Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Feb 28.
Published in final edited form as: J Nerv Ment Dis. 2006 Feb;194(2):91–97. doi: 10.1097/01.nmd.0000198140.02154.32

Characteristics and Predictors of Full and Partial Recovery From Generalized Anxiety Disorder in Primary Care Patients

Benjamin F Rodriguez *, Risa B Weisberg †, Maria E Pagano †, Steven E Bruce †, Michael A Spencer ‡, Larry Culpepper §, Martin B Keller †
PMCID: PMC3289410  NIHMSID: NIHMS355186  PMID: 16477186

Abstract

The current study examined the naturalistic course of generalized anxiety disorder (GAD) in a sample of 113 primary care patients across a 2-year period. Initial diagnoses were established using structured clinical interviews according to DSM-IV diagnostic criteria. Results indicated that the majority of patients meeting DSM-IV diagnostic criteria for GAD were still symptomatic to some degree after 2 years of follow-up. Rates of full and partial recovery from GAD, however, were found to be higher than those reported for previous studies of GAD in psychiatric patients. Diagnostic comorbidity, severity of psychosocial impairment, and gender were found to be significantly associated with achieving full or partial recovery from GAD. Psychiatric treatment was not found to be associated with time to full or partial recovery from GAD symptoms, likely due to a treatment-biasing effect. These results underscore that GAD is a chronic and persistent illness in primary care patients.

Keywords: GAD, psychiatric treatment, GAD course, anxiety, primary care

Generalized anxiety disorder (GAD) is a highly prevalent mental illness with 1-year and lifetime prevalence rates in the general population at 3.1% and 5.1%, respectively (Kessler et al., 1994). As many researchers have noted (e.g., Culpepper, 2002), GAD represents a particular problem in the primary care field, where GAD patients are frequently encountered but often unrecognized (Fifer et al., 1994; Niesenson et al., 1998; Wittchen et al., 2002). Anxiety disorder patients in general, and GAD patients in particular, are also among the highest utilizers of medical services (Greenberg et al., 1999; Katon et al., 1990; Rice and Miller, 1998; Roy-Byrne and Katon, 1997). Thus, studying GAD specifically within the primary care population is particularly timely and relevant.

Despite the prevalence of GAD, few studies have examined the illness’s natural course. Most of the data currently available on the course of GAD are retrospective and have come from epidemiological studies (e.g., Kessler et al., 1999; Wittchen et al., 1994) and/or from short-term follow-ups of patients involved in treatment studies (e.g., Mancusco et al., 1993; Rickels and Schweizer, 1993). One of the only naturalistic, prospective studies of the course of GAD conducted to date found low rates of recovery from GAD among psychiatric outpatients, with high rates of subsequent recurrence among those who do recover (Yonkers et al., 1996, 2000). This same study identified better psychosocial functioning in general, and in specific functioning domains (e.g., relationships with spouse, and life satisfaction, etc.) as significant predictors of GAD recovery.

While the results of the studies by Yonkers et al. (1996, 2000) provide important insights into the naturalistic course of GAD, several aspects of that study’s sample need to be considered when interpreting the course findings. First, recruiting the GAD sample from psychiatric treatment settings may represent a significant departure from the typical GAD patient who is likely untreated or receives treatment in a nonpsychiatric, general medical setting (Wittchen et al., 1994). All subjects in the sample by Yonkers et al. were also necessarily engaged in psychiatric treatment of some kind, at least at study intake. Even though this treatment was not systematic or controlled as it would be in a randomized clinical trial, it is possible that being in treatment affected the course of GAD, relative to untreated GAD patients. Patients who seek and/or are engaged in psychiatric treatment often have more severe symptoms and impairment than patients not engaged in treatment (i.e., Berkson’s bias; Berkson, 1946) and may have suffered from symptoms for longer periods of time than untreated patients. Because of these factors, additional studies of the course of GAD in nonpsychiatric or even untreated samples are necessary.

Building on previous research, the present study will examine clinical and demographic characteristics that may be predictive of the 2-year course of GAD in a sample of primary care patients. GAD is a chronic condition in which patients have been found to experience symptoms and impairment over periods of years and even decades (Kessler and Wittchen, 2002; Yonkers et al., 2000). However, GAD patients do not necessarily experience the same severity of symptoms or impairment throughout the course of their illness (Fifer et al., 1994; Maier et al., 2000; Yonkers et al., 1996). Rather, evidence indicates that the typical GAD patient, while experiencing a chronic course, shows a fluctuation in symptom severity during that course. Further, the results of controlled treatment studies indicate that GAD patients rarely experience full or sustained recovery from symptoms (Brown et al., 1996; Durham et al., 1997; Yonkers et al., 2003). Therefore, the current study will examine both full and partial recovery from GAD symptoms as separate clinical outcomes.

METHODS

The Primary Care Anxiety Project (PCAP) is an ongoing longitudinal study of the clinical course and outcomes of patients with anxiety disorders, with the main inclusion criteria being that patients had a general medical appointment on the day of recruitment and were determined to have an anxiety disorder. PCAP enrolled 539 patients with a recognized or unrecognized anxiety disorder originally identified by screening patients visiting their primary care providers. A spectrum of urban and rural; small and large group; and academic and nonacademic primary care, family medicine, and internal medicine sites participated. Descriptions of participant recruitment and screening procedures can be found in other published reports (e.g., Rodriguez et al., 2004; Weisberg et al., 2002). PCAP was approved by the institutional review board of Brown University and participating hospitals. All participants provided written informed consent prior to enrollment in the study. Once enrolled, participants were contacted for an in-person or telephone follow-up interview at 6 and 12 months, and annually thereafter. This report focuses on the first 2 years of longitudinal observation.

Participants

The PCAP sample included 135 primary care patients diagnosed with GAD based on DSM-IV (American Psychiatric Association, 1994) diagnostic criteria established by diagnostic interview employing the Structured Clinical Interview for the DSM-IV (SCID-IV; First et al., 1996). This report focuses on a subsample of 113 of those GAD participants who had at least a full 24 months of follow-up data postintake. There were no significant demographic or clinical differences between patients without a full 24 months of follow-up data and those with full data who were included in this report.

Assessment Measures

Intake Interview

All clinical diagnoses were established by means of diagnostic interviews that employed the SCID-IV (First et al., 1996). In PCAP, the psychotic screen, mood, anxiety, substance use, and eating disorders modules of the SCID-IV were administered.

Follow-Up Assessments

The Longitudinal Interval Follow-Up Evaluation (LIFE; Keller et al., 1987) is an interviewer-administered assessment that collects detailed information on anxiety disorder symptoms, psychosocial functioning, and treatment status. The LIFE employs a 6-point psychiatric status rating (PSR) scale to indicate the severity of psychiatric pathology. A PSR of 5 or 6 indicates the participant meets full DSM-IV diagnostic criteria for the given disorder with low to moderate and severe functional impairment, respectively (i.e., in episode). A PSR of 3 or 4 indicates the participant does not meet full DSM-IV diagnostic criteria for the disorder, but still exhibits notable residual symptoms and impairment to a mild or moderate degree, respectively (i.e., partial recovery). A PSR of 1 or 2 indicates the participant is either completely without symptoms of the disorder, or experiences a negligible number of symptoms on an occasional and transient basis (i.e., full recovery).

In this study, a period of full recovery was defined as minimal or no symptoms of GAD (PSR of 1 or 2) for at least 8 consecutive weeks. Partial recovery was defined as 8 or more consecutive weeks at less than full criteria for GAD but with significant residual symptoms and impairment, defined as PSR 4 or less for GAD, without a period of 8 consecutive weeks at PSR 1 or 2. Patients were considered unrecovered for GAD if they failed to experience either a full or partial recovery from GAD during the 2-year period of observation.

The LIFE employs a change point method to anchor participant reports of symptom levels to relevant life events such as birthdays, holidays, family vacations, etc., resulting in weekly ratings of psychiatric symptom severity. Three substudies have been conducted analyzing the reliability and validity of the LIFE and are described in Warshaw et al., (1994). These studies found interrater reliability and long-term test-retest reliability for the LIFE diagnostic ratings to be good to excellent for all anxiety disorders and major depressive disorder (MDD).

The LIFE also assesses participants’ psychosocial functioning in several domains including employment, housework, student work, interpersonal relationships with spouse, children, parents, siblings, and friends, recreation, life satisfaction, and overall social adjustment (as determined by the interviewer). The Longitudinal Interval Follow-up Evaluation—Range of Impaired Functioning Tool (LIFE-RIFT; Leon et al., 1999; 2000) was used to assess for psychosocial functioning. The LIFE-RIFT is a brief composite scale assessing psychosocial impairment, derived from the psychosocial domains assessed by the LIFE. Ratings of participant functioning in the work, recreation, interpersonal relationships, and satisfaction domains are summed to yield a single score ranging from 4 (no impairment, high functioning) to 20 (severe impairment). In the case of work, the area of worst impairment from the employment, student work, and housework domains is used in the LIFE-RIFT calculation. For relationship functioning, the rating for the worst relationship category is included in the LIFE-RIFT calculation. The psychometric properties of the LIFE-RIFT were evaluated and the scale was found to have good reliability, concurrent validity, and predictive validity (Leon et al., 1999; 2000).

Mental Health Treatment

Information on mental health treatment was gathered in two ways. First, information on psychotropic medication usage was obtained using the Psychotropic/Auxiliary Drug Treatment Schedule, administered as part of the LIFE interview. Participants are asked what psychotropic medications they are currently taking and the dosages, with the data and dosage recorded on a weekly basis. At intake, the data on medications and dosage was gathered both currently and retrospectively for 3 months prior to study intake. Second, current psychosocial treatment was assessed using a Types of Mental Health Treatment Received form developed for PCAP. The form asks participants if they were currently receiving treatment in specific psychotherapeutic modalities, including individual therapy, group therapy, family/couples therapy, self-help groups, day treatment, inpatient hospital treatment, residential treatment, or medication management. Data were gathered both currently and for the 6 months prior to intake. Both the Psychotropic/Auxiliary Drug Treatment Schedule and Types of Mental Health Treatment Received form are also administered at each follow-up interview to assess patient treatment during the intervening time period.

Statistical Analyses

Course data were examined using standard survival analysis methods (Kalbfleish and Prentice, 1980). Cox regressions (Cox, 1972) were used to test for various predictors of time to improvement from GAD (i.e., full recovery or partial recovery) during the 2-year follow-up period. The following intake variables were among the clinical course predictors examined: age, age at GAD onset, current MDD, number of current comorbid anxiety disorders, treatment participation, overall psychosocial impairment (LIFE-RIFT), and gender. Additional Cox regressions were conducted to explore whether the selected predictors were related to time to full recovery from GAD relative to partial recovery as well as relative to unrecovered GAD.

Analyses were conducted using SAS version 8.2 (SAS Institute, Inc., 1999) PROC FREQ, PROC MEANS, PROC ANOVA, PROC PHREG, and PROC NPAR1WAY.

RESULTS

Sample Characteristics

The majority (76%, N = 86) of the present sample was female, with a mean age of 39.2 years (SD = 12.01); 60% (N = 68) were married or cohabitating as if married, 16% (N = 18) were divorced, separated, or widowed, and 24% (N = 27) were never married. Participants were fairly well educated, with 96% (N = 108) of the sample reporting at least a high school degree or GED and 31% (N = 35) having a 4-year college degree or better. The sample was 86% (N = 97) Caucasian, 6% (N = 7) African American, 3% (N = 3) Hispanic, 2% (N = 2) Asian, and 3% (N = 4) other ethnic groups.

Regarding the diagnostic status of the current GAD sample, 12 (11%) also had comorbid panic disorder, 24 (21%) had comorbid panic disorder with agoraphobia, 1 (1%) had comorbid agoraphobia without history of panic disorder, 14 (12%) had comorbid posttraumatic stress disorder, and 23 (20%) had comorbid social anxiety disorder. MDD was a comorbid condition in 47 (42%) of GAD participants at the time of intake.

Longitudinal Course of GAD

A standard survival analysis employing Kaplan-Meier life tables revealed a 0.39 probability of full recovery from GAD and a 0.54 probability of partial recovery from GAD. Figure 1 depicts probability of full or partial recovery from GAD across the first 2 years of observation.

FIGURE 1.

FIGURE 1

Open in a new tab

Cumulative probability of recovery for generalized anxiety disorder.

Of the 44 GAD patients who achieved a full recovery during the first 2 years of follow-up, 10 (22%) experienced a partial recurrence of GAD symptoms (defined as at least 4 weeks at PSR 3 or 4 for GAD), while 13 (30%) had a full recurrence of GAD (defined as at least 4 weeks at PSR 5 or 6 for GAD). Only 21 (48%) did not experience a return of more than minimal GAD symptoms during the time period under observation.

Demographic and Clinical Characteristics of GAD Outcome Groups

Table 1 presents the intake characteristics of the full recovery, partial recovery, and unrecovered GAD participants. There were significant differences between the three groups with regard to the amount of time in episode with GAD prior to study intake (F[2,110] = 7.27; p = 0.001), where fully recovered patients experienced GAD for less time prior to entering the study than both partially recovered and unrecovered patients. Patients who achieved a full recovery from GAD also tended to be older at the time of GAD episode onset than both the partially recovered and not recovered patients (F[2,110] = 5.52; p = 0.005); however, the groups did not differ in terms of overall age at study intake (F[2,110] = 1.12; p = 0.33). There was a nonsignificant trend for not recovered patients to have more comorbid current anxiety disorders than either full recovery or partial recovery patients (F[2,102] = 2.85; p = 0.06), and a significantly smaller proportion of partial recovery patients had comorbid MDD (χ2 = 11.94; p = 0.003). There were no significant differences in current diagnostic comorbidity between full and partially recovered participants. A smaller proportion of the GAD patients who achieved a partial recovery were engaged in any type of treatment (χ2 = 6.16; p = 0.05) at intake or were receiving psychopharmacotherapy (χ2 = 6.47; p = 0.04) as compared with full recovery and unrecovered patients. There was a nonsignificant trend for more participants in the not recovered sample to be engaged in some form of psychotherapy (χ2 = 5.39; p = 0.07) as compared with individuals in either recovery group. The unrecovered sample also was significantly more likely to be receiving both psychotherapy and pharmacotherapy in combination (χ2 = 7.87; p = .02) than were the recovered groups. Finally, unrecovered GAD patients had significantly more psychosocial impairment at intake than those patients who had a full or partial recovery from GAD (F[2,102] = 8.17; p = 0.0005).

TABLE 1.

Intake Clinical and Demographic Characteristics of GAD Patients Who Achieved Full or Partial Recovery or Were Unrecovered After 2 Years of Follow-Up

Full Recovery N = 44 Partial Recovery N = 37 Unrecovered N = 32
Age of GAD onset 28.70 (15.93) 20.68 (14.16) 18.09 (13.66)
Weeks in episode prior to intake 67.05 (91.37) 142.26 (112.06) 146.48 (113.81)
Diagnostic comorbidity
 MDD 50% 19% 56%
 Current anxiety disorders, N .82 (1.04) .78 (1.00) 1.31 (1.03)
Treatment status
 Any treatment 45% 30% 59%
 Any medication 39% 30% 59%
 Any psychotherapy 16% 19% 38%
 Combined treatment 9% 11% 31%
Psychosocial functioning LIFE-RIFT 10.23 (3.15) 9.95 (3.10) 12.74 (3.04)
Open in a new tab

Medication Treatment Characteristics

To explore potential treatment participation differences among the three groups during the follow-up period, we examined how many patients in each group reported taking a selective serotonin reuptake inhibitor (SSRI) prior to their recovery from GAD. SSRIs were selected because this class of medications is frequently prescribed to anxiety and depression patients and has some empirical validation of efficacy in the treatment of GAD. Among those patients who experienced a full recovery from GAD, 14 (32%) had been taking an SSRI prior to their recovery from GAD. Additionally, five (14%) of those patients who achieved a partial recovery from GAD were so treated prior to their change in symptom status. Fifteen (50%) of the unremitted GAD patients were taking an SSRI during the time under observation. The difference in SSRI usage among fully recovered, partially recovered, and unrecovered GAD patients was statistically significant (χ2 = 10.71; p = 0.005).

The differential usage of benzodiazepines across full recovery, partial recovery, and unremitted GAD patients was also examined. Similar to the SSRI analysis, patients were considered if they were taking a benzodiazepine (regularly or PRN) of any dosage around the time of full or partial recovery from GAD. Among those patients who experienced a full recovery from GAD, six (14%) had been taking at a benzodiazepine prior to their recovery from GAD. Additionally, three (8%) of those patients who achieved a partial recovery from GAD were so treated immediately prior to their change in symptom status. Finally, eight (25%) of the unremitted GAD patients had been taking a benzodiazepine during the time under observation. There was no significant difference in benzodiazepine usage among fully remitted, partially remitted, and unremitted GAD patients (χ2 = 3.94; p = 0.14).

Predictors of GAD Course

A series of proportional hazards (Cox) regressions was conducted to examine which clinical characteristics at intake were predictive of achieving subsyndromal symptom status and recovery. Since all participants in this study met full criteria for GAD at intake, individuals with episodes of GAD of a long-term chronic nature may have been overrepresented in the sample. To adjust for this potential bias, age of GAD onset was included in the analyses. In addition, current MDD, number of current anxiety disorders, treatment with an SSRI, psychosocial impairment assessed by the LIFE-RIFT at intake, and gender were examined (Table 2). Greater psychosocial impairment at enrollment predicted a reduced likelihood of achieving subsyndromal symptom status (i.e., either full recovery or partial recovery) from GAD. Examination of the hazard ratio suggests that each 1-point increase in overall psychosocial impairment resulted in an 8% reduction in the likelihood of GAD patients experiencing a period of subsyndromal symptom status. In addition, there was a nonsignificant trend for men to have an increased likelihood of achieving subsyndromal symptom status.

TABLE 2.

Proportional Hazards Regression Analyses Examining Predictors of Full and Partial Recovery From GADa

Predictors Wald χ2 P Value B (SE) χ2 P Value Hazard Ratio 95% CI
Full or rartial recovery 17.17 0.009
 SSRI .21 (.27) .58 .45 1.22 .72–2.09
 MDD −.37 (.26) 2.09 .15 .69 .42–1.14
 Anxiety DO −.13 (.13) .98 .32 .88 .68–1.14
 LIFE-RIFT −.09 (.04) 5.25 .02 .92 .85–.99
 Age at onset .01 (.01) 1.44 .22 1.01 1.00–1.02
 Gender .47 (.26) 3.20 .07 1.60 .96–2.68
Full recovery 16.97 0.009
 SSRI .45 (.35) 1.65 .20 1.57 .79–3.12
 MDD .53 (.33) 2.58 .11 1.69 .89–3.21
 Anxiety DO −.06 (.18) .13 .72 .94 .67–1.33
 LIFE-RIFT −.11 (.05) 4.12 .04 .90 .81–.99
 Age at onset .03 (.01) 7.50 .01 1.03 1.01–1.05
 Gender .22 (.36) .38 .54 1.25 .62–2.50
Partial recovery 19.55 0.003
 SSRI .23 (.44) .29 .59 1.26 .54–2.96
 MDD −1.24 (.48) 6.69 .01 .29 .11–.74
 Anxiety DO −.16 (.20) .65 .42 .85 .57–1.27
 LIFE-RIFT −.09 (.06) 2.31 .13 .92 .82–1.03
 Age at onset .01 (.01) .54 .46 1.01 .99–1.03
 Gender .88 (.42) 4.36 .04 2.40 1.06–5.47
Open in a new tab

Note: Anxiety DO, Number of comorbid anxiety disorders; Age at onset, age at time of onset of GAD episode.

a

For all analyses, df = 6.

Two additional Cox regressions examined predictors of achieving full and partial recovery from GAD as separate clinical outcomes. These analyses employed the same predictors described. Both an older age of GAD onset and less severe psychosocial impairment were significantly predictive of the likelihood of achieving a full recovery from GAD. In the second regression, both major depression at intake and female gender were significantly predictive of reduced likelihood of partial recovery from GAD.

DISCUSSION

Our results indicated that over a 2-year period, many primary care patients with full criteria GAD will experience periods of symptom recovery at a rate higher than reported in studies of patients enrolled through psychiatric settings. However, well over half of primary care patients with GAD continued to experience significant symptoms of the disorder after 2 years of prospective follow-up. Further, a substantial proportion of those who did experience a period of full recovery from GAD had a recurrence of symptoms just within the 2-year period of observation. Diagnostic comorbidity, severity of psychosocial impairment, and female sex were all associated with lack of recovery. These findings add to the mounting evidence that GAD is a chronic and recurrent anxiety disorder (Yonkers et al., 2000, 2003).

It is notable that the 2-year probabilities of both full and partial recovery from GAD found in PCAP’s primary care population are much larger than those reported for GAD patients in psychiatric populations (e.g., Yonkers et al., 1996). Indeed, the primary care patients in our study were more likely to achieve both a full or partial recovery from GAD at 2 years than were psychiatric patients at 5 years (e.g., Yonkers et al., 2000). This suggests that the type of GAD patient encountered in primary care settings may be a less severe version of the GAD patient encountered and treated in a more traditional psychiatric treatment setting. However, it must also be noted that the studies by Yonkers et al. employed DSM-III-R diagnostic criteria for GAD with the psychiatric patients in their sample. Thus, it is possible that the differences in recovery rates may be in part attributable to changes in GAD diagnostic criteria between DSM-III-R and DSM-IV. Therefore, these potential differences in full and partial recovery rates between primary care patients with GAD and psychiatric patients with GAD must be viewed as tentative pending future studies of GAD in psychiatric samples employing DSM-IV diagnostic criteria.

Diagnostic comorbidity was among the few factors related to achieving full or partial recovery from GAD in primary care patients. Patients who did not recover from GAD were more likely to have comorbid MDD and have significantly more comorbid anxiety disorders at intake than patients who achieved full or partial recovery. MDD was also found to predict significantly the likelihood of partial recovery from GAD. Results from prior studies examining the influence of comorbidity on GAD symptom recovery have been inconsistent. Some studies of GAD course have reported associations between comorbidity and recovery (e.g., Durham et al., 1997; Mancuso et al., 1993), while others have not (e.g., Yonkers et al., 2000). Although differences in sample characteristics and may in part explain the inconsistencies in results across studies, more research is needed to clarify the complex relationship between Axis I psychiatric comorbidity and recovery from GAD.

At intake, individuals showing more severe psychosocial impairment as assessed by the LIFE-RIFT were less likely to achieve full or partial recovery from GAD. These results are consistent with findings of prior studies showing that GAD patients with more impairment in interpersonal relationships and life satisfaction predicted a reduced likelihood of full or partial recovery (e.g., Yonkers et al., 2000). Moreover, the findings are in line with other results showing that worsening impairment in psychosocial functioning domains is related to likelihood of mood and anxiety disorder recurrence (Leon et al., 1999, 2000; Rodriguez et al., 2005). These results highlight the potential role that psychosocial impairment may play in GAD course and should be of particular note to clinicians and treatment researchers. Most treatment strategies, cognitive-behavioral or otherwise, make the reduction of symptom severity the central focus of treatment as well as the primary measure of outcome. Our findings suggest that such symptom-focused approaches ultimately may not be sufficient and that interventions that specifically try to improve psychosocial functioning in addition to reducing symptoms may be more effective strategies for treating GAD and other clinical syndromes.

A notable finding of the current study was that men were significantly more likely than women to achieve a partial recovery from GAD. There was also a nonsignificant trend for men with GAD to experience a period of time at less than full diagnostic criteria for GAD, in general (i.e., either full or partial recovery). In a comparable short-interval follow-up study using psychiatric outpatients with DSM-III-R GAD, there was a nonsignificant trend for men to be more likely to recover from GAD after a period of 8 years (Yonkers et al., 2003). However, there was also a significantly higher recurrence rate among men in the same study. In total, these data suggest that GAD may be a different clinical phenomenon when present in men versus women, but additional studies are needed to clarify the nature of any gender differences in the manifestation of GAD.

Even though some significant differences were found in the rates and types of treatment received by the GAD patients who achieved full recovery or partial recovery or remained symptomatic across 2 years of follow-up, overall, being in treatment was not found to be a significant predictor of time to recovery from GAD. While it might be interpreted from these findings that psychiatric treatment ultimately bears no relationship to GAD recovery, the results need to be taken in appropriate context. A common finding in naturalistic, longitudinal research is a treatment bias effect, wherein patients with episodes of greater severity and duration are more likely to be enrolled and receive psychiatric treatment. Since severity of illness and impairment are inversely associated with likelihood of recovery, these same patients are less likely to recover than the less severe, untreated patients. Consequently, the effects of treatment are typically washed out, as may have been the case in the current study.

The present study had several limitations. First, the sample is mostly Caucasian and female. As a result, questions may arise as to how well the findings would generalize to a more gender-diverse and ethnically diverse sample. In addition, both the patient symptomatology data and the data on the type and dosage of psychotropic medication rely on retrospective patient self-report. Although the measures employed to collect these data have been shown to be reliable and well-validated instruments (Keller et al., 1987; Warshaw et al., 1994), additional data from more objective sources would add to the strength of our findings. Finally, even though all patients examined had 2 years to achieve a full or partial recovery, they did not have an equal amounts of time after symptom recovery potentially to have a recurrence. Thus, the data on proportion of recovered patients whose symptoms recur are likely to be an underestimate of the number of GAD patients who will experience a recurrence.

Acknowledgments

The PCAP is supported by an unrestricted grant from Pfizer Pharmaceuticals, Inc.

Footnotes

Drs. Risa B. Weisberg, Martin B. Keller, and Larry Culpepper have disclosed financial relationships with pharmaceutical companies. Detailed disclosure information is available from The Journal of Nervous and Mental Disease.

Portions of this paper were presented at the Annual Meetings of the Anxiety Disorders Association of America, Toronto, Ontario, Canada (March 2003), and Seattle, Washington (March 2005).

References

  1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4. Washington DC: American Psychiatric Association; 1994. [Google Scholar]
  2. Berkson J. Limitations of the application of fourfold table analysis to hospital data. Biometrics. 1946;2:47–53. [PubMed] [Google Scholar]
  3. Brown C, Schulberg HC, Madonia MJ, Shear MK, Houck PR. Treatment outcomes for primary care patients with major depression and lifetime anxiety disorders. Am J Psychiatry. 1996;153:1293–1300. doi: 10.1176/ajp.153.10.1293. [DOI] [PubMed] [Google Scholar]
  4. Cox DR. Regression models and life-tables. J R Stat Soc. 1972;34:187–220. [Google Scholar]
  5. Culpepper L. Generalized anxiety disorder in primary care: emerging issues in management and treatment. J Clin Psychiatry. 2002;63(suppl 8):35–42. [PubMed] [Google Scholar]
  6. Durham RC, Allan T, Hackett CA. On predicting improvement and relapse in generalized anxiety disorder following psychotherapy. Br J Clin Psychiatry. 1997;36:101–119. doi: 10.1111/j.2044-8260.1997.tb01234.x. [DOI] [PubMed] [Google Scholar]
  7. Fifer SK, Mathias SD, Patrick DL, Mazonson PD, Lubeck DP, Buesching DP. Untreated anxiety among adult primary care patients in a health maintenance organization. Arch Gen Psychiatry. 1994;51:740–750. doi: 10.1001/archpsyc.1994.03950090072010. [DOI] [PubMed] [Google Scholar]
  8. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for the DSM-IV Axis I Disorders. Biometrics Research Department, New York State Psychiatric Institute; New York, NY: 1996. [Google Scholar]
  9. Greenberg PE, Sisitsky T, Kessler RC, Finkelstein SN, Berndt ER, Davidson JR, Ballenger JC, Fyer AJ. The economic burden of anxiety disorders in the 1990’s. J Clin Psychiatry. 1999;60:427–735. doi: 10.4088/jcp.v60n0702. [DOI] [PubMed] [Google Scholar]
  10. Kalbfleish JG, Prentice RL. The Statistical Analysis of Failure Time Data. New York: Wiley; 1980. [Google Scholar]
  11. Katon W, Von Korff M, Lin E, Lipscomb R, Russo J, Wagner E, Polk E. Distressed high utilizers of medical care: DSM-III-R diagnoses and treatment needs. Gen Hosp Psychiatry. 1990;12:355–362. doi: 10.1016/0163-8343(90)90002-t. [DOI] [PubMed] [Google Scholar]
  12. Keller MB, Lavori PW, Friedman B, Nielsen E, Endicott J, McDonald-Scott NC, Andreason NC. The longitudinal interval follow-up evaluation: A comprehensive method for assessing outcome in prospective longitudinal studies. Arch Gen Psychiatry. 1987;44:540–548. doi: 10.1001/archpsyc.1987.01800180050009. [DOI] [PubMed] [Google Scholar]
  13. Kessler RC, DuPont RL, Berglund P, Wittchen HU. Impairment in pure and comorbid generalized anxiety disorder and major depression at 12 months in two national surveys. Am J Psychiatry. 1999;156:1915–1923. doi: 10.1176/ajp.156.12.1915. [DOI] [PubMed] [Google Scholar]
  14. Kessler RC, Wittchen HU. Patterns and correlates of generalized anxiety disorder in community samples. J Clin Psychiatry. 2002;63(suppl 8):4–10. [PubMed] [Google Scholar]
  15. Kessler RC, McGonagle KA, Zhao S, Nelson CB, Hughes M, Eshleman S, Wittchen HU, Kendler KS. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: Results from the National Comorbidity Survey. Arch Gen Psychiatry. 1994;51:8–19. doi: 10.1001/archpsyc.1994.03950010008002. [DOI] [PubMed] [Google Scholar]
  16. Leon A, Solomon DA, Mueller TI, Endicott J, Posternak M, Judd LL, Schettler PJ, Akiskal HS, Keller MB. The Range of Impaired Functioning Tool (LIFE-RIFT): A brief measure of functional impairment. Psychol Med. 1999;29:869–878. doi: 10.1017/s0033291799008570. [DOI] [PubMed] [Google Scholar]
  17. Leon A, Solomon DA, Mueller TI, Endicott J, Posternak M, Judd LL, Schettler PJ, Akiskal HS, Keller MB. A brief assessment of psychosocial functioning of subjects with bipolar I disorder: The LIFE-RIFT. J Nerv Ment Dis. 2000;188:805–812. doi: 10.1097/00005053-200012000-00003. [DOI] [PubMed] [Google Scholar]
  18. Maier W, Gaensicke M, Freyberger HJ, Linz M, Heun R, Lecrubier Y. Generalized anxiety disorder (ICD-10) in primary care from a cross cultural perspective: A valid diagnostic entity. Acta Psychiatr Scand. 2000;101:29–36. doi: 10.1034/j.1600-0447.2000.101001029.x. [DOI] [PubMed] [Google Scholar]
  19. Mancuso DM, Townsend MH, Mercante DE. Long-term follow-up of generalized anxiety disorder. Compr Psychiatry. 1993;34:441–446. doi: 10.1016/0010-440x(93)90072-c. [DOI] [PubMed] [Google Scholar]
  20. Nisenson LG, Pepper CM, Schwenk TL, Coyne JC. The nature and prevalence of anxiety disorders in primary care. Gen Hosp Psychiatry. 1998;20:21–28. doi: 10.1016/s0163-8343(97)00096-0. [DOI] [PubMed] [Google Scholar]
  21. Rice DP, Miller LS. Health economics and cost implications of anxiety and other mental disorders in the United States. Br J Psychiatry. 1998;173(suppl 34):4–9. [PubMed] [Google Scholar]
  22. Rickels K, Schweizer E. The treatment of generalized anxiety disorder in patients with depressive symptomatology. J Clin Psychiatry. 1993;54(suppl 1):20–23. [PubMed] [Google Scholar]
  23. Rodriguez BF, Bruce SE, Pagano ME, Keller MB. Relationship between psychosocial functioning and recurrence in generalized anxiety disorder, panic disorder and major depression. J Anxiety Disord. 2005;19:752–766. doi: 10.1016/j.janxdis.2004.10.002. [DOI] [PubMed] [Google Scholar]
  24. Rodriguez BF, Weisberg RB, Pagano ME, Machan JT, Culpepper L, Keller MB. Frequency and patterns if diagnostic comorbidity in a sample of primary care patients with anxiety disorders. Compr Psychiatry. 2004;45:129–137. doi: 10.1016/j.comppsych.2003.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Roy-Byrne PP, Katon W. Generalized anxiety disorder in primary care: The precursor/modifier pathway to increased health care utilization. J Clin Psychiatry. 1997;58:34–38. [PubMed] [Google Scholar]
  26. SAS Institute, Inc. SAS Procedure Guide (version 8.2) Cary (NC): SAS Institute, Inc; 1999. [Google Scholar]
  27. Warshaw MG, Keller MB, Stout RL. Reliability and validity of the longitudinal interval follow-up evaluation for assessing outcome of anxiety disorders. J Psychiatr Res. 1994;28:531–545. doi: 10.1016/0022-3956(94)90043-4. [DOI] [PubMed] [Google Scholar]
  28. Weisberg RB, Bruce SE, Machan JT, Kessler RC, Culpepper L, Keller MB. Nonpsychiatric medical illness among primary care patients with trauma histories and posttraumatic stress disorder. Psychiatr Serv. 2002;53:848–854. doi: 10.1176/appi.ps.53.7.848. [DOI] [PubMed] [Google Scholar]
  29. Wittchen HU, Kessler RC, Beesdo K, Krause P, Hoefler M, Hoyer J. Generalized anxiety and depression in primary care: Prevalence, recognition and management. J Clin Psychiatry. 2002;63(suppl 8):24–34. [PubMed] [Google Scholar]
  30. Wittchen HU, Zhao S, Kessler RC, Eaton WW. DSM-III-R generalized anxiety disorder in the National Comorbidity Survey. Arch Gen Psychiatry. 1994;51:355–364. doi: 10.1001/archpsyc.1994.03950050015002. [DOI] [PubMed] [Google Scholar]
  31. Yonkers KA, Bruce SE, Dyck IR, Keller MB. Chronicity, relapse, and illness-course of panic disorder, social phobia and generalized anxiety disorder: Findings in men and women from 8 years of follow-up. Depress Anxiety. 2003;17:173–179. doi: 10.1002/da.10106. [DOI] [PubMed] [Google Scholar]
  32. Yonkers KA, Dyck IR, Warshaw M, Keller MB. Factors predicting the clinical course of generalized anxiety disorder. Br J Psychiatry. 2000;176:544–550. doi: 10.1192/bjp.176.6.544. [DOI] [PubMed] [Google Scholar]
  33. Yonkers KA, Warshaw MG, Massion AO, Keller MB. Phenomenology and course of generalized anxiety disorder. Br J Psychiatry. 1996;168:308–313. doi: 10.1192/bjp.168.3.308. [DOI] [PubMed] [Google Scholar]

RESOURCES