|Home | About | Journals | Submit | Contact Us | Français|
Several studies have demonstrated that obsessive-compulsive disorder (OCD) is associated with interference in quality of life (QOL) and functional impairment. However, these studies did not compare individuals in remission to individuals who continue to have the disorder, predominantly used comparisons with norms and not with a matched normal sample, and did not always consider the impact of comorbidity. We administered multiple measures that assess QOL and functional impairment to 66 OCD patients who had previously consented for a clinical trial, and to 36 age and sex matched individuals who denied any psychiatric history. Results confirm that OCD was associated with significantly lower QOL and functional impairment compared to healthy controls in areas of work, social life, and family life. Individuals with OCD and other comorbid psychiatric diagnoses showed the poorest QOL and functioning, with comorbid depression accounting for much of the variance. The levels of QOL and functioning in individuals in remission tended to lay in between healthy controls and individuals with current OCD: their QOL or functioning did not differ significantly from healthy controls, nor did they consistently differ significantly from who had current OCD. These results suggest that individuals who are in remission have improved levels of QOL and functioning while individuals with OCD are significantly impaired, and individuals with OCD and comorbid disorders are the most impaired. Treatment strategies should be focused on achieving remission of all symptoms in order to have the greatest impact on functioning and QOL.
Research suggests that obsessive-compulsive disorder (OCD) is associated with disability, increases in use of health care services, and financial difficulties (e.g., Bobes et al., 2001; Koran, 2000). Therefore, it is important to examine the factors related to functional impairment and quality of life (QOL) in OCD, and to understand whether individuals who are in remission from OCD continue to have impaired functioning or QOL1.
Indeed, despite the growing literature on functional impairment and QOL in OCD, several important questions remain. One important question pertains to the relationship of OCD symptoms to these constructs. For example, Eisen et al. (2006) reported that a Y-BOCS of greater than 20 is associated with significant worsening in a measure of QOL (though this relationship was not found for a measure of psychosocial functioning). Moreover, several clinical trials found that treatment reduces OCD symptoms and improves functioning and/or QOL (e.g., Moritz et al., 2005; Pallanti et al., 2002) However, no studies to date have examined whether OCD patients in remission differ in QOL or functioning from patients with clinically significant OCD symptoms.
Another question is whether decreased functioning and QOL is due to OCD itself or to comorbidity (see Eisen et al., 2006 vs. Quilty et al., 2003; Moritz et al. 2005, and Rodriguez-Salgado et al. 2006 for conflicting evidence). Given the high rates of comorbidity of OCD with other anxiety disorders and depression (Kessler et al., 2005), it would be helpful to examine whether OCD symptoms per se, and which symptom subtypes, are most related to impairment in QOL or functioning.
The present study examined data from OCD patients who had previously enrolled in a double-blind, randomized clinical trial (Foa et al., 2005) and compared their QOL and functioning to that of matched healthy controls. Individuals with OCD were divided into three groups: in remission, current OCD only, and OCD with other comorbid psychiatric disorders. We hypothesized that OCD patients in remission would report similar QOL and functioning to matched healthy controls (HCs), while individuals with OCD would report poorer QOL and functional impairment. Additionally, we predicted that patients with OCD and comorbid psychiatric disorders would report the worst QOL and functional impairment. Finally, we predicted that within individuals with a history of OCD (current or past), increased severity of OCD would be related to decreased QOL and increased functional impairment, even when controlling for depression.
Adult outpatients with OCD who had previously enrolled in a multi-site randomized clinical trial (Foa et al., 2005) were re-contacted on average six years later (mean [SD]=5.7 [2.9]; range= 1 to 9.6) for an assessment that occurred by phone or in person. Of 122 participants, 105 (86%) were located. Thirty-two of 105 (30%) refused (n=25) or could not participate in the follow-up assessment (n=7, i.e., they had died, were overseas, or were hospitalized). Additionally, 7 patients were eliminated from all analyses (due to: meeting current criteria for social phobia (n=2), denying ever having OCD (n=1), or inconsistent diagnostic criteria vs. OCD severity (n=4)). Thus, 66 patients comprised the current sample. In addition, 36 healthy controls (HCs) were included. HCs were matched to the OCD sample on age, sex, and ethnicity. In addition to the clinical assessment, most of the participants received neuropsychological testing (reported in Simpson et al., 2006). Demographic features collected at the time of the original study did not differ between current participants and non-participants (all p values > 0.12). All subjects provided written informed consent following a full description of IRB-approved procedures.
One-day evaluations included an assessment by a trained rater of current Axis I DSM-IV disorders and symptoms using the Structured Clinical Interview for the DSM-IV (SCID-IV; First et al. 1996), the Yale-Brown Obsessive Compulsive Scale (Goodman et al., 1989), the Hamilton Rating Scale for Depression (Hamilton, 1960), and the Hamilton Rating Scale for Anxiety (Hamilton, 1959). The interviewer measures are reliable and valid for measuring OCD symptoms, depression, and anxiety respectively. Subjects also completed a self-report packet, which included psychometrically valid measures of functional impairment and QOL, as well as measures of OCD (OCI-R; Foa et al., 2002; Huppert et al., 2007), and depression (BDI, Beck, et al., 1979). On the basis of this evaluation, all HCs had no history of Axis I disorders or treatment.
Quality of Life Enjoyment and Satisfaction Questionnaire (QLESQ; Endicott et al., 1993). The short-form of this scale contains 16 items that assess QOL and has found to be reliable, valid, and sensitive to change (Kocsis, et al., 1997). We focused on the total percent score, and individual items related to family, work, and social life.
Social Adjustment Scale-Self Report (SAS; Weissman & Bolliwell, 1976) is a psychometrically validated self-report scale with 54 questions (Weissman et al., 1990). We focused on the total score, the average scores for the work and social subscales, and the average of the family members, partners, and children subscales as an overall family subscale.
Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36; Ware et al. 1992) is a reliable and valid self-report scale that measures eight domains including physical and mental dimensions of health-related functioning (McHorney et al., 1993). We focused on two domains that were related to mental health and did not overlap with symptoms: social functioning and role-functioning due to emotional problems (hereafter work).
All statistical procedures were performed with the SPSS, version 15. 0. Four groups were created based on SCID and Y-BOCS responses: 1) healthy controls who reported no history of psychiatric disorders (hereafter HC); 2) patients in remission who did not meet current SCID criteria for OCD (hereafter Hx OCD); 3) patients who met current SCID criteria for OCD only (hereafter current OCD), and 4) individuals who met SCID criteria for OCD and another current Axis I psychiatric disorder (hereafter comorbid OCD). The four groups were compared on demographic and symptom measures. ANOVAs were then conducted for each impairment or QOL domain (family, work, social) with follow-up comparisons conducted using Bonferroni corrections. These were followed with ANCOVAs using depression (BDI) as a covariate. Next, Pearson correlations were conducted examining the relationship between Y-BOCS/OCI-R total scores, HAM-A, and BDI and measures of QOL and functional impairment. Then, partial correlations were calculated using all of the same measures while controlling for depression using the BDI. Correlational analyses used alpha p < .01. Finally, to determine whether the results of Eisen et al. (2006) were replicated in our sample, we used SPSS version 15 to calculate LOESS curves for Y-BOCS and the QLESQ.
The 102 individuals were divided into the following groups: HCs (n = 36), Hx OCD (n = 15), current OCD (n = 32), and comorbid OCD (n = 19). The comorbid group had additional (one or more) current diagnoses of: major depressive disorder (n=9), dysthymia (n=3), social phobia (n=4), generalized anxiety disorder (n=1), panic disorder (n=1), and alcohol dependence (n=1).
There were no significant differences between the groups in sex, age, race, or IQ (see Table 1; all p’s > .50). There also were no significant differences in the medication use in the three OCD groups (Hx OCD = 27%, current OCD = 36%, comorbid OCD = 42%; χ2 (2) = .87, p = .65); no HCs were on medications. Furthermore, there were no differences in QOL and functioning measures between individuals on and not on medications (all p’s >.20). Table 1 also includes comparisons of symptom severity among groups. HCs had the lowest Y-BOCS, OCI-R, BDI, HAM-D, and HAM-A scores. Comorbid OCDs had higher ratings of OCD, anxiety, and depression than any of the other groups, and the current OCD group had higher OCD and depression ratings than the Hx OCD and HC groups. HCs and Hx OCDs did not differ on anxiety, depression, or OCI-R scores, but did differ on Y-BOCS scores.
ANOVAS were conducted to compare the groups on the various QOL and functional impairment measures. Results are reported in Table 2.
Regarding QOL, the QLESQ, HCs reported better QOL than OCDs and comorbid OCDs, and Hx OCD had better QOL than comorbid OCDs. Regarding functioning, there was variability in the findings depending on the measure. The Sheehan Disability Scale showed the same pattern across each subscale (i.e., work, social, family): Hx OCDs did not differ from HCs, but both groups were significantly less impaired than OCDs who, in turn, were less impaired than comorbid OCDs. On the SAS total and social subscale, HCs had better functioning than OCDs and comorbid OCDs, and Hx OCD had better functioning than comorbid OCDs. On the SAS work subscale comorbid OCD significantly differed from all the other groups (which did not significantly differ from each other). On the SAS family, HCs had better functioning than OCDs and comorbid OCDs, and Hx OCD did not significantly differ from any group. On the MOS, comorbid OCDs had significantly worse work and social functioning than the other groups, and the other groups did not differ significantly from each other.
Examining continuous data by combining the three OCD groups (Hx OCD, current OCD, comorbid OCD) suggested that greater severity of OCD was related to worse QOL and greater functional impairment (see Table 3)2. Given recent findings by Eisen et al (2006), we also examined the relationship between Y-BOCS severity and QOL to determine if the relationship was linear by using LOESS curves (see Figure 1). In contrast to Eisen et al., we found a linear relationship best fit the data. When controlling for depression using the BDI (Table 4), many of the relationships between OCD severity and QOL or functional impairment became non-significant, except impairment in work, social life, and family life as measured with the SDS3.
Depression and anxiety were also significantly related to QOL and functional impairment and in most domains (Table 3). These relationships were only slightly changed when controlling for OCD severity by either the Y-BOCS or OCI-R. This suggests that anxiety and depression are significant contributors to problems with QOL and functional impairment in OCD.
Following Masellis, et al., (2003), we also examined the relationship among obsessions and compulsions as measured by the Y-BOCS and QOL and functional impairment. The Y-BOCS Obsession and Compulsion subscale scores correlated .81 with each other, suggesting a high overlap between severity of obsessions and severity of compulsions. Results were extremely similar for both obsessions and compulsions, and reflected the Y-BOCS total results in Table 3 (all results were r = +/-.09 of the correlations with Y-BOCS total). Only 2 correlations yielded somewhat different results. The SAS social subscale was significantly correlated with Y-BOCS compulsions (r (63) = .35, p < .01), but not the obsession subscale (r (63) =.26, p = .04). In contrast, the MOS work functioning subscale was significantly correlated with obsessions (r (63) = -.44, p < .01), but not compulsions (r (63) = -.27, p = .03).
OCD symptom subtypes as measured by the OCI-R showed some different relationships with QOL and functioning. In particular, checking and neutralizing were not significantly correlated with any measure at p <.01. Hoarding was correlated with impairments on SAS total and social (r’s = .48, .39), all SDS scales (r = .40 to .54), MOS work (r = -.48) and social (r = -0.35) and QLESQ total, social and family scales (r’s = -.34 to -.56). Obsessing was correlated with impairment on all SDS scales (r’s = .45 to .49), MOS work (r = -.57) and social (r = -0.50), and QLESQ total and social (r’s =-.41, -.34). Washing was related to more impairment on the SAS total (r = .37), all SDS scales (r’s = .40 to .57), the MOS social (r = -0.39) and QLESQ total, social, and family scales (r’s = -.34 to -.44). Ordering was only related to MOS social (r = - 0.33). After controlling for depression, only the correlations between the washing and SDS social (r = .43) and the obsessing and MOS work (r = -.44) subscales remained significant.
Consistent with previous studies, the present study found that individuals with OCD and comorbid disorders tend to be significantly more impaired than individuals with OCD without comorbidity, both of whom have significantly worse QOL and functioning and than healthy controls. The level of functioning and QOL in individuals in remission tended to fall between that of healthy controls and individuals with OCD only. In terms of severity of QOL and functional impairment, individuals with comorbid OCD were in the moderate to severe range, individuals with OCD only were in the mild to moderate range, and OCD patients in remission were in the very mild range. While QOL and functional impairment were generally related to severity of OCD, the relationship was attenuated when controlling for depression. The simplest scales for measuring functional impairment (the Sheehan Disability Scales) seemed to be quite sensitive to OCD severity and less influenced by depression. These findings extend the literature by including healthy controls and individuals with various levels of OCD symptoms, and therefore provide some evidence for the clinical significance of ranges of scores on various measures of QOL and functional impairment in patients with OCD.
One could interpret the findings regarding the remitted OCD group as either promising (individuals in remission were similar to healthy individuals who never had a psychiatric disorder), or more guardedly (individuals in remission did not differ from patients with OCD except on the SDS). The lack of differences with either group could be due to lack of power or insensitivity of the QOL and functioning measures. Alternatively, the lack of difference from the OCD group could suggest that while improvements in QOL and functioning occur, there are persistent deficits in OCD even after successful treatment. This would not be surprising in a chronic condition such as OCD that is likely to have impacted development in such a way that continues even if symptoms remit (i.e., new skills that others have learned in childhood may have to be developed). This deserves additional study in a larger sample.
Our finding that comorbid patients had the poorest QOL and greatest functional impairment is consistent with the studies showing that comorbidity in general negatively influences most areas of life (e.g., Sanderson & Andrews, 2002). Our findings are also consistent with the reports by Quilty et al. (2003), Rapaport et al., (2005), and Rodriguez-Salgado et al., (2006), though inconsistent with those of Masellis et al. (2003) and Eisen et al. (2006). Specifically, our results suggest that depression accounts for the relationship between either obsessions or compulsions and QOL or functional impairment. We also examined the relationship between symptom subtypes and QOL and functional impairment. Obsessing (e.g., harm thoughts, sexual thoughts, scrupulosity), washing, and hoarding were all more related to QOL than checking, ordering, and neutralizing. However, depression seemed to account for much of the relationship, even in hoarders. More research is needed to examine impairment in subtypes, and its possible interaction with treatment outcome (c.f., Mataix-Cols, et al., 2002, Abramowitz et al., 2003).
Our findings suggested a linear relationship between the QLESQ and OCD severity as well as with measures of functional impairment and OCD. These results are inconsistent Eisen et al., (2006), who found a nonlinear relationship between QLES-Q and YBOCS and suggested a cutscore of 20 on the Y-BOCS on the basis of the nonlinear relationship. The best method of determining a cutscore to determine diagnostic thresholds would be by using ROC analyses after administering the Y-BOCS to a large, unselected population along with a structured diagnostic interview to determine presence or absence of the diagnosis of OCD (see Swets, 1996). Measuring QOL and functioning in the same population could be useful additional information for further analysis.
There are several limitations of this study. First, the sample size for individuals in remission and those with comorbidity was small, and could result in an inability to detect some differences among groups. In addition, QOL and functioning measures were self-report, and there was no independent verification of functioning. Finally, this study used a sample of convenience of individuals who had consented to participate in a clinical trial and evaluated them years after their consent. Thus, this was a mixed sample of individuals with varied history and levels of current treatment, making it difficult to ascertain the role that treatment played in changing QOL and functioning. On the other hand, the data were collected substantially after study treatments occurred, and therefore allowed the full impact of symptom reduction on QOL and functioning to take place. Given our cross-sectional data, we can not exclude the possibility that those in remission had better functioning and QOL even prior to treatment, and that better functioning could be a predictor of remission instead of remission leading to better functioning. Studies examining immediate effects of treatment on QOL and functional impairment may underestimate the long-term effects of treatments. To better determine this, it would have been useful to determine duration of remission to see if this was related to QOL and functional impairment. Prospective studies will be the best method to determine this in future research.
A number of tentative conclusions may be reached from the current study. Patients in remission from OCD appear to be able to function better than OCDs, but may not function as well as extremely healthy controls who have no history of psychiatric disorders. Next, the Sheehan Disability Scales, which are the simplest and easiest to administer seemed to hold up quite well in comparison to other measures of disability. In addition, it is important to consider co-occurring symptoms of anxiety and depression when attempting to optimize functioning and QOL in individuals with OCD. More research is needed on the relationships among long-term course of OCD symptoms and QOL and functioning, and the best methods to optimize both.
We would like to thank staff at each site for their contributions (e.g., clinical evaluators, neuropsychological examiners, research assistants, data managers) and Drs. Michael J. Kozak and Martin E. Franklin for help in the initial study design and procurement of funds. This work was supported by the National Institutes of Mental Health (K23MH064491 to Dr. Huppert, MH45436-06S1 to Drs. Liebowitz and Simpson, MH45404-06S1 to Dr. Foa, and K23 MH001907 to Dr. Simpson).
1Quality of life and functional impairment are distinct concepts. Quality of life here means subjective quality of life, or how content (or dissatisfied) an individual is in a variety of areas of life because of their psychiatric symptoms. Functional impairment is defined as difficulty in engaging in activities in various domains of life due to their psychiatric symptoms. These concepts have some overlap, especially when subjectively rated by self-report, but they are theoretically distinct. For example, one could be functioning well in their work, but still be dissatisfied, or one could be working significantly below potential, but be satisfied.
2Correlations were similar when examining the zero order and partial correlations among only those individuals with current OCD or comorbid OCD and excluding those in remission
3Similar results were found when using the HAM-D instead of the BDI to control for depression