3.2.1. Elimination of the Multi-Axial System

One of the key changes from DSM-IV to DSM-5 is the elimination of the multi-axial system. DSM-IV approached psychiatric assessment and organization of biopsychosocial information using a “multi-axial” formulation [2]. There were five different axes. Axis I consisted of mental health and SUDs; Axis II was reserved for coding personality disorders and mental retardation; Axis III was used for coding medical conditions; Axis IV was to note psychosocial and environmental problems (e.g., housing, employment); and Axis V was a numerical assessment of overall functioning known as the GAF. Although the impact of removing the overall multi-axial structure in DSM-5 is unknown, there is concern among clinicians that eliminating the disciplined approach for gathering and organizing clinical assessment data that was required under the multi-axial system will hinder clinical practice [57]. However, the direct impact of eliminating the multi-axial structure on NSDUH/MHSS is likely to be negligible since it will not affect the characteristics of diagnoses. Of note, Axis II developmental disorders are specifically excluded from the definition of SMI. The removal of the axis system, in addition to disorder re-categorization, which is discussed in more detail later, may necessitate a more specific delineation of what is defined as a “developmental disorder” for exclusion.

3.2.2. Removal of the GAF score

The removal of the GAF score, an assessment of overall functioning, introduces a level of uncertainty in the measurement of SMI in MHSS and NSDUH. The federal government defines adults with SMI as people aged 18 or older who currently have or at any time during the past year had a diagnosable mental behavioral or emotional disorder of sufficient duration to meet diagnostic criteria (excluding SUDs, dementia, and developmental disorders) that has resulted in functional impairment, which substantially interferes with or limits one or more major life activities [58]. In addition, those who had a psychiatric disorder in the past that resulted in serious impairment, but are now improved due to treatment are considered to have an SMI. MHSS/NSDUH follows this definition, using the GAF score to indicate impairment, with the exception that NSDUH and MHSS only assess past year disorders. At the end of the MHSS interview, the clinical interviewer assigned a GAF score to the respondent. Representing the DSM-IV Axis V assessment, the GAF score ranges from 1 to 100, based on a two-dimensional continuum of (1) psychiatric symptoms—from severe mental illness to superior mental health, and (2) psychosocial functioning—from severe impairment to superior functioning. For example, a GAF score of 5 would indicate very serious psychiatric symptoms and/or grossly impaired functioning; a GAF score of 95 would indicate no symptoms and superior functioning across a wide range of psychosocial and occupational activities; and a GAF score of 50 would indicate serious symptoms or serious impairment in social, occupational, or school functioning. Impairment in functioning caused by physical or environmental limitations is not included when assigning a GAF score.

In the MHSS SCID, the clinical interviewer was directed to record the GAF score that represented the respondent’s worst symptoms or level of functioning in the past year, reflecting whichever dimension was lower. As mentioned previously, a respondent would be classified as having SMI if (1) the GAF score was 50 or lower and (2) one or more nonsubstance use disorders were present in the past year.

The GAF scale was dropped from the DSM-5 because of its conceptual lack of clarity (i.e., including symptoms, suicide risk, and disabilities in the descriptors) and questionable psychometric properties [2]. The American Psychological Association’s (APA’s) observation of ongoing problems with the GAF is supported by the MHSS data, which found high interrater reliability between the clinical interviewers’ and clinical supervisors’ consensus ratings for symptoms, but not for the GAF scale (Table 3.1). Symptom-level agreement was calculated using the total number of variables in the clinical interview as the denominator, and without penalty for omissions and commissions from earlier errors within the same disorder. As shown in Table 3.1, symptom-level agreement improved and was consistently high after the first year, averaging 96 percent. There was also consistently high agreement for having one or more mental disorders, with an average of 99 percent. Clinical interviewers’ kappa scores steadily improved each year, with 100 percent of clinical interviewers having high kappa scores for the last seven quarters of the study. Less consistent agreement was found for GAF (0 to 95 percent) and SMI (0 to 100 percent). The majority of error for GAF and SMI ratings was caused by clinical interviewers’ GAF ratings that were lower than the clinical supervisors’ consensus GAF ratings. Discussions during the interrater reliability prediction conference calls indicated that clinical interviewers’ biases toward lower GAF scores were because of overestimating the severity of the respondent’s symptoms (e.g., the seriousness of suicidal thoughts, the danger of driving while impaired by alcohol).

Table 3.1

Interrater Reliability Exercise Performance: 2008–2012 MHSS.

Several steps were taken to improve consistency within the MHSS team of clinical interviewers and clinical supervisors. First, quarterly interrater reliability data collection was conducted with group conference calls to discuss and “calibrate” the team’s interrater reliability ratings. Second, all clinical interviewer and clinical supervisor training sessions incorporated GAF lectures and exercises. Third, the clinical interviewers and clinical supervisors were provided with training and/or supervision by Dr. Michael First (the SCID developer). Despite these efforts, the interrater reliability remained fair to poor.

To address this limitation and still provide a global measure of disability, the World Health Organization Disability Assessment Schedule version 2.0 (WHODAS 2.0) is included, for further study, in DSM-5 and was included in the DSM-5 field trial. The WHODAS 2.0 assesses disability among adults aged 18 or older across six domains: cognition, mobility, self-care, getting along with others, life activities (e.g., household, work and/or school), and participation in society. In each item the individual is asked to rate how much difficulty he or she has had in specific areas of functioning in the past 30 days: “none” (1), “mild” (2), “moderate” (3), “severe” (4), and “extreme” (5). The WHODAS 2.0 comes in two lengths. The 36-item WHODAS 2.0 yields subscale scores for each of the six domains of functioning along with the overall functioning score and requires approximately 20 minutes to administer. The 12-item WHODAS 2.0 yields an overall functioning score that explains 81 percent of the variance of the 36-item version and requires approximately 5 minutes to administer. The WHODAS 2.0 is available in interviewer-, self-, and proxy-administered forms. Domain scores and summary scores are calculated by adding the ratings using the simple or complex rating system. Using the simple rating system, scores are summed according to the assigned scores for each item, from “none” (1) to “extreme” (5), while the complex scoring system differentially weights the items and levels of severity using a computer program (available from the WHO website). The scoring is done by summing the recorded item scores within each domain, adding the six domain scores, and converting the summary score into a metric ranging from 0 to 100 (where 0 = no disability and 100 = full disability). Population norms for item response theory (IRT)-based scores of the WHODAS 2.0 are also available at the WHO website.

The WHODAS has several strengths:

• It is easy to administer by either the respondent or interviewer.
• It is applicable in both clinical and general population settings.
• It is applicable across cultures and in all adult populations.
• It produces standardized disability levels and profiles.
• It is a direct conceptual link to the International Classification of Functioning, Disability and Health (ICF).
• It has good psychometric properties (e.g., 0.82 to 0.98 reliability coefficient for the full-length WHODAS: http://www.who.int/classifications/icf/whodasii/en/index2.html).
• The WHODAS 2.0 allows for a clinician to override the respondent’s answers in cases where information gathered in the clinical interview indicates that another answer is more accurate.
• It is the method increasingly used by researchers.

Likewise, there are several limitations to using the WHODAS 2.0 in place of the GAF. One potential limitation is that the WHODAS is used as a predictive variable in the MHSS prediction model. If the WHODAS were to replace the GAF, then the outcome (severity of mental illness as indicated by the WHODAS) would include one of the model predictors (the presence of impairment as indicated by the WHODAS). This may not be problematic because the score from the WHODAS used in the clinical interview may be altered to reflect the clinical judgment of the interviewer rather than the respondent’s perceptions; however, there is likely to be substantial overlap. This overlap may improve AMI and SMI measurement if the WHODAS is a psychometrically strong indicator of impairment (which is suggested by the literature [59,60]), or it may increase imprecision in areas of measurement weakness. Therefore, while replacing the GAF with the WHODAS 2.0 in the MHSS may improve the reliability and ease of classifying the severity of mental illness cases in the MHSS (e.g., AMI, SMI), it may create difficulties in measuring mental illness (i.e., SMI and AMI) in NSDUH. A stronger limitation of replacing the GAF with the WHODAS to be considered is that the two assessments are dissimilar in what they measure. The GAF provides a measure of the respondent’s symptoms and functioning whereas the WHODAS measures only disability. Replacing the GAF with the WHODAS 2.0 would require creating new prediction model for SMI and AMI to reflect the differences between these two assessments. Given the concerns of this replacement, it may be more efficient to retain using the GAF score in MHSS, but problems with the GAF interrater reliability may make replacement preferable. Other assessments of disability might be considered to replace either the WHODAS in NSDUH or the GAF in the MHSS (e.g., the Short Form 36 [SF-36] [61]; the Sheehan Disability Scale [SDS] [62]; the Life Functioning Questionnaire [LFQ] [63]) after considering factors like context effects, timing, and breaks in trends.

3.2.3. Disorder Reclassification

DSM-IV and DSM-5 categorize disorders into “classes” with the intent of grouping similar disorders (particularly those that are suspected to share etiological mechanisms or have similar symptoms) to help clinician and researchers use the manual. From DSM-IV to DSM-5, there has been a reclassification of many disorders that reflects advancements in the field. Table 3.2 lists the disorder classes included in DSM-IV and DSM-5. In DSM-5, six classes were added and four were removed. As a result of these changes in the overall classification system, numerous individual disorders were reclassified. Table 3.3 lists DSM-IV and DSM-5 disorders and their classification scheme.

Table 3.2

Disorder Classes Presented by the DSM-IV and DSM-5, as Ordered in DSM-IV.

Table 3.3

Disorder Classification in the DSM-IV and DSM-5.

The reclassification of disorders does not have a direct impact on estimates produced by MHSS or NSDUH; however, it does warrant consideration when documenting disorders that may have changed classes. For example, an estimate for “any anxiety disorder,” using the DSM-IV classification would include PTSD and OCD, while the DSM-5 classification would not. Therefore, the DSM version and included disorders should be clearly documented in all relevant SAMHSA products that report class-level estimates.

3.3.1. Major Depressive Episode/Disorder (NSDUH and MHSS)

An MDE is characterized by the combination of depressed mood or loss of interest or pleasure (anhedonia) lasting for most of the day, nearly every day, or both for 2 weeks or more [2]. The primary symptom (depressed mood or anhedonia) must be accompanied by four or more additional symptoms (Table 3.4), or 3 or more if both depressed mood and anhedonia are present, and must cause clinically significant distress or impairment. NSDUH and the MHSS directly assess MDE, and the MHSS also assesses MDD. The primary difference between MDE and MDD is that MDD includes all of the criteria for MDE but also includes the exclusionary criteria for bereavement, mania, and hypomania (discussed in Section 3.2.3); that is, an individual with MDD cannot ever have experienced a manic or hypomanic episode.

Table 3.4

DSM-IV to DSM-5 Major Depressive Episode/Disorder Comparison.

Changes in the MDE/MDD criteria from DSM-IV to DSM-5 have been minimal (Table 3.4). There have been some wording changes in the way that “mixed states” are described for diagnostic coding (mixed states now fall under the specifier “with mixed features”). In addition, the examples provided to describe a depressed mood have been expanded in DSM-5 from “e.g. feels sad or empty” [1, p. 327] to “e.g. feels sad, empty, hopeless” [2, p. 160] This change in wording has not received much attention [64]. However, the wording change has the possibility of increasing the prevalence of MDE/MDD if survey respondents and clinicians were not already equating feeling hopeless with feeling sad, empty, or depressed. NSDUH’s assessment of MDE in adults and youth (aged 12 to 17) varies slightly to take into account maturation differences in these groups. Assessment of part of Criterion A1 in adults asks: “Have you ever had a period of time lasting several days or longer when most of the day you were very discouraged about how things were going in your life.” In youth this question is phrased: “Have you ever had a period of time lasting several days or longer when most of the day you felt very discouraged or hopeless about how things were going in your life?” Thus, hopelessness is already being assessed among youth in NSDUH and no changes in the youth MDE module are indicated to maintain alignment with DSM criteria. Among adults, NSDUH could replace the language used in the adult module with that used in the youth module and thereby maintain alignment with DSM. This has the potential to increase the prevalence estimates generated by NSDUH and alter trend estimates. It may also introduce context effects. However, in a review of DSM-5 criteria changes, Uher et al. noted that in a treatment sample, only 8 percent of patients with MDD reported feeling hopelessness but not sadness [64]. Moreover, NSDUH also assesses feelings of emptiness, depression, and discouragement, thereby increasing the likelihood that individuals experiencing feelings of hopelessness will also endorse one of the other symptoms and therefore be included in the diagnostic tally.

The more substantive change is that the formal bereavement exclusion for MDE/MDD in DSM-IV has been removed from DSM-5. The bereavement exclusion criterion has been a longstanding feature of MDE/MDD, designed to allow clinicians to distinguish between normal grieving and a mental illness [65]. It has been replaced with text noting that MDE/MDD should not be confused with normal and appropriate grief but that the presence of bereavement is not prohibitive of an MDE/MDD diagnosis. This change had several rationales. First, there was the need to remove the implication that bereavement typically lasts only 2 months, whereas the duration is more commonly years [66]. Second, bereavement is a psychosocial stressor, similar to other nonexcluded stressors that can precipitate an MDE [67]. Third, when MDD occurs in the context of bereavement, it adds an additional risk for suffering, feelings of worthlessness, suicidal ideation, poorer somatic health, worse interpersonal and work functioning, and an increased risk for persistent complex bereavement disorder, which is now described with explicit criteria in Conditions for Further Study in DSM-5 Section III [68]. Fourth, bereavement-related major depression is associated with similar genetic and biological risk factors, personality characteristics, patterns of comorbidity, and risks of chronicity and/or recurrence as nonbereavement-related MDD [69]. Finally, the depressive symptoms associated with bereavement-related depression respond to the same treatments as nonbereavement-related depression [69]. It should be noted, however, that the removal of this exclusion criterion is one of the more controversial changes made in DSM-5 [70,71].

The changes in exclusion criteria are not anticipated to affect NSDUH estimates. NSDUH already does not apply the bereavement or mixed-episode criteria. Rather, the elimination of these exclusions may lead to NSDUH generating population prevalence estimates that are closer to the population prevalence of DSM-5 MDE. Therefore, changes do not appear to be needed to the NSDUH MDE module at this time. However, it should be noted that DSM-5 divided the DSM-IV mood disorders into two groups, depressive disorders and bipolar and related disorders. This change makes it impossible to classify an individual with MDE in NSDUH into either group because NSDUH does not assess mania/hypomania which enables the determination of which class of disorders MDE falls under (depressive disorders or bipolar and related disorders).

The MHSS assesses MDD, which is impacted by the changes from DSM-IV to DSM-5 because it includes the exclusion criterion for bereavement and mixed episodes. The removal of these exclusion criteria is expected to increase the prevalence of MDE/MDD. Quantitative analysis of the impact of bereavement-related criterion changes on MDE prevalence using nationally representative data from the National Epidemiologic Study of Alcohol and Related Conditions (NESARC) indicated that an estimated 7.7 percent of participants with depressive episodes had bereavement-related episodes [72]. This corresponded to an estimated increase of MDE/MDD prevalence of 1.17 percent (unweighted). In the MHSS, this would lead to an underestimate of MDD prevalence. In the AMI/SMI predictive model, there would be some misclassification of the AMI/SMI outcome of “caseness” (individuals excluded for bereavement would be misclassified as not having MDD). However, the impact resulting from the change in bereavement-related exclusionary criteria on AMI/SMI estimates in NSDUH is likely to be minimal because MDE from NSDUH is part of the predictive algorithm, with an estimated 98.8 percent (weighted) of NSDUH respondents who have MDE reach the cut-point for AMI. Of more concern is the change in mixed episode-related exclusionary criteria. A mixed episode is defined as co-occurring symptoms of mania and depression lasting at least 1 week in DSM-IV (Table 3.9); quantitative estimates of the impact of removing this criterion have not been identified, but a review of the available evidence suggests this change will lead to a slight increase in the prevalence estimates of MDE [73]. A 2011 critical review of the bipolar literature suggested that 40 to 50 percent of people with MDD have met subthreshold features of mania/hypomania in their lifetime, although the use of past year diagnoses in MHSS would reduce the number of cases being captured that had both MDD and a mixed episode [74]. Data from the DSM-5 field trials was mixed, with some sites noting a slight decrease, others noting a slight increase, and one site recording a substantial increase in the point prevalence of DSM-IV and DSM-5 defined MDD (range of DSM-IV MDD: 0.21 to 0.49; range of DSM-5 MDD: 0.19 to 0.37) [75]. The variation in this data and their reliance on patient samples limits their direct applicability to NSDUH/MHSS. Moreover, the test-retest reliability for MDD in the field trials averaged in the questionable range (sites ranged from unacceptable [Intraclass Kappa=0.13] to good [Intraclass Kappa=0.42], no sites had very good reliability for MDD).

Table 3.9

DSM-IV to DSM-5 Mixed Episode Criteria Comparison.

3.3.2. Dysthymic Disorder (MHSS)

Dysthymic disorder is a disorder characterized by a persistently depressed mood that occurs most of the day, for more days than not, for a period of at least 2 years (at least 1 year or more in children and adolescents; Table 3.5) [2]. In the DSM it has been re-named persistent depressive disorder. This name change reflects the consolidation of DSM-IV chronic major depressive disorder and dysthymic disorder that occurred in DSM-5. Previously, in DSM-IV, a diagnosis of dysthymic disorder was contraindicated if the patient met criteria for MDD in the first 2 years after the symptoms arose. In DSM-5 this exclusion has been removed. DSM-5 estimates the 12-month prevalence of persistent depressive disorder at approximately 2 percent, a combination of the estimated prevalence of dysthymia (1.5 percent) and chronic MDD (0.5 percent) [2]. In a nationally representative study of Australian households, the estimated lifetime prevalence of persistent depressive disorder was 4.6 percent [76].

Table 3.5

DSM-IV to DSM-5 Dysthymic Disorder Comparison.

The NSDUH main interview and the MHSS do not assess chronic major depression; therefore, estimates of persistent depressive disorder will be an underestimate of the population prevalence of persistent depressive disorder by approximately 0.5 percent for past year estimates [2]. However, the impact on AMI/SMI estimates would likely be minimal because most if not all individuals with chronic major depression would be classified as having MDE/MDD, and therefore, would be counted as having a mental illness for inclusion in the predictive algorithm. If the DSM-IV to DSM-5 criteria changes for persistent depressive disorder are not included in MHSS, SAMHSA should ensure clarity of language in MHSS documentation to prevent confusion in estimates for only dysthymia rather than the new term, persistent depressive disorder.

3.3.3. Manic Episode and Bipolar I Disorder (MHSS)

Bipolar I disorder, at one time referred to as manic-depressive disorder, is defined by the occurrence of at least one manic episode, which is a period of abnormally and persistently elevated, expansive, or irritable mood that is accompanied by increased energy or activity, which results in clinically significant impairment in functioning or the need for hospitalization [2]. Manic episodes and bipolar I disorder are assessed in MHSS and thereby incorporated into the SMI estimate based on the clinical sample that is used in the development of the prediction model.

The diagnostic criteria for manic episodes have undergone several changes between DSM-IV and DSM-5 (Table 3.6). Criterion A now requires that mood changes are accompanied by abnormally and persistently goal-directed behavior or energy. Second, wording has been added to clarify that (1) symptoms must represent a noticeable change from usual behavior, and (2) these changes have to be present most of the day, nearly every day during the minimum 1 week duration.

Table 3.6

DSM-IV to DSM-5 Manic Episode Criteria Comparison.

Exclusion criteria for manic episodes have also changed. Mania that emerge after antidepressant treatment can be classified as bipolar I disorder diagnosis in the DSM-5, whereas this was a substance-induced manic disorder in DSM-IV. The criteria for bipolar I disorder has also undergone a slight change (Table 3.7). In DSM-IV, six subtypes of bipolar I diagnoses were presented according to the features of the most recent mood episode. However, in DSM-5 these subtypes were converted to specifiers instead (i.e., specify most recent episode type according to its features).

Table 3.7

DSM-IV to DSM-5 Bipolar I Disorder Comparison.

Manic episodes and bipolar I disorder are assessed in the MHSS directly and are included in the operational definition of AMI/SMI used in NSDUH. The literature on how diagnostic changes will impact prevalence estimates is limited. Data from the DSM-5 field trials suggest that the impact will be minimal. Prevalence estimates from the two clinical sites where bipolar I was evaluated both found the same point prevalence estimates of 0.25 to 0.28 percent under both DSM-IV and DSM-5 criteria. However, the DSM-5 field trials used patient populations for estimation, and therefore, these findings may not be generalizable to the general population [75].

In addition to the changes in manic episode criteria, there have been changes to the overall diagnostic criteria for bipolar I disorder. In DSM-IV, bipolar I disorder was diagnosed by “type,” which was characterized by the nature of the most recent episode (bipolar I disorder, single manic episode; bipolar I disorder, most recent episode hypomanic; bipolar I disorder, most recent episode manic; bipolar I disorder, most recent episode mixed; bipolar I disorder, most recent episode depressed; and bipolar I disorder, most recent episode unspecified). Each of these “types” had slightly varying criteria (Table 3.7). In DSM-5 the diagnostic description has been simplified and these “types” have been relegated to the role of specifiers. Diagnostic procedure indicates that clinicians should first provide the bipolar I diagnosis then specify the characteristics of the most recent episode, in addition to several other specifiers. Although important to understanding the general change in diagnostic approach, these changes are geared toward communicating a more streamlined diagnostic description rather than reflecting a change in the diagnostic criteria and will not have an impact on NSDUH or MHSS beyond those changes discussed with regard to manic episode criteria changes.

3.3.4. Panic Disorder and Agoraphobia (MHSS)

Panic disorder is an anxiety disorder characterized by panic attack(s) and the ongoing concern about experiencing additional panic attacks [2] (Table 3.10). A panic attack is an abrupt, but quickly peaking, surge of intense fear or discomfort, accompanied by a series of physical symptoms. Agoraphobia is an anxiety disorder characterized by an intense fear or anxiety triggered by the real or anticipated exposure to a number of situations (i.e., using public transportation, being in open spaces, being in enclosed spaces, standing in line, being in a crowd, or being outside the home), which causes clinically significant distress or impairment (Table 3.10) [2]. In DSM-IV, panic disorder and agoraphobia were conceptually linked. The diagnoses in DSM-IV included panic disorder with agoraphobia, panic disorder without agoraphobia, and agoraphobia without history of panic disorder. In DSM-5, however, panic disorder and agoraphobia have been separated and individuals meeting criteria for both disorders are considered to have comorbid mental illnesses.

Table 3.10

Panic Disorder and Agoraphobia Criteria Changes from DSM-IV to DSM-5.

Panic disorder and agoraphobia are both assessed in the MHSS and criteria changes may affect AMI/SMI estimates. Examining the comparison of panic disorder criteria specifically (Table 3.10), with the exception of the disaggregation of agoraphobia, the criteria are similar between DSM-IV and DSM-5. There are minor wording changes to the description of panic attacks that may have slight implications to the prevalence of panic disorder under DSM-5. Under DSM-IV the specification was made that panic attacks were discrete periods of intense fear or discomfort that peaked within 10 minutes. In DSM-5, panic attacks are described as an abrupt surge of intense fear or intense discomfort that peak within a few minutes. The wording changes reflect two conceptual issues. First, the change in wording from a discrete event to an abrupt surge broadens criteria based on evidence that panic attacks do not necessarily arise “out of the blue,” but can arise during periods of anxiety or other distress and that it is the sudden increase in fear/discomfort that is the hallmark of a panic attack. In addition, they have removed the 10-minute criterion in favor of the less precise but implicitly shorter descriptive of “within minutes” [2 p. 214]. The diagnostic changes in panic disorder will impact NSDUH and the MHSS insofar as they alter the prevalence of the disorder or the population being classified as having a disorder. Literature on the impact of these changes is limited at this time, the wording changes are minimal, and the overall concept and threshold for the disorder remain unchanged; therefore, the impact on MHSS and AMI/SMI estimates will likely be nominal.

More substantive changes have occurred with the criteria for agoraphobia (Table 3.10). In addition to being separated conceptually from panic disorder (meaning those who were excluded from having agoraphobia cause by panic disorder will now qualify for the agoraphobia diagnosis), a more formalized set of criteria have been developed. First, a formalized threshold for the number of fear or anxiety provoking situations has been two or more situations from a list of five. Second, in DSM-5 the anxiety is not restricted to a fear of having a panic attack, as it was in DSM-IV, rather the anxiety is related to a fear of having trouble escaping or obtaining help should panic-like symptoms or other incapacitating or embarrassing symptoms occur. Third, a new criterion is that the fear or anxiety must be consistent and that the same situation almost always provokes the reaction. Fourth, the fear or anxiety has to be out of proportion to the situation. Fifth, a duration criterion has been added stating that the fear, anxiety, or avoidance typically lasts 6 months or more. Finally, the agoraphobia symptoms must cause significant distress and impairment.

All of these changes make it difficult to determine the change in prevalence or characteristics of individuals meeting DSM-5 diagnostic criteria for agoraphobia. Although the disaggregation of agoraphobia from panic disorder would increase the prevalence of agoraphobia, the more stringent diagnostic criteria may counter that increase to some extent. In addition to the alterations in prevalence, a direct concern for MHSS and subsequently how the definition of SMI and AMI is operationalized is that individuals who met DSM-5 criteria for agoraphobia may differ substantively from individuals who met DSM-IV agoraphobia (with or without panic disorder). Currently, there is no quantitative data on this possibility in children or adults.

3.3.5. Specific Phobia (MHSS)

Specific phobia is an anxiety disorder characterized by fear or anxiety about the presence of a specific object or situation [2]. DSM-IV to DSM-5 criteria changes for specific phobia consists of numerous minor wording changes (Table 3.11). The two major changes include the elimination of the DSM-IV requirements that the person recognizes that the fear is excessive or unreasonable and a specification that the duration for everyone is typically 6 months or longer (as opposed to requiring that minimum duration just for children). The elimination of the DSM-IV criterion that specifies that the person has insight that the fear is excessive or unreasonable is expected to increase the prevalence of specific phobia; however, research into the degree of impact is inconsistent [77]. In a study of 3,000 psychiatric outpatients (average age = 38.5, range not specified) and 1,800 prospective bariatric surgical patients (average age = 42.3, range not specified) receiving care through a Rhode Island hospital system, comparisons of diagnostic rates with and without the insight criterion indicated that less than 1 additional percent of each population met criteria for lifetime-specific phobia as a result of dropping the insight criterion [78]. Of the 12.3 percent of psychiatric outpatients who met either DSM-IV or DSM-5 criteria for lifetime specific phobia, 4.1 percent met criteria without having insight into the excessive/unreasonableness of their fear. This is consistent with other research that identified age differences in criterion endorsement for specific phobia [77]. In a study of older, community-sampled adults (aged 65 or older), the prevalence of past year specific phobia was 2.0 percent, and over half of those meeting DSM-IV specific phobia criteria (excluding the insight criterion) did not have insight into the excessive/unreasonableness of their fear [79]. This suggests that the overall prevalence of specific phobia may be slightly increased (approximately 1 percent) under DSM-5 criteria (because of the removal of the insight criterion), thereby affecting AMI/SMI estimation, but this effect may not be uniform across all ages.

Table 3.11

DSM-IV to DSM-5 Specific Phobia Comparison.

The change in the duration criterion was recommended to increase reliability of the diagnosis among adults [77]; however, there are little data about how significant this change will be with respect to population prevalence estimates and any resulting misclassification, thereby making it difficult to determine the impact on MHSS and subsequently NSDUH. Anecdotal evidence suggests that if the duration of specific phobia is sufficiently long that this change is likely to have little impact on prevalence estimates [77].

3.3.6. Social Phobia (MHSS)

Social phobia is an anxiety disorder characterized by fear of social situations wherein the individual may be exposed to scrutiny by others [2]. Diagnostic criteria for social phobia have undergone several minor wording changes from DSM-IV to DSM-5 (Table 3.12). One additional change is anticipated to have a broader impact on estimates of social phobia. In DSM-IV, criteria for social phobia required that an individual “recognizes that the fear is excessive or unreasonable” (i.e., the insight criterion). In DSM-5 this has been changed to note that “the fear or anxiety is out of proportion to the actual threat posed by the social situation and to the sociocultural context.” This means that the patient does not need to have insight that their fear is unreasonable, so long as the clinician can determine that the fear is unreasonable.

Table 3.12

DSM-IV to DSM-5 Social Phobia Comparison.

In a study of 3,000 psychiatric outpatients (average age = 38.5, range not specified) and 1,800 prospective bariatric surgical patients (average age = 42.3, range not specified) receiving care through a Rhode Island hospital system, comparisons of diagnostic rates for social phobia with and without the insight criterion indicated that less than 1 additional percent of each population met criteria for lifetime social phobia after dropping this criterion [78]. Among the 31.2 percent of psychiatric outpatients who had lifetime social phobia (without applying the insight criterion restriction), only 2.1 percent did not have insight into the excessive or unreasonableness of their fear. The results were slightly lower among the 12.1 percent of bariatric surgery patients who met lifetime criteria for social phobia. Only 3.7 percent of bariatric patients meeting lifetime criteria for social phobia did not have insight into their condition. This suggests that the removal of the insight criterion will lead to a slight increase in the prevalence of social phobia; however, this study was conducted in a sample of patients (psychiatric and bariatric) from one hospital system and makes direct applicability to MHSS difficult.

3.3.7. Obsessive-Compulsive Disorder (MHSS)

OCD, formerly classified as an anxiety disorder and given its own classification under DSM-5 (Obsessive-Compulsive and Related Disorders, which also includes hoarding, trichotillomania, and excoriation disorders), is characterized by the presence of obsessions (recurrent, persistent thoughts that cause anxiety or distress) and/or compulsions (repetitive behaviors or mental acts) that cause clinically significant distress or impairment [1,2]. OCD criteria have undergone significant changes in diagnostic criteria that are anticipated to have a substantial impact on prevalence estimates (Table 3.13). The primary changes include a slight but significant wording change in the definition of obsessions, the removal of three exclusionary criteria, and a more detailed explanation of the remaining exclusionary diagnoses. In DSM-IV obsessions were defined as “recurrent persistent thoughts, impulses, or images…” under DSM-5 criteria, obsessions were defined as “recurrent and persistent thoughts, urges, or images….” Two of the eliminated exclusion criteria also apply to the assessment of obsessions. In DSM-IV, the obsessions could not be simply excessive worries about real-life problems. In DSM-5 this criterion has been removed. Second, DSM-IV criteria required that a person recognize that the obsessions were a product of the person’s own mind, whereas DSM-5 has eliminated this exclusion criterion. Third, DSM-5 also dropped the overall (not obsession specific) criterion that the person recognize the obsessions or compulsions were excessive or unreasonable. Finally, DSM-5 expanded the examples of exclusionary diagnoses (other diagnoses that might better explain the symptoms) to note a much larger list of potentially similar disorders.

Table 3.13

DSM-IV to DSM-5 Obsessive-Compulsive Disorder Comparison.

The removal of several exclusionary criteria may contribute to an increase the population prevalence of OCD under DSM-5 criteria. However, to date, prevalence estimates for the DSM-5 revised diagnosis of OCD have not been identified to evaluate the impact of diagnostic changes on population estimates. Regarding the impact on AMI/SMI estimation, the existing DSM-IV estimates would underestimate the prevalence of DSM-5 AMI/SMI because individuals who were classified as not having OCD because of the exclusion criteria would have been classified as having no mental illness, unless they met criteria for another disorder assessed in MHSS. Comorbidity of OCD is high; reports from the National Comorbidity Survey-Replication (NCS-R) indicate that 90.0 percent of people with a lifetime DSM-IV diagnosis of OCD also met criteria for another lifetime DSM-IV diagnosis [80]. This finding suggests that the diagnostic revisions would have only minimal impact on AMI and SMI estimation [80].

3.3.8. Posttraumatic Stress Disorder (MHSS)

PTSD, which was classified as an anxiety disorder under DSM-IV and has been reclassified into a new class called Trauma- and Stressor-Related Disorders, is characterized by the development of a pattern of symptoms following exposure to a traumatic event (i.e., the stressor). These symptom types include re-experiencing the traumatic event (e.g., recurrent intrusive and distressing recollections of the event), avoidance of stimuli associated with the trauma, persistent negative alterations in cognition (e.g., feelings of detachment or estrangement from others), and a numbing of general responsiveness (e.g., diminished interest in activities), and increased symptoms of arousal (e.g., exaggerated startle response).

DSM-5 criteria for PTSD differ significantly from those in DSM-IV (Table 3.14). First, the stressor criterion (Criterion A) is more explicit about how an individual experienced “traumatic” events. Second, the subjective reactive criterion (Criterion A2) has been eliminated. Third, whereas there were three major symptom clusters in DSM-IV—re-experiencing, avoidance/numbing, and arousal—there are now four symptom clusters in DSM-5 because the avoidance/numbing cluster is divided into two distinct clusters: avoidance and persistent negative alterations in cognitions and mood. Fourth, for the latter cluster of symptoms, persistent negative emotional states have been added to the criteria. Fifth, the final cluster—alterations in arousal and reactivity—now also includes irritable or aggressive behavior and reckless or self-destructive behavior. Finally, PTSD is now developmentally sensitive in that diagnostic thresholds have been lowered for children and adolescents. Furthermore, separate criteria have been added for children aged 6 years or younger with PTSD.

Table 3.14

DSM-IV to DSM-5 Posttraumatic Stress Disorder Comparison.

Results from the DSM-5 field trials of patient populations suggests a 0.4 to 0.5 percent decrease in PTSD point prevalence under the revised DSM-5 criteria [51]. However, since the DSM-5 field trial relied on patient populations that may limit the generalizability to household-based samples. Moreover, MHSS uses past year diagnoses rather than point-in-time diagnoses so results should be interpreted with caution.

3.3.9. Generalized Anxiety Disorder (MHSS)

Generalized anxiety disorder (GAD) is an anxiety disorder characterized by excessive anxiety and worry that is not focused on a single trigger (e.g., fear of social situations, fear of having a panic attack, or fear of a specific event/situation). There have been very few changes made to GAD criteria in DSM-IV (Table 3.15). The DSM-IV criteria for GAD included that the anxiety and worry does not occur exclusively due to PTSD, a mood disorder, a psychotic disorder, or a pervasive developmental disorder. In DSM-5 this has been replaced with text indicating that “the disturbance is not better explained by another mental disorder.” According to reports from the DSM-5 field trials of people seeking treatment, there was a small but significant decrease in DSM-5 prevalence estimates of GAD compared with the DSM-IV clinical screening diagnosis [75]. This may suggest that current estimates for GAD generated by the MHSS are a slight overestimate of the DSM-5 population prevalence. Moreover, this is the first disorder examined thus far for which estimates are expected to decrease. However, these findings were based on assessments of treatment seeking participants at only one of the seven field trial sites; therefore, results must be interpreted with caution as site-specific effects may affect the estimates and the use of a treatment sample may not be generalizable to a nationally representative household-based survey like NSDUH.

Table 3.15

DSM-IV to DSM-5 Generalized Anxiety Disorder Comparison.

3.3.10. Anorexia Nervosa (MHSS)

Anorexia nervosa is an eating disorder characterized by an intense fear of gaining weight and the refusal to maintain a minimally normal body weight. Individuals with anorexia also exhibit a misperception of body shape and/or size. There have been several DSM-5 criteria changes (Table 3.16). In DSM-IV, a diagnosis of anorexia nervosa was excluded if the patient maintained bodyweight at or above the 85th percentile for his or her height/age. In DSM-5 this criteria is similar, but adds sex, developmental norms, and physical health and uses body mass index data. The DSM-5 adds “persistent behavior that interferes with weight gain” as an added way to meet a criteria. The DSM-5 does not include criteria on menstruating females’ absence of three consecutive menses, as the DSM-IV does. The restrictive type and binge-eating/purging types differ in that DSM-IV specifies “during the current episode” and DSM-5 specifies “during the past 3 months.” The DSM-5 adds criteria for partial and full remission, while the DSM-IV does not include this information.

Table 3.16

DSM-IV to DSM-5 Anorexia Nervosa Comparison.

There is a paucity of research into the impact of these criteria changes on disorder prevalence in the population. In a small study of 364 women, Brown et al., reported that DSM-5 criteria for anorexia nervosa produced higher lifetime frequency estimates than reported for DSM-IV anorexia nervosa [81]. However, these results may not be generalizable to findings in MHSS (and thus NSDUH) as they were drawn from a sample of female university students and not a nationally representative sample.

3.3.11. Bulimia Nervosa (MHSS)

Bulimia nervosa is an eating disorder characterized by binge eating followed by inappropriate compensatory behaviors designed to prevent weight gain. In addition, the self-evaluation of individuals with bulimia nervosa is excessively influenced by weight and body shape. DSM-IV to DSM-5 criteria changes may have a substantial impact (Table 3.17). The major change in criteria for diagnosis of bulimia nervosa is reducing the binge frequency threshold from twice per week in DSM-IV to once per week in DSM-5. The other differences include the DSM-IV differentiating between purging and nonpurging type (the DSM-5 does not) and the DSM-5 specifying criteria for partial remission, full remission, and severity, while the DSM-IV does not.

Table 3.17

DSM-IV to DSM-5 Bulimia Nervosa Comparison.

Data from an Australian cohort study of 2,822 adolescents and young adults (57.0 percent female) whose parents were recruited from antenatal clinics at a single hospital and followed through age 20, indicate that rates of bulimia nervosa are higher when applying the DSM-5 criteria versus the DSM–IV [82]. The prevalence of past month DSM-IV diagnosed bulimia nervosa in males aged 20 was 0.3 percent and the corresponding DSM-5 based rate was 0.7 percent. In females, the past month prevalence of DSM-IV bulimia nervosa was 2.4 percent among 20 year olds but 7.9 according to DSM-5 criteria. This suggests that the MHSS, if left unchanged, would underestimate DSM-5 prevalence of bulimia nervosa, particularly among females. However, the degree of underestimation is unclear because of differences in the population covered by the regional Australian study and the nationally representative (of the United States) sample drawn by NSDUH.

3.3.12. Intermittent Explosive Disorder (MHSS)

Intermittent explosive disorder (IED) is an impulse-control disorder characterized by episodes in which an individual fails to resist aggressive impulses, resulting in serious aggressive acts such as assault or significant property damage. Moreover, the degree of expressed anger is disproportionate to the degree of provocation or experienced stressor. The DSM-IV to DSM-5 revision presents important criteria changes for IED (Table 3.18). Specific criteria underwent significant wording changes, including a more precise definition of “several episodes” to twice weekly for at least 3 months. Furthermore, DSM-5 added the opportunity to meet diagnostic criteria through acts of verbal aggression, which was not available in DSM-IV, thereby slightly expanding the definition of outbursts. Finally, DSM-5 contains three additional criteria:

Table 3.18

DSM-IV to DSM-5 Intermittent Explosive Disorder Comparison.

• The recurrent aggressive outbursts specified in DSM-IV have been changed to note that they are not premeditated and are not committed to achieve some tangible objective.
• The recurrent outbursts cause either marked distress in the individual or impairment in occupational or interpersonal functioning, or are associated with financial or legal consequences.
• Patients must be at least 6 years old or the developmental equivalent (aggressive behavior that occurs as part of an adjustment disorder in 6 to 18 year olds should not be considered for this diagnosis).

According to a study of a nationally representative household-based sample of 9,282 adults aged 18 or older carried out in 2001 to 2003 (NCS-R), the lifetime prevalence of DSM-IV IED in adults was 7.3 percent and the 12-month prevalence was 3.9 percent [83]. Current literature could not be found on estimates of prevalence for IED using the DSM-5 criteria. The majority of criteria added to the DSM-5 IED diagnosis were restrictive in nature (i.e., they do not add ways to meet criteria, but add additional requirements to meet criteria), which would suggest that prevalence estimates under DSM-5 would be lower than that of DSM-IV. However, it is difficult to determine how much adding verbal aggression (the only expansive criteria change) would expand inclusion, making it impossible to determine the impact on MHSS estimates based on current literature.

3.3.13. Adjustment Disorder (MHSS)

Adjustment disorder, which has been reclassified from its own class in DSM-IV to the trauma- and stressor-related disorders class in DSM-5 (which also includes acute and PTSD), is a mental illness characterized by the development of clinically relevant emotional or behavioral symptoms (e.g., anxiety or depression) resulting from experiencing a psychosocial stressor(s). There have been only minor criteria changes to adjustment disorder in DSM-5, primarily adding clarification to the DSM-IV language and making note of potential cultural differences (Table 3.19). There is only one change that has the possibility of affecting disorder prevalence. The DSM-IV broadly states that the symptoms do not represent bereavement and the DSM-5 specifies the symptoms do not represent normal bereavement. This might increase the prevalence of this disorder; however, no studies were identified that quantified the impacts of this change. Of concern is the broad range of prevalence estimates produced under DSM-IV criteria, suggesting that the underlying prevalence for either version is unknown. In a cross-sectional survey of a representative sample of 3,815 patients from 77 primary health care centers in Catalonia, Spain, the point prevalence was 2.9 percent [84]. According to the DSM-5, the prevalence of adjustment disorders in outpatient mental health treatment patients ranges from approximately 5 to 20 percent. In hospital psychiatric consultation settings, it can often reach 50 percent [2]. Estimates from the 2008 to 2012 MHSS Clinical Study suggest a past year prevalence of 6.9 percent in the U.S. adult population [85].

Table 3.19

DSM-IV to DSM-5 Adjustment Disorders Comparison.

3.3.14. Psychotic Disorders (MHSS)

In DSM-5, the psychotic disorders class includes: schizophrenia; schizophreniform disorder; schizoaffective disorder; delusional disorder; brief psychotic disorder; psychotic disorder due to another medical condition; substance/medication-induced psychotic disorder; unspecified schizophrenia spectrum and other psychotic disorder; and other specified schizophrenia spectrum and other psychotic disorder. These disorders share a common set of characteristic symptoms or key features that include delusions (fixed beliefs that are not amenable to change in light of conflicting evidence); hallucinations (perception-like experiences that occur without an external stimulus); disorganized thinking/speech (e.g., frequent derailment or incoherence); grossly disorganized (e.g., childlike silliness or unpredictable agitation) or catatonic behavior (a marked decrease in reactivity to the environment, or purposeless and excessive motor activity without obvious cause); and negative symptoms such as affective flattening (diminished emotional expression), avolition (lack of motivation to achieve meaningful goals), or alogia (diminished speech output) [2].

Changes between DSM-IV and DSM-5 include the dropping of shared psychotic disorder and the reclassification of psychotic disorder not otherwise specified into unspecified schizophrenia spectrum and other psychotic disorder, and other specified schizophrenia spectrum and other psychotic disorder (Table 3.20). Other changes between the two editions of the DSM include minor wording changes and larger changes, including

Table 3.20

DSM-IV to DSM-5 Psychotic Disorders.

• removing the note in Criterion A of schizophrenia that specifies, in cases where there are bizarre hallucinations, or those consisting of a voice keeping a running commentary, or two voices conversing that only one Criterion A symptom is necessary;
• adding a stipulation in Criterion A of schizophreniform disorder that at least one of the two or more schizophrenia Criterion A symptoms must be delusions, hallucinations, or disorganized thinking/speech;
• removing the stipulation in Criterion A of delusional disorder that the delusions be nonbizarre; and
• adding a criterion requiring clinically significant distress or impairment for the diagnoses of psychotic disorder due to another medical condition, and substance/medication-induced psychotic disorder.

The MHSS does not assess specific psychotic disorders, instead it includes a screener for psychotic symptoms (i.e., delusions and hallucinations) that are psychogenic in nature (i.e., delusions or hallucinations that are not possibly or definitely the result of the use of medications, drugs, or alcohol or a general medical condition such as a high fever). Because these symptoms may also be present in mood disorders with psychotic features, their presence alone cannot be used make a diagnosis of a psychotic disorder. The purpose of the screener for psychotic symptoms in the MHSS has been to identify people experiencing delusions and/or hallucinations and classify them as having SMI. Whereas the changes reflected in DSM-5 do have an effect on the hierarchy within this class of disorders, there are no changes that will have an effect on the use of a screener of this type for psychotic symptoms as an indicator of SMI.

3.4.1. Neurodevelopmental Disorders

Attention Deficit/Hyperactivity Disorder

Attention-deficit/hyperactivity disorder (ADHD) is a chronic neurodevelopmental disorder characterized by a persistent and pervasive pattern of inattention and/or hyperactivity-impulsivity that interferes with functioning or development. ADHD among adults is associated with significant functional impairment. Community-based studies have repeatedly shown that adults with ADHD were less likely to have graduated high school or completed college and were more likely to have been arrested, abuse substances, be unemployed, and report dissatisfaction with their personal and professional lives [58,86–89].

The diagnostic criteria for ADHD in DSM-5 are similar to those in DSM-IV (Table 3.21). The same 18 symptoms are used as in DSM-IV, and continue to be divided into two symptom domains (inattention and hyperactivity/impulsivity), of which at least 6 symptoms in 1 domain are required for diagnosis. The primary reason for excluding ADHD from MHSS was because ADHD was largely conceptualized as a childhood disorder. However, several changes have been made in DSM-5 that might suggest considering it for inclusion in future MHSS’s. First, examples have been added to criteria to facilitate application across the life span rather than focusing primarily on children. Second, the onset criterion has been changed from “symptoms that caused impairment were present before age 7 years” to “several inattentive or hyperactive-impulsive symptoms were present prior to age 12.” Third, a symptom threshold change has been made for adults to reflect substantial evidence of clinically significant ADHD impairment, with the cutoff for ADHD of five symptoms, instead of the six required for younger people (this includes inattention, hyperactivity, and impulsivity). Finally, ADHD was placed in the neurodevelopmental disorders chapter to reflect brain developmental correlates with ADHD and the DSM-5 decision to eliminate the DSM-IV chapter that includes all diagnoses usually first made in infancy, childhood, or adolescence.

Table 3.21

DSM-IV to DSM-5 Attention-Deficit/Hyperactivity Disorder Comparison.

The majority of ADHD criteria changes focused on improving detection of ADHD among adults [90], thereby increasing the relevance to the estimation of AMI and SMI. DSM-IV ADHD was the most common neurodevelopmental disorder among children and adults and has the capacity to significantly raise the prevalence of mental illness in MHSS if it were included. Prevalence estimates of DSM-5 diagnosed ADHD among adults have not been identified. However, a point prevalence estimate of DSM-IV ADHD in adults aged 18 to 44, was 4.4 percent in 1990 to 1992, based on a subsample of respondents in the nationally representative NCS-R [84]. This is likely to be an underestimate given the relaxed diagnostic threshold under DSM-5 (five symptoms instead of six) and expanded age of onset (from age 7 to age 12) [90]. Notably, the comorbidity of ADHD with other disorders was estimated to be up to 47.1 percent (for any anxiety disorder [84]) indicating that the prevalence increase in the MHSS estimates, should ADHD be included, will not increase to the total prevalence of ADHD since about 50 percent of cases are being identified as cases in another diagnostic module currently assessed. However, the prevalence of this disorder and evidence of its clinical implications among adults makes this disorder a strong candidate for inclusion in any future DSM 5-based NSDUH clinical study of mental disorders and AMI/SMI.

3.4.2. Schizophrenia Spectrum and Other Psychotic Disorders

Schizophrenia

Schizophrenia is a mental disorder characterized by delusions, hallucinations, and/or disorganized speech, which cause significant distress and functional impairment. These symptoms involve a range of cognitive, behavioral, and emotional dysfunctions although there is great variation across individuals presenting with these symptoms. Schizophrenia is considered a chronic mental illness with lifetime prevalence between 0.3 percent and 0.7 percent. Despite its low prevalence, schizophrenia is responsible for over $7.6 billion annually in direct costs (e.g., mental health service use, law enforcement costs) and an additional $32.4 billion in annual indirect costs (e.g., loss of productivity) [91].

Two changes have been made to DSM-5 Criterion A for schizophrenia (Table 3.22). First, the special attribution of bizarre delusions and Schneiderian first-rank auditory hallucinations (e.g., two or more voices conversing) has been eliminated. Under DSM-IV criteria, if the delusions were bizarre or if there were Schneiderian first-rank auditory hallucinations then only this one symptom was necessary to meet Criterion A. This special attribution was removed because of the nonspecificity of Schneiderian symptoms and the poor reliability in distinguishing bizarre from nonbizarre delusions. Therefore, in DSM-5, two Criterion A symptoms are required for any diagnosis of schizophrenia. The second change in DSM-5 is the addition of a requirement in Criterion A that the individual must have at least one of these three symptoms: delusions, hallucinations, and disorganized speech. At least one of these core “positive symptoms” is necessary for a reliable diagnosis of schizophrenia. In addition to these two Criterion A changes, the DSM-IV subtypes of schizophrenia (i.e., paranoid, disorganized, catatonic, undifferentiated, and residual types) were eliminated because of their limited diagnostic stability, low reliability, and poor validity. Instead, a dimensional approach to rating severity for the core symptoms of schizophrenia was included to capture the heterogeneity in symptom type and severity expressed across individuals with psychotic disorders.

Table 3.22

DSM-IV to DSM-5 Schizophrenia Comparison.

The MHSS clinical study already assesses some symptoms of schizophrenia through its use of a psychotic symptom screener. Therefore, the consideration is whether to add a separate module to specifically assess schizophrenia. Despite being one of the most debilitating and costly mental illnesses with a substantial public health impact, schizophrenia has a very low prevalence. Also, its assessment is complicated by one important factor. Although the DSM-IV and DSM-5 SCID have modules for assessing schizophrenia, it was designed to be administered in person by a clinician. The MHSS uses a telephone interviewing procedure. This works well for almost all psychiatric symptoms, which are based on patient self-report. However, disorganized speech, disorganized behavior, and negative symptoms (symptom 5 of Criteria A) are based on clinician observation, not by patient self-report. These symptoms are difficult to assess over the phone without visual cues. This concern, combined with the very low prevalence of the disorder and the fact that the estimation SMI includes psychotic symptoms, continues to justify the exclusion of the schizophrenia-specific module in the MHSS.

3.4.3. Bipolar and Related Disorders

Bipolar II Disorder

Bipolar II disorder shares many similarities with bipolar I; it is characterized by periods of major depression and periods of elevated mood. The primary difference between bipolar II and bipolar I is that a person with bipolar II has only met criteria for a hypomanic episode and never a full manic episode (as in bipolar I) in their lifetime. The differences between hypomanic (Table 3.8) and full manic episodes (Table 3.6) are in duration (the duration criterion for hypomanic episodes is shorter [4 days vs. 1 week]) and in functional impairment. Hypomanic episodes are not accompanied by significant functional impairment.

Table 3.8

DSM-IV to DSM-5 Hypomania Criteria Comparison.

Criteria changes to the overall bipolar II disorder (Table 3.23) from DSM-IV to DSM-5 have included minor wording changes; however, there have been changes to the criteria for MDE (discussed in Section 3.2.1) and to hypomanic episodes (Table 3.8), which are integral to the diagnosis of bipolar II. The changes in MDE criteria, as already discussed, have an as yet undetermined effect, but they are expected to raise the prevalence slightly. The primary change to hypomanic episodes is a restrictive change. Under DSM-IV, the main feature of a hypomanic episode was described as “a distinct period of persistently elevated, expansive, or irritable mood, lasting throughout at least 4 days, that is clearly different from the usual nondepressed mood” [1, p. 338]. Under DSM-5, this has been changed to “a distinct period of abnormally and persistently elevated, expansive, or irritable mood and abnormally and persistently increased activity or energy, lasting at least 4 consecutive days and present most of the day, nearly every day” [2, p. 133]. These changes, which add an additional qualifier to the criteria, will likely result in a decrease in the prevalence of hypomanic episodes in the population. Results of the DSM-5 field trials suggest that the decrease in hypomania prevalence will result in a substantial reduction in bipolar II diagnosis, with the point prevalence of bipolar II disorder decreasing from 0.18 percent under DSM-IV criteria to 0.09 percent under DSM-5 criteria [75]. However, the results provided by the field trial also note that there was insufficient sample size (bipolar II was only assessed at one clinical site) to generate a stable prevalence estimate. National estimates derived from the NESARC suggest a lifetime DSM-IV prevalence of bipolar II disorder to be 1.12 percent. Considering the results of the field trials, the DSM-5 lifetime prevalence of bipolar II disorder to be very roughly between 0.56 percent (half that of the DSM-IV estimate) and the DSM-IV NESARC [92]. However, even these approximations must be made with caution because of the large variance around the estimate produced in the field trials.

Table 3.23

DSM-IV to DSM-5 Bipolar II Disorder Comparison.

The potential inclusion of bipolar II in the MHSS has three major considerations. First, is the overall prevalence, although a lifetime prevalence 1.12 percent is considered rare by epidemiological standards (rare is usually defined as less than 5 percent [93]), it translates into a large number of affected people. Second, lifetime treatment seeking among individuals with bipolar II is as high as 90 percent, suggesting that the disorder may place a demand on treatment services that exceeds that of other disorders (e.g., 27 percent among those with separation anxiety disorder) [94]. Third, there is the practical consideration of how many individuals with bipolar II are being identified as “cases” in the MHSS even though bipolar II is not assessed because they have a comorbid disorder that is assessed (this was the primary reason for bipolar II exclusion in prior MHSS). Evidence suggests that up to 48.0 percent of respondents in NESARC with a lifetime history of DSM-IV bipolar II disorder also met criteria for one of the anxiety disorders that is already assessed in MHSS [92]. Moreover, even though the hypomania for bipolar II is not being assessed in MHSS, MDE is assessed and individuals with bipolar II in the past year experiencing depression would be counted even if hypomania is not assessed.

3.4.4. Depressive Disorders

Premenstrual Dysphoric Disorder

Premenstrual dysphoric disorder (PMDD) had been included in DSM-IV as a “criteria set provided for further study,” and has been included in DSM-5 as an official disorder. PMDD is characterized by low mood, anxiety, and irritability symptoms during the premenstrual phase of a woman’s cycle that remits around the onset of menses. Criteria changes to PMDD (Table 3.24) from DSM-IV to DSM-5 include minimal wording changes to several of the criteria. This diagnosis continues to include several diagnostic exclusions for other medical conditions, mental health disorders, and the effects of substances used that may better explain the occurrence of symptoms. Past year prevalence estimates among menstruating women are between 1.8 and 5.8 percent. However, it is noted that estimates are significantly inflated if they are based on retrospective rather than prospective reports. This is an important consideration in determining whether PMDD should be included in the MHSS now that it is considered a verified condition.

Table 3.24

DSM-IV to DSM-5 Premenstrual Dysphoric Disorder Comparison.

3.4.5. Anxiety Disorders

Separation Anxiety Disorder

Separation anxiety disorder (SAD) is a psychological condition in which an individual experiences excessive anxiety, fear, or distress regarding separation from home or from people to whom the individual has a strong emotional attachment (e.g., a parent, grandparents, or siblings; Table 3.25). SAD is the inappropriate and excessive display of fear and distress when faced with situations of separation from the home or from a specific attachment figure. The anxiety that is expressed is categorized as being atypical of the expected developmental level and age and can be diagnosed among adults and children. The severity of the symptoms ranges from anticipatory uneasiness to full-blown anxiety about separation. SAD may cause significant negative effects within a person’s everyday life, as well. These effects can be seen in areas of social and emotional functioning, family life, physical health, and within the academic context. The duration of this problem must persist for at least 4 weeks and must present itself before a person is 18 years of age to be diagnosed as SAD [2].

Table 3.25

DSM-IV to DSM-5 Separation Anxiety Disorder Comparison.

The primary change from DSM-IV to DSM-5 has been the reclassification of SAD from the Disorders Usually Diagnosed in Infancy, Childhood, and Adolescence section to Anxiety Disorders. Criteria changes from DSM-IV to DSM-5 are largely wording changes rather than major conceptual differences. The largest differences are related to broadening the criteria to better align with the presentation of SAD among adults. This is noteworthy as there is evidence that while prevalent among treatment samples, SAD is often not diagnosed among adults because of a clinical description that aligns only with childhood-based manifestations despite evidence of significant impairment among adults [95]. Data from the nationally representative NCS-R indicated the prevalence of DSM-IV SAD among adults to be 1.9 percent in the past 12 months and 6.6 percent across the lifetime [96]. Females were more likely to have SAD than males and approximately a third of SAD cases in childhood persisted into adulthood. Comorbidity in SAD is high, with an estimated 88.5 percent of people meeting criteria for another disorder assessed in the NCS-R. This suggests that a large number of adult with SAD would be classified as having AMI through their endorsement of criteria for another disorder.

3.4.6. Obsessive-Compulsive and Related Disorders

Body Dysmorphic Disorder

Body dysmorphic disorder (BDD) is a mental illness characterized by a preoccupation with a perceived physical defect or flaw that is excessive and causes significant distress or functional impairment (Table 3.26). There have been several important changes in BDD criteria from DSM-IV to DSM-5. First, BDD has been reclassified from somatoform disorders in DSM-IV to obsessive-compulsive and related disorders under DSM-5. Second, DSM-5 BDD has an added diagnostic criterion indicating that the patient must have had repetitive behaviors or mental acts that were in response to preoccupations with perceived defects or flaws in physical appearance. Third, a “with muscle dysmorphia” specifier has been added to reflect a growing literature on the diagnostic validity and clinical utility of making this distinction in individuals with BDD. Finally, the delusional variant of BDD (which identifies individuals who are completely convinced that their perceived defects or flaws are truly abnormal in appearance) is no longer coded as both a delusional disorder (somatic type) and BDD. Under DSM-5 this presentation is designated only as BDD with the “absent insight/delusional beliefs” specifier and not as a delusional disorder.

Table 3.26

DSM-IV to DSM-5 Body Dysmorphic Disorder Comparison.

Under DSM-IV criteria, BDD was a fairly common disorder affecting approximately 2.4 percent of the general population at any time (point prevalence) as estimated by a random sample telephone survey conducted in 2004 among 2,513 adults residing in the United States [97]. As a comparison, this is similar to point prevalence estimates of DSM-IV defined generalized anxiety disorder. Estimates under DSM-5 criteria are not currently available. However, criteria-induced changes in the estimates are likely to be minor. Over 90 percent of people with BDD report repetitive behaviors or mental acts in response to their preoccupation with a perceived physical deficit [98], and this was the only restrictive change to diagnostic criteria. BDD affects both males and females and has been identified in children as young as age 5 and as old as age 80 [98]. Moreover, patients with BDD report that unwanted, anxiety provoking obsessions related to BDD cause significant distress (e.g., avoidance of social situations due to anxiety of being ridiculed) and the repetitive behaviors are time consuming, functionally impairing (e.g., being late for work due to compulsive compensatory behavior), and can be dangerous (e.g., skin picking leading to infection). Suicidal thoughts is a significant concern among people with BDD; suicidal ideation is as high as 80 percent in this population and one in four make a suicide attempt [98]. There are very few prospective studies of BDD, but retrospective studies suggest a mean onset of BDD around age 16 and indicate a chronic course with a low probability of remission without treatment [98]. The overall prevalence and significant clinical implications suggest that BDD may be suitable for inclusion in the MHSS. However, approximately 75 percent of those with BDD meet criteria for MDD as well, and therefore are also being captured as having SMI. However, the level of distress and severity of consequences associated with this disorder may merit BDD being included even if the comorbidity with other disorders is high.

Trichotillomania

Trichotillomania (or hair-pulling disorder) is the compulsive urge to pull out (and in some cases, eat) one’s own hair leading to noticeable hair loss, distress, and social or functional impairment (Table 3.27). Common areas for hair to be pulled out are the scalp, eyelashes, eyebrows, legs, arms, hands, and the pubic areas. Criteria changes from DSM-IV to DSM-5 include the reclassification of the disorder from impulse-control disorders to obsessive-compulsive and related disorders, the elimination of several criteria (sensations of tension prior and relief after the hair pulling), and the addition of a criteria specifying that there have been repeated attempts to decrease or stop hair pulling. The prevalence of lifetime or past year DSM-5 trichotillomania is unknown. Moreover, owing to the social implications the disorder is often underreported. The 12-month DSM-IV prevalence was estimated to be 1 percent to 2 percent among adults and adolescents [2,99]. Lifetime prevalence is estimated to be between 0.6 and 3.4 percent, with the disorder much more common among females [100]. However, these estimates come from studies that assessed college students and are not representative of the national population. Estimates of lifetime prevalence of trichotillomania in inpatient psychiatric samples have ranged from 1.3 to 4.4 percent [101]. The overall prevalence in the United States is unknown. Trichotillomania can cause significant distress to individuals with this disorder who may avoid social situations and cause permanent damage to the hair growth and quality and follows a chronic course with waxing and waning symptom levels [101].

Table 3.27

DSM-IV to DSM-5 Trichotillomania (Hair-Pulling Disorder) Comparison.

It is difficult to evaluate the utility of adding trichotillomania to the MHSS because of the limited data available. Although there is evidence of a prevalence and clinical significance comparable to other disorders assessed in the MHSS, the prevalence estimates are based on data that may not be generalizable to the national population and therefore NSDUH and MHSS. Moreover, data on the comorbidity of trichotillomania with other disorders is lacking. This makes it difficult to assess how many respondents with trichotillomania would be diagnosed with another, already assessed, disorder.

Excoriation (Skin Picking) Disorder

Excoriation disorder is an obsessive-compulsive–related disorder that is characterized by repetitive skin picking (Table 3.28). This disorder is associated with clinical distress and impairment as individuals often spend a great deal of time engaging in skin picking behaviors that results in lesions, possible infection, and scaring. The course of excoriation disorder is typically chronic, with waxing and waning in intensity in the absence of treatment. Excoriation is a new disorder in DSM-5 with an estimated lifetime prevalence of at least 1.4 percent among adults in the general population. The level of comorbidity in the general population is unknown but expected to be high, particularly with OCD, depression, and anxiety. Currently, there is little data available with regard to this disorder beyond general prevalence estimates. Prior to DSM-5, individuals with this disorder would likely have been characterized as having OCD or an impulse-control disorder not otherwise specified.

Table 3.28

Excoriation (Skin Picking) Disorder.

Hoarding Disorder

Hoarding disorder is a new diagnosis in the DSM-5, and therefore was not a consideration in the prior development of the MHSS. Hoarding disorder is defined as a persistent difficulty in parting with possessions, which is independent of their value and results in a congestion of possessions that clutter living areas and compromise their intended use to such a degree that there is significant distress and impairment (Table 3.29). Compulsive hoarding behavior has been associated with health risks, impaired functioning, economic burden, and adverse effects on friends and family members. When clinically significant enough to impair functioning, hoarding can prevent typical uses of space so as to limit activities such as cooking, cleaning, moving through the house, and sleeping. It can also be dangerous if it puts the individual or others at risk from fire, falling, poor sanitation, and other health concerns. Hoarding is considered to be closely aligned with other obsessive and compulsive disorders (under DSM-IV hoarding was considered a symptom of OCD) and has been included in that disorder class. However, there are distinct differences that led to the disorder being separated from OCD [102]. For example, thoughts about hoarding are not usually perceived to be intrusive or unwanted but rather a normal part of the individual’s thoughts. Moreover, distress is usually as a result of a third party’s interference, such as relatives or the authorities, and as a result of the thought or action of removing the item rather than the acquisition.

Table 3.29

DSM-5 Hoarding Disorder.

DSM-5–based estimates for hoarding disorder range from 2 percent to 5 percent among adults based on a review of available epidemiological literature [2,102]. Descriptive epidemiological studies suggest that hoarding is found in similar proportion among males and females, with some sign of a slightly higher prevalence in males. Symptoms usually worsen across the lifespan and people with hoarding disorder frequently lack insight into their behavior. The average age of onset, based on retrospective studies, is estimated at age 12 to 13 with symptoms worsening over the lifespan and typically reaching clinical significance around the mid-30s [102]. Comorbidity among adults with hoarding disorder, like most disorders, is high, with approximately 75 percent having a comorbid depressive or anxiety disorder at some point in their lifetime [2].

3.4.7. Trauma- and Stressor-Related Disorders

Acute Stress Disorder

Acute stress disorder (ASD) is the development of symptoms following exposure to a traumatic event (i.e., an event resulting in extreme, disturbing, or unexpected fear; stress or pain; and that involves or threatens serious injury, perceived serious injury, or death). ASD is a potential physical and mental response to feelings (both perceived and real) of intense helplessness. In addition to exposure to a traumatic event, the criteria for an ASD diagnosis include 14 potential symptoms across five categories:

• intrusion symptoms (e.g., involuntary thoughts, memories, or dreams relating to the traumatic event);
• negative mood (i.e., the inability to experience positive emotions);
• dissociative symptoms (e.g., the inability to remember an important part of the traumatic event);
• avoidance symptoms (e.g., efforts are made to avoid external reminders of the traumatic event); and
• arousal symptoms (e.g., sleep disturbance or an exaggerated startle response).

A diagnosis of ASD requires 9 or more of the 14 symptoms developing or worsening after the traumatic event and persisting for 3 days to 1 month after the exposure. (Note: if the symptoms persist beyond 1 month after the exposure, a diagnosis of PTSD may be appropriate). The symptoms must cause clinically significant distress or impairment in at least one essential area of functioning.

In DSM-5, the criterion relating to experiencing a traumatic event is more limited than in DSM-IV (Table 3.30). DSM-IV required that exposure to a traumatic event may include “being confronted” with an event that involved actual or threatened death or serious injury or threat to the physical integrity of self or others. DSM-5 provides a more definitive description of being “confronted with an event” by stipulating that when a traumatic event involves learning about something that has happened to a close family member or close friend, the event must have been violent or accidental. Also, the DSM-IV criterion regarding the subjective reaction to the traumatic event (e.g., “the person’s response involved intense fear, helplessness, or horror”) has been eliminated in DSM-5. Based on evidence that acute posttraumatic reactions are very heterogeneous and that DSM-IV’s emphasis on dissociative symptoms is overly restrictive, individuals may meet diagnostic criteria in DSM-5 for ASD if they exhibit at least 9 of 14 listed symptoms cross any of the categories.

Table 3.30

DSM-IV to DSM-5 Acute Stress Disorder Comparison.

Prevalence rates for ASD among the population are unavailable; however, prevalence among people exposed to trauma is estimated to be between 6 and 50 percent, depending on the nature of traumatic event [2]. A very rough estimate of the past year prevalence of ASD can be computed using a published estimate of the percentage of people who have experienced an event that qualifies as a traumatic event (21.0 percent [87]) by a published mean rate of ASD following various traumatic events (13.0 percent [103]) to arrive at a past year prevalence rate for ASD of 2.7 percent. ASD was introduced in DSM-IV as a means of identifying people having extreme stress reactions immediately following exposure to a traumatic event.

The extant literature indicates that ASD has some overlap with two other disorders that have been measured in the MHSS—PTSD and adjustment disorder—although it is noted that ASD is not reliably predictive of PTSD nor does ASD always precede PTSD [103,104]. Likewise, in the absence of an assessment for ASD, adjustment disorder may identify people who would have received an ASD diagnosis, but the diagnostic criteria for adjustment disorder are much broader than those for ASD [105]. Because the symptoms of ASD are relatively parallel to those of PTSD with the exception of the duration, it may be possible to assess for symptoms of ASD without adding appreciably to the burden of the clinical interview (by providing a coding option for an ASD diagnosis in the PTSD module if the requisite number of symptoms have been experienced and the duration has been at least 3 days but not more than 1 month).

3.4.8. Somatic Symptom and Related Disorders

Somatic Symptom Disorder

Somatic symptom disorder in DSM 5 reflects a significant change from the somatoform disorders included in DSM-IV (Table 3.31). Somatic symptom disorder is characterized by distressing somatic symptoms (i.e. physical symptoms such as pain or discomfort), that result in excessive thoughts, anxiety, or concern about health. In contrast, the somatoform disorders in DSM-IV had required that there be multiple somatic symptoms that were medically unexplained. Because of this dramatic change in the diagnostic criteria, prevalence rates of this disorder are unknown.

Table 3.31

DSM-IV to DSM-5 Somatic Symptom Disorder Comparison.

The potential inclusion of somatic symptom disorder to the MHSS has two major considerations. First, there are multiple mental health disorders that may include somatic symptoms (e.g., depressive disorders, panic disorder, and obsessive-compulsive disorder). If the somatic complaints are experienced solely in the context of another disorder, then a diagnosis of somatic symptom disorder is not appropriate. Second, in cases where the somatic symptoms are associated with a physical illness, the degree of impairment must exceed the impairment expected to result from the physical illness. This may present challenges within the context of a telephone assessment conducted by a person not trained in impairment associated with physical illnesses.

Illness Anxiety Disorder

Formerly known as hypochondriasis in DSM-IV, illness anxiety disorder is a somatic symptom disorder characterized by the excessive preoccupancy or worry about having a serious illness in the absence of actual medical condition. Illness anxiety disorder often involves fears that minor bodily symptoms may indicate a serious illness, constant self-examination and self-diagnosis, and a preoccupation with one’s body. Many individuals with illness anxiety disorder express doubt and disbelief in the doctors’ diagnosis, and report that doctors’ reassurance about an absence of a serious medical condition is unconvincing or short-lasting. The disorder can become a disabling torment for the individual, as well as his or her family and friends.

Illness anxiety disorder has undergone a number of criteria changes from its predecessor, hypochondriasis (Table 3.32). These changes reflect not only changes in understanding of the disorder itself, but the understanding of the entire class of somatoform disorders. In addition, hypochondriasis was renamed largely because the name was perceived as pejorative and not conducive to an effective therapeutic relationship. As a result of the criteria changes, most individuals who would have been diagnosed with hypochondriasis under DSM-IV criteria would now receive a DSM-5 diagnosis of somatic symptom disorder because of the significant somatic symptoms they report. In DSM-5, individuals with high health anxiety without somatic symptoms would receive a diagnosis of illness anxiety disorder (unless their health anxiety was better explained by a primary anxiety disorder, such as generalized anxiety disorder). These changes would result in significant reduction in the prevalence estimates of illness anxiety disorder compared to hypochondriasis. The DSM-5 estimates that illness anxiety disorder affects between 1.3 percent to 10.0 percent of adults in the general population over a 1- to 2-year period with an equal distribution between males and females [5]. However, there are little data on DSM-5 illness anxiety disorder and no data on levels of comorbidity with other disorders. Although the prevalence of illness anxiety disorder prompts consideration for inclusion in MHSS, the lack of data makes it difficult to compare the benefit of its addition against the added respondent burden.

Table 3.32

DSM-IV to DSM-5 Illness Anxiety Disorder Comparison.

3.4.9. Feeding and Eating Disorders

Avoidant/Restrictive Food Intake Disorder

Avoidant/restrictive food intake disorder is a new disorder in DSM-5. It is derived from feeding disorder of infancy or early childhood in DSM-IV, but has been broadened to also include restrictive or avoidant eating beginning after the age of 6 and adapted to include symptoms of restrictive or avoidant food intake that may be seen in adolescence and adulthood. Avoidant/restrictive food intake disorder is differentiated from other feeding and eating disorders (i.e., anorexia nervosa and bulimia nervosa). In avoidant/restrictive food intake disorder, food is restricted or avoided because of sensory characteristics of the food (e.g., texture, odor, color) rather than concerns about weight or body weight/shape. The majority of changes in disorders and diagnostic criteria for eating disorders have been aimed at providing categorizations that reflect clinically meaningful differences between diagnoses (Table 3.33). Using the DSM-IV classifications, many people with an eating disorder received a diagnosis of eating disorder—not otherwise specified (ED-NOS) rather than either anorexia nervosa or bulimia nervosa. Several studies identified homogeneous groupings of symptoms among people diagnosed with ED-NOS; these groupings are reflected in the new disorders introduced in DSM-5 such as avoidant/restrictive food intake disorder and binge eating disorder [106–108].

Table 3.33

DSM-IV to DSM-5 Avoidant/Restrictive Food Intake Disorder Comparison.

Criterion A, which concerns how this disorder may be manifested, has been made more descriptive in the DSM-5 than its counterpart in the DSM-IV. Specifically, it calls for either significant weight loss, significant nutritional deficiency, dependence on tube feeding or nutritional supplements, or marked interference with psychosocial functioning. As noted previously, the restriction that onset must have occurred by age 6 has been removed. Additional diagnostic exclusions have been added; the symptoms should not be due to another medical condition (e.g., allergies or gastrointestinal disease) or another mental health disorder (e.g., autism spectrum disorder or specific phobia).

Prevalence rates for this disorder among adults are unavailable. With the exception of pica (persistent eating of nonfood substances), disorders categorized as feeding or eating disorders are hierarchical and mutually exclusive. A diagnosis of anorexia nervosa precludes diagnoses of bulimia nervosa, binge eating disorder, and avoidant/restrictive food intake disorder; and a diagnosis of bulimia nervosa precludes a diagnosis of binge eating disorder and avoidant/restrictive food intake disorder.

Inclusion of avoidant/restrictive food intake disorder among disorders assessed in the MHSS may present several challenges because of the nature of several of the symptoms. Criterion A calls for evidence of an eating disturbance that results in a persistent desire to meet appropriate nutritional and/or energy needs. Potential signs of this include potential weight loss or a significant nutritional deficiency. Both of these indicators are typically assessed through a physical exam, often including laboratory tests. The MHSS is a telephone interview that precludes the ability of interviewers to observe visual signs of unhealthy weight loss or nutritional deficiencies. Although the interviewers will have training in psychiatric assessment, few will likely have training in nutrition or be able to effectively probe for physical symptoms related to deficiencies of specific nutrients. Likewise, Criteria B and D exclude eating disturbances that result from cultural practices, lack of food, or other health conditions. Participants who do not have access to physical and/or mental health treatment may be less knowledgeable about whether they have other conditions that may contribute to their eating disturbance.

Binge Eating Disorder

Binge eating disorder had been included in DSM-IV as a “criteria set provided for further study,” and has been included in DSM-5 as a disorder. This disorder is characterized by binge or out of control eating accompanied by significant distress about eating. Binge eating disorder is differentiated from bulimia nervosa in that there are no inappropriate compensatory behaviors (e.g., purging or excessive exercise) seen in binge eating disorder. As noted previously, this disorder was added into DSM-5 because a significant subset of people presenting with an eating disorder had exhibited binge eating behaviors that were not accompanied by any behaviors intended to compensate for the binge eating [106,109,110].

Changes between the criteria enumerated in DSM-IV and those in DSM-5 are minimal (Table 3.34). The only change, which represents a less stringent requirement in DSM-5, reduces the minimum frequency/duration of the binge eating behavior to at least once a week for 3 months (it had been at least 2 days a week for 6 months). The 12-month prevalence of binge eating disorder under DSM-IV among people aged 12 or older is 1.6 percent among females and 0.8 percent among males [2], which is higher than the 12-month prevalence of both eating disorders included in the MHSS. The 12-month prevalence of anorexia nervosa among young females is approximately 0.4 percent and the 12-month prevalence of bulimia nervosa among young females is between 1.0 and 1.5 percent [2]. The male-to-female ratio for binge eating disorder is far less skewed than those of anorexia nervosa and bulimia nervosa [111]. These are important considerations in making the decision about the inclusion of binge eating disorder in the assessment of feeding and eating disorders. The inclusion of binge eating in a DSM 5-based MHSS would likely improve the accuracy of prevalence rates for feeding and eating disorders derived from the MHSS data.

Table 3.34

DSM-IV to DSM-5 Binge-Eating Disorder Comparison.

3.4.10. Sleep-Wake Disorders

Hypersomnolence Disorder

Hypersomnolence disorder is characterized by excessive sleepiness despite having gotten at least 7 hours of sleep that results in significant distress or impairment. Criteria changes to hypersomnolence disorder (Table 3.35) from DSM-IV to DSM-5 have been made to better define the extent of the sleep/waking disturbance. Minimal wording changes have been made to several of the criteria. Criterion A has been more explicitly defined, and the frequency/duration have been revised to require that hypersomnolence occurs at least three times per week for at least 3 months (from almost daily for at least a month). This diagnosis continues to include several diagnostic exclusions for other medical conditions, mental health disorders, and the effects of substances used that may better explain the occurrence of symptoms.

Table 3.35

DSM-IV to DSM-5 Hypersomnolence Disorder Comparison.

The point prevalence of DSM-5 hypersomnolence disorder among U.S. adults is estimated at about 1 percent [2]. Mental health disorders (e.g., depression and anxiety) and other medical disorders (cancer, heart disease, Parkinson’s disease) often accompany symptoms of sleep-wake disorders. The sleep-wake class of disorders have clinical significance in that they identify sleep-wake symptoms that will not be improved through treatment of the accompanying disorders and require targeted treatment [112]. There is frequently a complex, bidirectional relationship between sleep-wake disorders and other mental health and medical problems. Identifying and understanding what factors may be affecting a participant’s sleep and waking behaviors is a formidable task. The MHSS is a telephone interview, which precludes the ability of interviewers to observe visual signs of physical disorders or other mental health disorders. MHSS participants who do not have access to physical and/or mental health treatment may be less knowledgeable about whether they have other conditions that may better explain their hypersomnolence. Based on the difficulty of adequately assessing all of the diagnostic exclusions that must be considered and the high comorbidity, it appears that hypersomnolence disorder should continue to be excluded in the MHSS.

Insomnia Disorder

Insomnia disorder is characterized by poor sleep quality or quantity, including having difficulty falling asleep, remaining asleep, or returning to sleep after awakening. Criteria changes to insomnia disorder (Table 3.36) from DSM-IV to DSM-5 have been made to better define the extent of the sleep disturbance. Minimal wording changes have been made to several of the criteria. Criterion A has been more explicitly defined, and the duration has been revised to require that insomnia is experienced at least 3 nights per week for at least 3 months (rather than 1 month as in DSM-IV). This diagnosis continues to include several diagnostic exclusions for other medical conditions, mental health disorders, and the effects of substances used that may better explain the occurrence of symptoms.

Table 3.36

DSM-IV to DSM-5 Insomnia Disorder Comparison.

The point prevalence of DSM-5 defined insomnia disorder among adults is estimated at about 6 to 10 percent [2]. Mental health disorders (e.g., depression and anxiety) and other medical disorders (cancer, heart disease, Parkinson’s disease) often accompany symptoms of sleep disorders. The sleep-wake class of disorders have clinical significance in that they identify sleep-wake symptoms that will not be improved through treatment of the accompanying disorders and require targeted treatment [112]. There is frequently a complex, bidirectional relationship between sleep-wake disorders and other mental health and medical problems. Identifying and understanding what factors may be affecting a participant’s sleep and waking behaviors is a formidable task. The MHSS is administered over the telephone, which precludes the ability of interviewers to observe visual signs of physical disorders or other mental health disorders. MHSS participants who do not have access to physical and/or mental health treatment may be less knowledgeable about whether they have other conditions that may better explain their insomnia. The potential inclusion of insomnia disorder to the MHSS has two major considerations. First, there is a high prevalence of this disorder (6 to 10 percent). Second, there is considerable difficulty of adequately assessing all of the diagnostic exclusions, which would likely add a significant burden to the clinical interview.

3.4.11. Disruptive, Impulse-Control, and Conduct Disorders

Conduct Disorder

Conduct disorder is a disorder classified in DSM-5 under “Disruptive, Impulse-Control, and Conduct Disorders.” Conduct disorder is usually first diagnosed in childhood and was categorized in DSM-IV under the category of the same name (Table 3.37). Conduct disorder is characterized by repetitive and persistent pattern of behavior that violates the rights of others or major societal norms. People with conduct disorder often show aggression to people and animals, destruction of property, deceitfulness or theft, and/or serious violations of rules. Criteria changes between DSM-IV and DSM-5 are expected to impact prevalence estimates.

Table 3.37

DSM-IV to DSM-5 Conduct Disorder Comparison.

Conduct disorder is believed to be on a spectrum between oppositional defiant disorder (see Table 3.39) and antisocial personality disorder. If a person meets criteria for antisocial personality disorder, they are precluded from receiving a diagnosis of conduct disorder. Data from wave 1 of NESARC indicate that approximately 75 percent of adults meeting conduct disorder criteria also meet antisocial personality disorder criteria [113]. Within the NESARC sample, the point prevalence of conduct disorder in the absence of antisocial personality disorder was 1.4 percent (unweighted, weighted data not reported) among males and 0.7 percent among females. Estimates of the comorbidity of conduct disorder (absent antisocial personality disorder) with other mental illnesses are available mainly in children and adolescents and it is unclear how this will generalize to adults. However, a review of studies on conduct disorder in children under age 18 suggested that 30 percent to 50 percent of youth with conduct disorder also met criteria for ADHD [99]. Conduct disorder has significant public health implications as it places a burden on both the treatment and criminal justice sectors [114] and is associated with increased risk-taking behaviors that lead to a higher prevalence of injuries, substance use, sexually transmitted infections, etc. [115]. Additional consideration might also be given to screening for antisocial personality disorder if disorder-specific estimates were to be generated using MHSS data. DSM-5 hierarchy specifies that individuals who meet both conduct disorder criteria and antisocial personality disorder criteria, only receive the antisocial personality disorder criteria. Without assessing antisocial personality disorder, it would be impossible to tease out the two disorders. Notably, not all national studies included the antisocial personality disorder exclusionary criteria (e.g., NCS-R) and for the purpose quantifying AMI this specific exclusion is not be necessary.

Table 3.39

DSM-5 Disorders Not Suggested for Inclusion in the MHSS.

3.4.12. Substance-Related and Addictive Disorders

Gambling Disorder

Gambling disorder (called pathological gambling in DSM-IV) is an addictive disorder characterized by the persistent and maladaptive risking of something of value in the hopes of obtaining something of greater value, which results in disruption of personal, family, or vocational pursuits (Table 3.38). In an important departure from DSM-IV, gambling disorder has been reclassified from impulse control disorders to the substance-related and addictive disorders class. This change reflects the increasing and consistent evidence that some behaviors, such as gambling, activate the brain reward system with effects similar to those of drugs of abuse [2].

Table 3.38

DSM-IV to DSM-5 Gambling Disorder Comparison.

There have been several important criteria changes from DSM-IV pathological gambling to DSM-5 gambling disorder. The main changes include:

• A decrease in the number of criteria required to meet the diagnostic threshold from five or more under DSM-IV to four or more under DSM-5
• The addition of a 12-month window for clustering of symptoms
• The addition of a clinically significant distress or impairment criterion
• The removal of a legal criterion.

Estimates of 12-month gambling disorder under DSM-IV ranged from 0.2 percent to 0.3 percent in the general population, with a higher prevalence among males [4]. The reduction in the number of symptoms required to meet diagnostic criteria will likely lead to an increased prevalence of DSM-5 gambling disorder. In a study of callers to a gambling helpline, the prevalence of gambling disorder increased by 9 percent because of the lowered threshold [116]. In a study by Petry et al., which included a small random sample as well as targeted sampling of intervention and treatment facilities, the prevalence of gambling disorder increased from 16.2 percent under DSM-IV to 18.1 percent under DSM-5 [117]. The effect of criteria changes on prevalence estimates in the general population are unknown, however, since the random sampled component of the Petry study was too small to derive reliable estimates.

3.4.13. Personality Disorders

DSM-5’s removal of the multiaxial system leads to the removal of the boundary between personality disorders and the other mental disorders. Personality disorders are associated with ways of thinking and feeling about oneself and others that significantly and adversely affect how an individual functions in many aspects of life. Personality disorders fall within 10 distinct types: paranoid personality disorder, schizoid personality disorder, schizotypal personality disorder, antisocial personality disorder, borderline personality disorder, histrionic personality, narcissistic personality disorder, avoidant personality disorder, dependent personality disorder, and obsessive-compulsive personality disorder. A personality disorder diagnosis requires characteristics described in the criteria to be pathological (i.e., outside the range of normal variation), persistent (i.e., frequently present over a period of at least the last 5 years with onset by early adulthood), and pervasive (i.e., apparent in a variety of contexts, such as at work and at home, or in the case of items concerning interpersonal relations, occurs in several different relationships). Aside from the removal of personality disorders from a separate axis, the 10 personality disorders did not undergo any changes to the diagnostic criteria.

Data from the NCS-R indicate that an estimated 9.1 percent of adults have met criteria for one or more personality disorders in the past year [118]. In this study, the prevalence of antisocial personality disorder was 1.0 percent, avoidant personality disorder was 5.2 percent, and borderline personality disorder was 1.6 percent. Data from the 2001–2002 NESARC suggests a prevalence rate of 4.4 percent for paranoid personality disorder, 3.9 percent for schizotypal personality disorder, 3.1 percent for schizoid personality disorder, 0.5 percent for dependent personality disorder, and 1.9 percent for histrionic personality disorder [119]. Prevalence rates for obsessive-compulsive personality disorder range from 2.1 percent and 7.9 percent, and from less than 0.1 percent to 6.2 percent for narcissistic personality disorder [2]. Criteria changes from DSM-IV to DSM-5 have included dropping two of the personality disorders that were included in Appendix B of DSM-IV (Criteria Sets and Axes Provided for Further Study): depressive personality disorder and passive-aggressive personality disorder. People with personality disorders often have significant functional impairment and a poorer treatment prognosis making these disorders of clinical importance. However, personality disorders are frequently comorbid with other disorders, meaning that their unique contribution to the assessment of AMI/SMI may be limited. Data from the NCS-R indicated that comorbidity with another DSM-IV Axis I disorder ranged from 40 percent to 85 percent depending on the specific personality disorder [118]. Development of a SCID personality disorders screener and assessments for specific disorders are under development.

3.4.14. Disorders Due to a Medical Condition, Substance Use, or Medications

Disorders caused by a medical condition can closely resemble the disorders not arising from medical conditions or substance use, including the level of impairment and need for treatment. Thus, knowing the extent of the prevalence of these disorders is important for estimating need for treatment and evaluating mental health system capacity. The assessment of several of these disorders is already a part of the MHSS procedure; therefore, their inclusion will not add substantially to the response burden. Moreover, the definition of SMI and AMI does not distinguish between the sources of these disorders. If additions are made to MHSS, the following modules could be considered.

• Anxiety (caused by a medical condition and substance induced)
• Bipolar (caused by a medical condition and substance induced)
• Depressive Disorder (caused by a medical condition and substance induced)
• Obsessive-Compulsive and Related Disorder (caused by a medical condition and substance induced)
• Sleep Disorder (substance-induced only)

Of these, only substance-induced sleep disorders are not currently assessed and would potentially increase the response burden.

3.4.15. Other Specified Disorders

Each class of disorders under DSM-IV included a “Not Otherwise Specified” diagnostic category. This diagnostic category has been broken into two categories under DSM-5: other specified disorders and unspecified disorders. Related to symptom characteristics, these categories are the same; the distinction is one of clinician information. The former category is used in situations where the clinician communicates the reason other diagnostic categories do not apply, whereas the latter is used in instances where the clinician does not specify the reasons (e.g., lack of time for full assessment in an emergency room). Otherwise, these diagnostic categories are identical and include disorder presentations characteristic of the disorder class (e.g., anxiety disorders or depressive disorders) that cause “clinically significant distress or impairment in social, occupational, or other important areas of functioning” [2], which do not meet the diagnostic criteria for a specific disorder. Exact prevalence estimates for these disorders are unknown. However, based upon the prevalence estimates of other disorders within the classes, inclusion of the following DSM-5 SCID modules may have an impact on SMI and other MI estimates.

• Other Specified Bipolar Disorder: This module includes both other specified Bipolar Disorder and Cyclothymia. Cyclothymia is characterized by chronic (2 or more years among adults), fluctuating mood disturbance involving numerous separate periods of hypomanic symptoms, and periods of depressive symptoms. It has low lifetime prevalence, between 0.4 and 1 percent making it more suitable to be assessed in combination with other specified bipolar disorder.
• Other Specified Depressive Disorder
• Other Specified Eating Disorder
• Other Specified Obsessive-Compulsive and Related Disorder

3.6.1. Impact of DSM Revisions on MHSS Estimates of SMI, AMI and Specific Disorders

The diagnostic revisions between DSM-IV and DSM-5 included both subtle (e.g., MDE adding the word hopeless) and some distinctive (e.g., somatic symptom disorder’s complete reworking) changes. These changes have the potential to affect AMI, SMI, and specific disorder estimates in MHSS. Although many of the disorders have had at least minor changes under DSM-5, in most cases there is very little data available on how the changes will affect prevalence estimates. In most cases where there are data, the prevalence estimates appear to remain unchanged or to have increased only slightly. This would suggest that MHSS DSM-IV–based estimates will underestimate the disorders under DSM-5 criteria and subsequently the AMI/SMI estimation in NSDUH would likely be an underestimate. However, the data available for these changes has primarily come from treatment-based samples and may not be generalizable to the nationally representative sample in MHSS.

The changes from DSM-IV to DSM-5 may also lead to reevaluating which disorders are currently assessed by the MHSS. New disorders would need to be considered for inclusion based on a number of factors, including prevalence, clinical severity, assessment length, and comorbidity with other disorders, all of which are relevant to accurate estimation of AMI and SMI. Changes to existing disorders might result in changes in prevalence or severity and may also prompt a reevaluation of their inclusion in MHSS. Based on these considerations, this report explored a number of potential inclusions for MHSS. Table 3.40 summarizes the benefits and drawbacks for each disorder.

Table 3.40

Summary of Options for Additional DSM-5 SCID Modules for the Future MHSS.

If additional disorders were to be added to the MHSS, ADHD, binge eating disorder, and conduct disorder may be the disorders with the highest potential return given the respondent burden. In addition, ASD, bipolar II, and body dysmorphic disorder may merit inclusion based on prevalence and/or clinical impairment. However, the addition of all of these disorders concurrently could significantly lengthen the assessment for individuals with high comorbidity; therefore, priority should be given to disorders of primary interest to NSDUH stakeholders and which would provide the greatest addition to the accuracy of AMI and SMI estimates in the population.

3.6.2. Impact on Estimates of Any Mental Illness/Serious Mental Illness

The model used to predict AMI and SMI status for the adult NSDUH sample are based, in part, on responses to the K6 scale and the WHODAS. The K6 scale has been validated against DSM-IV diagnostic criteria. However, assuming that the same variables would continue to be optimal to predict DSM-5-based SMI status, the K6 has not yet been validated against the DSM-5 diagnostic criteria. Therefore, either the K6 would need to be validated or another general mental distress scale that has been validated against DSM-5 would need to be identified. In addition to the impact to the NSDUH main interview because of a break in trends and contexts effects, it is unclear whether a measure of general mental distress that has been validated against DSM-5 exists. Further exploration in this area (with potentially a pilot study) would be needed.

The removal of the GAF score from DSM-5 brings into question how functional impairment should be measured in MHSS. NSDUH could use the WHODAS 2.0, as described in DSM-5. This would provide a validated measure of disability for the clinical interview. However, a version of the WHODAS is also used to measure disability in the NSDUH interview. This may improve the precision of AMI and SMI estimation. However, such changes should be evaluated carefully. Other measures of disability, including retaining the GAF, could also be considered. However, the GAF has demonstrated issues with interrater reliability.

3.6.3. Impact on NSDUH Major Depressive Episode Estimates

In addition to being assessed in the MHSS, MDE is also assessed as part of the NSDUH interview. Although the diagnostic changes from DSM-IV to DSM-5 regarding mixed episodes and the bereavement exclusion will have no effect on NSDUH estimates of MDE, the addition of the term “hopeless” as a subjective descriptor of low mood (Criterion A1) may have an effect on these estimates. NSDUH could revise the MDE question about low mood in the adult depression module to include the word “hopeless.” This would result in the questionnaire more accurately reflecting the revised wording of this symptom in DSM-5 (the adolescent depression module already contains the term “hopeless”). However, based on Uher’s report that only 8 percent of a treatment sample reported feeling discouraged/hopeless but not feeling sad [64], this change is anticipated to have a minimal effect on prevalence rates for the general population.