							.65	.3–1.2
Stable housing							1.20	.9–1.7
Violence exposure							.81	.5–1.3
Safe neighborhood							.56*	.4–.9
Social support: Teacher							.83*	.7–1.0
Social support: Parent							.81	.6–1.1
Social support: Peer							1.05	.8–1.3
Fair/poor health							.881	.5–1.7
Age: 11–13^c							1.35	.9–1.9
Access to physician							.88	.6–1.2
N	9,421		9,416		9,409		8,197
Pseudo R²	.04		.06		.06		.09

Reference = male.

Reference = black.

Reference = ages 14 and up.

p < .05. ***p < .001.

Discussion

This study investigated the relationship between anxiety symptoms and anxiety disorder diagnosis across social groups with a particular focus on race and gender. Specifically, we assessed whether some social groups experience higher rates of anxiety symptoms (physical symptomology) and feeling fear (named fear) but a lower likelihood of receiving an anxiety disorder diagnosis. We also studied whether some social groups experience the opposite: lower rates of anxiety symptoms (physical symptomology) and feeling fear (named fear) but a higher likelihood to receive an anxiety disorder diagnosis. Lastly, after adjusting for known predictors of anxiety disorder, we conducted an analysis to identify significant factors for receiving an anxiety disorder diagnosis. The results suggest that anxiety symptoms are important predictors for an anxiety disorder diagnosis. This is the ideal situation: Symptoms strongly predict the related diagnosis. This, however, was not the entire story.

We found that factors beyond anxiety symptoms were also important predictors of anxiety disorder diagnosis and that there was not a tidy relationship between symptoms and diagnosis across social groups. First, our study offered a critical examination of the symptom–to–diagnosis gap for anxiety disorders within various social groups. The findings suggest that while several disadvantaged groups are more likely to experience symptoms of anxiety, they are not more likely to receive a diagnosis. Specifically, people who have experienced violence, live in neighborhoods that they do not feel are safe, live in poverty, and have unstable housing are more likely to experience symptoms of anxiety but not more likely to receive an anxiety disorder diagnosis. These initial finding suggest that, as we should expect, various stressors and life circumstances are associated with anxiety, but somewhere along the way—be it variations in medical care seeking, through access barriers, or with physician or diagnostic bias—these disadvantaged groups are not receiving a diagnosis at the same rate as other social groups. This symptom–to–diagnosis gap is critical because diagnosis represents the gateway to both social acknowledgement and medical care. Importantly, untreated anxiety has been shown to be detrimental to social, emotional, and physical functioning and is particularly problematic for socially disadvantaged groups (Fifer et al. 1994; van Beljouw et al. 2010), which underscores the importance of equal access to appropriate diagnosis.

Second, our research also showed that gender and race, even after taking into account anxiety symptoms, are among the strongest predictors of an anxiety disorder diagnosis. That is, above and beyond symptoms, a person’s race and gender are important factors in whether one receives an anxiety disorder diagnosis. Focusing specifically on race, we found that whether a respondent expressed anxiety symptomology or described feeling fearful, Native American, white, and Hispanic respondents were significantly more likely to receive an anxiety disorder diagnosis compared to black respondents. This suggests potential underdiagnosis among black respondents and/or overdiagnosis among Native American, white, and Hispanic respondents. Either way, these data suggest that race is an important factor in whether one receives an anxiety disorder diagnosis. These findings on both race and gender in diagnosis reflect literature in gender studies that suggests the medicalization of women while also reflecting literature on racial bias in medicine (Everett et al. 2016; Hines-Martin et al. 2003; McLean et al. 2011).

The multivariate findings broadly suggest inequity in the process of diagnosing an anxiety disorder. We were able to include a variable in the final model to help account for access to medical care, which allows us to make increasingly confident claims about bias in anxiety disorder diagnosis. Specifically, we can better refute the claim that access to care is the key explanation for the symptom to diagnosis gap. Adjusting for access allows for more targeted claims about bias in anxiety disorder diagnosis, which our research suggests is situated, at least in part, in diagnostic and physician bias.

A final contribution of this study is that it bridges stress process theory and the sociology of diagnosis to answer pertinent research questions on how diagnoses are made. While stress process theory helps us understand why certain groups are more likely to experience the symptoms of anxiety, the sociology of diagnosis helps us understand the uneven distribution of anxiety disorder diagnosis that is not well matched to symptomology. In a parallel fashion, stress process theory emphasizes the injustice of the unequal distribution of stress and disorder, while the sociology of diagnosis suggests the injustice of unequal access to diagnosis, medical acknowledgement, and thus, treatment.

Limitations

There are key measurement and design limitations that are important to acknowledge. Given that this study is based on secondary data, the variables are not designed specifically for our research purposes. As such, several variables are not ideally constructed. Nonetheless, they still provide valuable information and serve as strong indicators and proxies. For instance, while the measure of anxiety symptomology functioned as we anticipated, it is not based on DSM criteria. Moreover, the measure does not capture symptoms of any specific anxiety disorder but rather a broad set of symptoms relevant to anxiety. Additionally, the named fear variable measures the frequency of experiencing fear but would be stronger if it also measured intensity of fear. In addition, the diagnosis status was assessed retrospectively in Wave 4, and it is quite possible that individuals are inaccurately remembering the age of diagnosis. While consistent with recommendations from the Carolina Population Center, the race variable is limited in several ways worth noting. While in the survey the respondents were asked two questions on race and ethnicity and were able to select all that applied, the analytical variable only considered one endorsement (based on an a priori ordering) and did not reflect any endorsements of multiracial identities. For example, if a responded marked Hispanic and black, his or her response is recoded simply as Hispanic. This is problematic in that it reduces the complexity of racial identity and imposes a cardinal racial identity that may not be the one experienced by the respondent. Future research on medical diagnosis would benefit from more identity–based race constructs since a respondent’s experienced racial identity may have important implications for medical bias. Lastly, although the survey is longitudinal and allows us to take into account time order, we are still unable to assess causality.

Future Research

This research suggests several avenues for future inquiry. First, there is still much work to be done to understand the causes of the symptom–to–diagnosis gap. Our research empirically demonstrates this gap, shows for which social groups this gap is most pronounced, and adjusts for access to care to begin honing in on causal mechanisms. Future research could qualitatively and experimentally explore and assess, respectively, the causal mechanisms of this symptom–to–diagnosis gap. Second, this research established that black respondents who experience anxiety symptoms often do not receive diagnoses. The next obvious empirical question to ask is: What is happening to black respondents who do not receive these diagnoses? There are a multitude of possibilities. For instance, they may receive a different mental health or medical diagnosis, experience a loss of social functioning due to untreated symptoms, or (successfully or unsuccessfully) self–treat. We also anticipate that social science research on race and mental health in the criminal justice system and education will be particularly informative for this line of inquiry (e.g., Cauffman 2004; Garland et al. 2005; Rawal et al. 2004; Vincent et al. 2008). Lastly, and relatedly, researchers should explore the social and health trajectory for white female respondents who are diagnosed with anxiety disorders. This current study clearly shows that white female respondents are diagnosed at a significantly higher rate than other groups—regardless of symptoms, named fear, and access—but it does not tell us if this is appropriate, functional, beneficial, or otherwise. Future research could explore the potential medicalization of white girls and women through anxiety disorder diagnosis.

Footnotes

Acknowledgements

We would like to thank Michelle Stransky, Jared Del Rosso, and faculty in the Sociology and Criminology Department at the University of North Carolina-Wilmington for their thoughtful feedback on this project. The data set used for this project was purchased with funds from the Charles L. Cahill fund, an internal grant at the University of North Carolina-Wilmington.This research uses data from Add Health, a program project designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris, and funded by a grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 17 other agencies. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design. Persons interested in obtaining Data Files from Add Health should contact Add Health, The University of North Carolina at Chapel Hill, Carolina Population Center, Carolina Square, Suite 210, 123 W. Franklin Street, Chapel Hill, NC 27516 (addhealth_contracts@unc.edu). No direct support was received from grant P01-HD31921 for this analysis.

Authors’ Note

The authors each contributed equally to this manuscript.

The author order was determined by a coin toss.

Notes

References

Abbo

Catherine

Kinyanda

Eugene

Kizza

Ruth B.

Levin

Jonathan

Ndyanabangi

Sheilla

Stein

Dan J.

2013. “Prevalence, Comorbidity and Predictors of Anxiety Disorders in Children and Adolescents in Rural North-eastern Uganda.” Child and Adolescent Psychiatry and Mental Health 7:21.

Ansseau

Marc

Fischler

Benjamin

Dierick

Michel

Albert

Adelin

Leyman

Sophie

Mignon

Annick

. 2008. “Socioeconomic Correlates of Generalized Anxiety Disorder and Major Depression in Primary Care: The GADIS II Study (Generalized Anxiety and Depression Impact Survey II).” Depression and Anxiety 25(6):506–13.

Asch

Steven M.

Kilbourne

Amy M.

Gifford

Allen L.

Burnam

M. Audrey

Turner

Barbara

Shapiro

Martin F.

Bozzette

Samuel A.

, for the HCSUS Consortium. 2003. “Underdiagnosis of Depression in HIV.” Journal of General Internal Medicine 18(6):450–60.

Bauer

Mark S.

Altshuler

Lori

Evans

Denise R.

Beresford

Thomas

Williford

William O.

Hauger

Richard

for the VA Cooperative Study #430 Team. 2005. “Prevalence and Distinct Correlates of Anxiety, Substance, and Combined Comorbidity in a Multi-site Public Sector Sample with Bipolar Disorder.” Journal of Affective Disorders 85(3):301–15.

Blaxter

Mildred

. 1978. “Diagnosis as Category and Process: The Case of Alcoholism.” Social Science & Medicine. Part A: Medical Psychology & Medical Sociology 12:9–17.

Brown

Phil

. 1995. “Naming and Framing: The Social Construction of Diagnosis and Illness.” Journal of Health and Social Behavior Spec No 34–52.

Brown

Phil

. 1990. “The Name Game: Toward a Sociology of Diagnosis.” The Journal of Mind and Behavior 11:385–406.

Cauffman

Elizabeth

. 2004. “A Statewide Screening of Mental Health Symptoms among Juvenile Offenders in Detention.” Journal of the American Academy of Child & Adolescent Psychiatry 43(4):430–39.

Caughy

Margaret O’Brien

O’Campo

Patricia J.

Muntaner

Carles

. 2003. “When Being Alone Might Be Better: Neighborhood Poverty, Social Capital, and Child Mental Health.” Social Science & Medicine 57(2):227–37.

10.

Chapman

Kenneth R.

Tashkin

Donald P.

Pye

David J.

2001. “Gender Bias in the Diagnosis of COPD.” Chest 119(6):1691–95.

11.

Chiong

Winston

. 2001. “Diagnosing and Defining Disease.” JAMA 285(1):89–90.

12.

Chow

Julian Chun-Chung

Jaffee

Kim

Snowden

Lonnie

. 2003. “Racial/Ethnic Disparities in the Use of Mental Health Services in Poverty Areas.” American Journal of Public Health 93(5):792–97.

13.

Christopher

Lane

. 2007. Shyness: How Normal Behavior Became a Sickness. New Haven, CT: Yale University Press.

14.

Connolly

Amanda

Gaehl

Ella

Martin

Helen

Morris

Julie

Purandare

Nitin

. 2011. “Underdiagnosis of Dementia in Primary Care: Variations in the Observed Prevalence and Comparisons to the Expected Prevalence.” Aging & Mental Health 15(8):978–84.

15.

Conrad

Peter

. 1992. “Medicalization and Social Control.” Annual Review of Sociology 18:209–32.

16.

Conrad

Peter

. 1975. “The Discovery of Hyperkinesis: Notes on the Medicalization of Deviant Behavior.” Social Problems 23(1):12–21.

17.

Conrad

Peter

Schneider

Joseph W.

2010. Deviance and Medicalization: From Badness to Sickness. Philadelphia: Temple University Press.

18.

Croskerry

Pat

. 2003. “The Importance of Cognitive Errors in Diagnosis and Strategies to Minimize Them.” Retrieved July 21, 2017 (http://journals.lww.com/academicmedicine/Fulltext/2003/08000/The_Importance_of_Cognitive_Errors_in_Diagnosis.3.aspx).

19.

Davidoff

Julia

Christensen

Scott

Khalili

David N.

Nguyen

Jaidyn

IsHak

Waguih William

. 2012. “Quality of Life in Panic Disorder: Looking beyond Symptom Remission.” Quality of Life Research 21(6):945–59.

20.

Deacon

Brett

Lickel

James

Abramowitz

Jonathan S.

2008. “Medical Utilization across the Anxiety Disorders.” Journal of Anxiety Disorders 22(2):344–50.

21.

Everett

Bethany G.

Onge

Jarron Saint

Mollborn

Stefanie

. 2016. “Effects of Minority Status and Perceived Discrimination on Mental Health.” Population Research and Policy Review 35(4):445–69.

22.

Fernandes

Vanessa

Osório

Flávia L.

2015. “Are There Associations between Early Emotional Trauma and Anxiety Disorders? Evidence from a Systematic Literature Review and Meta-analysis.” European Psychiatry 30(6):756–64.

23.

Fifer

Sheila K.

Mathias

Susan D.

Patrick

Donald L.

Mazonson

Peter D.

Lubeck

Deborah P.

Buesching

Don P.

1994. “Untreated Anxiety among Adult Primary Care Patients in a Health Maintenance Organization.” Archives of General Psychiatry 51(9):740–50.

24.

Garland

Ann F.

Lau

Anna S.

Yeh

May

McCabe

Kristen M.

Hough

Richard L.

Landsverk

John A.

2005. “Racial and Ethnic Differences in Utilization of Mental Health Services among High-risk Youths.” American Journal of Psychiatry 162(7):1336–43.

25.

Graber

Mark L.

Franklin

Nancy

Gordon

Ruthanna

. 2005. “Diagnostic Error in Internal Medicine.” Archives of Internal Medicine 165(13):1493–99.

26.

Green

Alexander R.

Carney

Dana R.

Pallin

Daniel J.

Ngo

Long H.

Raymond

Kristal L.

Iezzoni

Lisa I.

Banaji

Mahzarin R.

2007. “Implicit Bias among Physicians and Its Prediction of Thrombolysis Decisions for Black and White Patients.” Journal of General Internal Medicine 22(9):1231–38.

27.

Haley

William E.

LaMonde

Laurie A.

Han

Beth

Burton

Allison M.

Schonwetter

Ronald

. 2003. “Predictors of Depression and Life Satisfaction among Spousal Caregivers in Hospice: Application of a Stress Process Model.” Journal of Palliative Medicine 6(2):215–24.

28.

Hines-Martin

Vicki

Malone

Mary

Kim

Sanggil

Brown-Piper

Ada

. 2003. “Barriers to Mental Health Care Access in an African American Population.” Issues in Mental Health Nursing 24(3):237–56.

29.

Hoffman

Deborah L.

Dukes

Ellen M.

Wittchen

Hans-Ulrich

. 2008. “Human and Economic Burden of Generalized Anxiety Disorder.” Depression and Anxiety 25(1):72–90.

30.

Hovens

J. G. F. M.

Wiersma

J. E.

Giltay

E. J.

van Oppen

Spinhoven

Penninx

B. W.

Zitman

F. G.

2010. “Childhood Life Events and Childhood Trauma in Adult Patients with Depressive, Anxiety and Comorbid Disorders vs. Controls.” Acta Psychiatrica Scandinavica 122(1):66–74.

31.

Jutel

Annemarie

. 2009a. “Doctor’s Orders: Diagnosis, Medical Authority and the Exploitation of the Fat Body.” Pp. 60–77 in Biopolitics and the “Obesity Epidemic”: Governing Bodies, edited by Wright

Harwood

New York: Taylor & Francis.

32.

Jutel

Annemarie

. 2009b. “Sociology of Diagnosis: A Preliminary Review.” Sociology of Health & Illness 31(2):278–99.

33.

Jutel

Annemarie Goldstein

Conrad

Peter

. 2014. Putting a Name to It: Diagnosis in Contemporary Society. Baltimore: JHU Press.

34.

Jutel

Annemarie

Nettleton

Sarah

. 2011. “Towards a Sociology of Diagnosis: Reflections and Opportunities.” Social Science & Medicine 73(6):793–800.

35.

Kessler

Ronald C.

Berglund

Patricia

Demler

Olga

Jin

Robert

Merikangas

Kathleen R.

Walters

Ellen E.

2005. “Lifetime Prevalence and Age-of-onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication.” Archives of General Psychiatry 62(6):593–602.

36.

Kessler

Ronald C.

Chiu

Wai Tat

Demler

Olga

Walters

Ellen E.

2005. “Prevalence, Severity, and Comorbidity of 12-Month DSM-IV Disorders in the National Comorbidity Survey Replication.” Archives of General Psychiatry 62(6):617–27.

37.

Kessler

Ronald C.

Petukhova

Maria

Sampson

Nancy A.

Zaslavsky

Alan M.

Wittchen

Hans-Ullrich

. 2012. “Twelve-month and Lifetime Prevalence and Lifetime Morbid Risk of Anxiety and Mood Disorders in the United States.” International Journal of Methods in Psychiatric Research 21(3):169–84.

38.

Klerman

Gerald L.

1989. “Psychiatric Diagnostic Categories: Issues of Validity and Measurement: An Invited Comment on Mirowsky and Ross (1989).” Journal of Health and Social Behavior 30(1):26–32.

39.

Latkin

Carl A.

Curry

Aaron D.

2003. “Stressful Neighborhoods and Depression: A Prospective Study of the Impact of Neighborhood Disorder.” Journal of Health and Social Behavior 44:34–44.

40.

Leddy

Meaghan

Haaga

David

Gray

James

Schulkin

Jay

. 2011. “Postpartum Mental Health Screening and Diagnosis by Obstetrician-gynecologists.” Journal of Psychosomatic Obstetrics & Gynecology 32(1):27–34.

41.

Lewinsohn

Peer M.

Gotlib

Ian H.

Lewinsohn

Mark

Seeley

John R.

Allen

Nicholas B.

1998. “Gender Differences in Anxiety Disorders and Anxiety Symptoms in Adolescents.” Journal of Abnormal Psychology 107(1):109–17.

42.

Link

Bruce G.

Phelan

. 1995. “Social Conditions as Fundamental Causes of Disease.” Journal of Health and Social Behavior 36:80–94.

43.

McKinlay

John B.

1996. “Some Contributions from the Social System to Gender Inequalities in Heart Disease.” Journal of Health and Social Behavior 37:1–26.

44.

McLean

Carmen P.

Asnaani

Anu

Litz

Brett T.

Hofmann

Stefan G.

2011. “Gender Differences in Anxiety Disorders: Prevalence, Course of Illness, Comorbidity and Burden of Illness.” Journal of Psychiatric Research 45(8):1027–35.

45.

Merikangas

Kathleen Ries

Jian-ping

Burstein

Marcy

Swanson

Sonja A.

Avenevoli

Shelli

Cui

Lihong

Benjet

Corina

Georgiades

Katholiki

Swendsen

Joel

. 2010. “Lifetime Prevalence of Mental Disorders in U.S. Adolescents: Results from the National Comorbidity Survey Replication–Adolescent Supplement (NCS-A).” Journal of the American Academy of Child & Adolescent Psychiatry 49(10):980–89.

46.

Mirowsky

John

Ross

Catherine E.

1989. “Psychiatric Diagnosis as Reified Measurement.” Journal of Health and Social Behavior 30(1):11–25.

47.

Nettleton

Sarah

. 2006. “‘I Just Want Permission to Be Ill’: Towards a Sociology of Medically Unexplained Symptoms.” Social Science & Medicine 62(5):1167–78.

48.

Nutt

David J.

Ballenger

James C.

Sheehan

David

Wittchen

Hans-Ulrich

. 2002. “Generalized Anxiety Disorder: Comorbidity, Comparative Biology and Treatment.” International Journal of Neuropsychopharmacology 5(4):315–25.

49.

Olatunji

Bunmi O.

Cisler

Josh M.

Tolin

David F.

2007. “Quality of Life in the Anxiety Disorders: A Meta-analytic Review.” Clinical Psychology Review 27(5):572–81.

50.

Ozer

Emily J.

Lavi

Iris

Douglas

Laura

Wolf

Jennifer Price

. 2017. “Protective Factors for Youth Exposed to Violence in Their Communities: A Review of Family, School, and Community Moderators.” Journal of Clinical Child & Adolescent Psychology 46(3):353–78.

51.

Pappas

Gregory

Queen

Susan

Hadden

Wilbur

Fisher

Gail

. 1993. “The Increasing Disparity in Mortality between Socioeconomic Groups in the United States, 1960 and 1986.” New England Journal of Medicine 329(2):103–09.

52.

Pascoe

Elizabeth A.

Richman

Laura Smart

. 2009. “Perceived Discrimination and Health: A Meta-analytic Review.” Psychological Bulletin 135(4):531–54.

53.

Pearlin

Leonard I.

Menaghan

Elizabeth G.

Lieberman

Morton A.

Mullan

Joseph T.

1981. “The Stress Process.” Journal of Health and Social Behavior 22:337–56.

54.

Perry

Brea L.

Harp

Kathi L. H.

Oser

Carrie B.

2013. “Racial and Gender Discrimination in the Stress Process: Implications for African American Women’s Health and Well-Being.” Sociological Perspectives 56(1):25–48.

55.

Rawal

Purva

Romansky

Jill

Jenuwine

Michael

Lyons

John S.

2004. “Racial Differences in the Mental Health Needs and Service Utilization of Youth in the Juvenile Justice System.” The Journal of Behavioral Health Services & Research 31(3):242–54.

56.

Rosenberg

Charles E.

2006. “Contested Boundaries: Psychiatry, Disease, and Diagnosis.” Perspectives in Biology and Medicine 49(3):407–24.

57.

Ross

Catherine E.

2000. “Neighborhood Disadvantage and Adult Depression.” Journal of Health and Social Behavior 41:177–187.

58.

Samaan

Rodney A.

2010. “The Influences of Race, Ethnicity, and Poverty on the Mental Health of Children.” Journal of Health Care for the Poor and Underserved 11(1):100–10.

59.

Sareen

Jitender

Cox

Brian J.

Afifi

Tracie O.

de Graaf

Ron

Asmundson

Gordon J. G.

Ten Have

Margreet

Stein

Murray B.

2005. “Anxiety Disorders and Risk for Suicidal Ideation and Suicide Attempts: A Population-based Longitudinal Study of Adults.” Archives of General Psychiatry 62(11):1249–57.

60.

Scheper-Hughes

Nancy

. 1993. Death without Weeping: The Violence of Everyday Life in Brazil. Berkeley: University of California Press.

61.

Schieman

Scott

Whitestone

Yuko Kurashina

Van Gundy

Karen

. 2006. “The Nature of Work and the Stress of Higher Status.” Journal of Health and Social Behavior 47(3):242–57.

62.

Shearer

Steven L.

2007. “Recent Advances in the Understanding and Treatment of Anxiety Disorders.” Primary Care: Clinics in Office Practice 34(3):475–504.

63.

Simon

Naomi M.

Otto

Michael W.

Wisniewski

Stephen R.

Fossey

Mark

Sagduyu

Kemal

Frank

Ellen

Sachs

Gary S.

Nierenberg

Andrew A.

Thase

Michael E.

Pollack

Mark H.

, for the STEP-BD Investigators. 2004. “Anxiety Disorder Comorbidity in Bipolar Disorder Patients: Data from the First 500 Participants in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD).” American Journal of Psychiatry 161(12):2222–29.

64.

Sorkin

Dara H.

Ngo-Metzger

Quyen

Billimek

John

August

Kristin J.

Greenfield

Sheldon

Kaplan

Sherrie H.

2011. “Underdiagnosed and Undertreated Depression among Racially/Ethnically Diverse Patients with Type 2 Diabetes.” Diabetes Care 34(3):598–600.

65.

Szasz

Thomas

. 2007. The Medicalization of Everyday Life: Selected Essays. Syracuse, NY: Syracuse University Press.

66.

Taylor

John

Turner

R. Jay

. 2002. “Perceived Discrimination, Social Stress, and Depression in the Transition to Adulthood: Racial Contrasts.” Social Psychology Quarterly 65(3):213–25.

67.

Thoits

Peggy A.

2010. “Stress and Health: Major Findings and Policy Implications.” Journal of Health and Social Behavior 51(1 suppl):S41–53.

68.

Turner

R. Jay

Lloyd

Donald A.

1999. “The Stress Process and the Social Distribution of Depression.” Journal of Health and Social Behavior 40(4):374–404.

69.

Turner

R. Jay

Wheaton

Blair

Lloyd

Donald A.

1995. “The Epidemiology of Social Stress.” American Sociological Review 60:104–25.

70.

Turney

Kristin

Wildeman

Christopher

Schnittker

Jason

. 2012. “As Fathers and Felons: Explaining the Effects of Current and Recent Incarceration on Major Depression.” Journal of Health and Social Behavior 53(4):465–81.

71.

van Beljouw

Ilse M. J.

Verhaak

Peter F. M.

Cuijpers

Pim

van Marwijk

Harm W. J.

Penninx

Brenda W. J. H.

2010. “The Course of Untreated Anxiety and Depression, and Determinants of Poor One-year Outcome: A One-year Cohort Study.” BMC Psychiatry 10:86.

72.

Vincent

Gina M.

Grisso

Thomas

Terry

Anna

Banks

Steven

. 2008. “Sex and Race Differences in Mental Health Symptoms in Juvenile Justice: The MAYSI-2 National Meta-analysis.” Journal of the American Academy of Child & Adolescent Psychiatry 47(3):282–90.

73.

Whiting

Penny

Rutjes

Anne W. S.

Reitsma

Johannes B.

Glas

Afina S.

Bossuyt

Patrick M. M.

Kleijnen

Jos

. 2004. “Sources of Variation and Bias in Studies of Diagnostic Accuracy: A Systematic Review.” Annals of Internal Medicine 140(3):189–202.

74.

Williams

David R.

Yan

Jackson

James S.

Anderson

Norman B.

1997. “Racial Differences in Physical and Mental Health: Socio-Economic Status, Stress and Discrimination.” Journal of Health Psychology 2(3):335–51.

75.

Wittchen

Hans-Ulrich

. 2002. “Generalized Anxiety Disorder: Prevalence, Burden, and Cost to Society.” Depression and Anxiety 16(4):162–71.