Family Accommodation in Obsessive Compulsive Disorder and Its Association With Insight

Abstract

Background:

Family accommodation (FA) is defined as a family members’ involvement by doing certain behavior in response to the daily rituals of patients with obsessive-compulsive disorder (OCD). FA is associated with more severe symptoms of OCD. Similarly, poor insight is associated with high severity of OCD symptoms. FA and insight are also related to each other, as more accommodative behavior of family members is associated with poor insight.

Methods:

This cross-sectional observational study assessed 103 adult OCD patients with an illness duration of at least one year. Participants were assessed using Yale–Brown Obsessive Compulsive Scale (Y-BOCS) symptom checklist, Y-BOCS to assess severity, and Family Accommodation Scale-SR (Hindi) to evaluate FA. Insight was assessed with the 11th item of Y-BOCS.

Results:

FA was common in the families of our participants. The mean±SD Y-BOCS score was 28.72±5.09, and the mean FAS-SR score was 44.1±12.03. A significant positive correlation existed between the FA and YBOCS scores, that is, with the increase in FA, YBOCS scores also significantly increased. A considerable number reported poor insight, and had high scores on YBOCS and their caregiver had high FA scores.

Conclusion:

FA is indicative of high symptom severity of OCD and higher FA is associated with poor insight, so FA needs further research for its interplay with OCD symptomatology and role in the maintenance of symptoms.

Key Messages:

Family accommodation is positively correlated with the severity of obsessive-compulsive symptoms and is more severe in patients with poor insight.

Obsessive-Compulsive  Disorder (OCD) is found in 1%–3% of the adult population. ¹ It causes significant interference in various aspects of life, including social, occupational, academic, and family.^2,3 Family accommodation (FA) includes the participation of family members and facilitation and avoidance of certain activities of them, to adjust the obsessive-compulsive symptoms of patients with OCD.^4,5,6 FA is highly prevalent among family members of patients with OCD and predicts more severe symptomatology. ⁷ FA also predicts poorer family functioning ⁶ and mediates its relationship with the severity of OCD. ⁸ High FA is also associated with significant disability, poor quality of life, and high expressed emotion and significantly impacts the outcome of OCD.^9,10 FA has also been associated with contamination compulsions, increased parental OCD, and anxiety symptoms. ¹¹ In family members, variables such as sensitivity to guilt, anxiety sensitivity, and passive communication style have been seen to be predictors of FA. ¹² Two meta-analyses^13,14 have also found a moderate effect size for the association between FA and symptom severity. A recent Indian study reported that FA is associated with greater symptom severity and functional impairment in OCD. ¹⁵ A recent systematic review found consistencies in FA and symptom severity in OCD, but the process of FA has not yet been clarified. Besides, there is no clarity on which of the different variables related to patient illness and caregivers play a significant role in FA. ¹⁶

Only a few studies have examined the relationship between FA, insight, and OCD severity, and most of these have been done in the pediatric population. These studies have reported that patients with poor insight had increased symptom severity, higher levels of OCD-related impairment, and parents of these patients had higher FA compared to the fair-insight group.^{8, 17}

Though ample studies have stated that FA positively correlates with the severity of OCD, there is a lack of studies assessing insight’s role in FA. Our objective was to examine the association between FA and the severity of OCD and the relationship of FA with insight. Based on available research, we hypothesized that FA is positively associated with symptom severity of OCD and insight.

Materials and Methods

Procedure

This was a cross-sectional and observational hospital-based study. The sample size was calculated based on an Indian study by Cherian ¹⁰ using the formula = Z ² × P(1–P)/e ² (power = 80%, level of significance = 0.05, Z = 1.96, prevalence = 50% (unknown), the margin of error = 0.2). A sample of 103 adult patients with OCD, who were visiting the OCD clinic under the Department of Psychiatry, were recruited. The study was approved by the Institutional Ethics Committee. All the patients diagnosed with OCD as per DSM IV-TR criteria ¹⁸ by senior consultants were approached to participate in the study. They were recruited via purposive sampling. The inclusion criteria for patients were age of 18–60 years, illness duration of at least one year, and the availability of a primary caregiver who was a family member continuously staying with the patient for the last year. Exclusion criteria for patients were the presence of any other comorbid psychiatric disorder, and for family members, the presence of any psychiatric disorder. Patients and their family members who agreed to participate in the study and met the inclusion and exclusion criteria were explained about the purpose of the study, following which written informed consent was taken.

Demographic details and clinical characteristics were noted in a specially designed sociodemographic pro forma. The caregivers were screened on General Health Questionnaire (GHQ) ¹⁹ to rule out common mental illnesses, and serious mental illnesses were ruled out by history and interview by a senior psychiatrist. Patients were assessed with Yale–Brown Obsessive Compulsive Scale (Y-BOCS) symptom checklist, Y-BOCS ²⁰ to assess severity, and FA Scale—Self-Rated (Hindi) (FAS-SR) ²¹ to evaluate FA. Insight was assessed using the 11th item of Y-BOCS.

Instruments

GHQ: It was developed by David Goldberg. The GHQ-28 is a 28-item measure of emotional distress in medical settings. It has been divided into four subscales. These are somatic symptoms (items 1–7), anxiety/insomnia (items 8–14), social dysfunction (items 15–21), and severe depression (items 22–28). Test–retest reliability has been reported to be high, 0.78–0.90. ¹⁹

Y-BOCS Symptom Checklist: This includes over 50 types of obsessions and compulsions divided into 15 larger categories according to the behavior expression of the symptoms. ²⁰

Y-BOCS: It is a widely used scale to measure the severity of patients with OCD. In general, the information is gathered from patients, and the final scoring is based on the clinical judgment of the interviewer. It includes 10 questions: five related to obsessions and five, to compulsions. Every question can be rated from 0 (no symptoms) to 4 (extremely severe compulsions and obsessions), with total scores ranging from 0 to 40. The scale has good psychometric properties, with good inter-rater reliability and internal consistency. ²⁰

Item 11 of the YBOCS: It assesses the patient’s capability to acknowledge absurdity and conviction in their obsessive thoughts and behavior. This is scored from 0 (excellent insight) to 4 (lack of insight) as per the individual’s ability to understand the irrationality of obsessive thoughts and the consequences of not following the compulsive behavior in response to these. ²⁰

FAS-SR is a 19-item scale. This was developed by Pinto (2012) as a self- report measure assessing accommodating behaviors in the past week, completed by the relative of the patient with OCD. ²² The FAS-SR is based on the finalized, gold-standard FAS-IR pioneered by Calvocoressi in 1999 with an initial attempt to systematically investigate FA in OCD.^4,23 The FAS-SR parallels the structure of the FAS-IR by including two separate sections; the first includes a detailed OCD symptom checklist and the second includes items that strictly address FA. The recorded responses are identical across both measures, utilizing a 5-point Likert scale assessing the frequency of the accommodating behaviors in the past week. The FAS-SR had demonstrated excellent internal consistency, discriminate validity, and inter-rater reliability. ²²

In this study, the Hindi version of FAS-SR was used. The FAS-SR (Hindi) had demonstrated excellent internal consistency reliability, higher than the internal consistency of the FAS-IR. ²¹

Statistical Analysis

The data were analyzed using Statistical Package for the Social Science (SPSS), Version 20. Descriptive statistics were used, and the Shapiro–Wilk test was used to test for normality. Spearman correlation and linear regression analysis were used to understand the relationship between variables.

Results

In the Shapiro–Wilk test, the value of P < 0.05 was obtained for most variables, indicating that the data was not normally distributed.

Table 1 shows the sociodemographic details and clinical characteristics of the patients. The mean age of the participants was 31.66 years (SD = 9.67).

Table 1.

Sociodemographic and Clinical Profile.

Sociodemographic Details		n (%)
Family type	Nuclear	58(56.31)
	Joint	45(43.69)
Marital status	Married	53(51.46)
	Unmarried	46(44.66)
	Separated	1(0.97)
	Widow	3(2.91)
Socioeconomic status	Lower	1(0.97)
	Upper lower	8(7.77)
	Lower middle	49(47.57)
	Upper middle	43(41.75)
	Upper	2(1.94)
Employment	Employed	28(27.18)
	Unemployed	75(72.82)
Sex	Female	45(43.69)
	Male	58(56.31)
Primary caregivers	Spouse	48(46.61)
	Parents	41(39.81)
	Siblings	7(6.8)
	Son	4(3.89)
	Other	3(2.92)
Obsessions (as per Y-BOCS checklist)	Contamination	72(69.9)
	Doubts	48(46.6)
	Aggressive	38(36.89)
	Religious	34(33.01)
	Symmetry	31(30.1)
	Sexual	23(22.33)
	Somatic	4(3.88)
	Hoarding	10(9.7)
	Miscellaneous	34(33.01)
Compulsions	Checking	50(48.54)
	Cleaning/Washing	73(70.87)
	Counting	18(17.47)
	Repeating	50(48.54)
	Ordering/Arranging	27(26.21)
	Hoarding	10(9.71)
	Miscellaneous	78(75.73)
Level	Mild	1(0.97)
	Moderate	20(19.42)
	Severe	61(59.22)
	Extreme	21(20.39)

Y-BOCS: Yale–Brown Obsessive Compulsive Scale.

The mean age of onset of OCD was 20.51 (SD = 8.32) years. The mean duration of illness was 8.21 (SD = 7.39) years. The duration of untreated illness was 6.11 (SD = 5.57) years. Y-BOCS obsession subscore was 15.41, compulsion subscore was 13.3, and total score was 28.71 (SD = 5.09), with an observed range of scores from 13 to 40. Level of OCD was mild in 0.97%, moderate in 19.42%, severe in 59.22%, and extreme in 20.39%. The mean score for FA was 44.09 (SD = 12.03), with an observed range of scores from 18 to 69. Insight into OCD symptoms was good in 28.16%, fair in 52.43%, and poor in 19.42% of the patients.

Table 2 shows the correlations between scores of Y-BOCS, FA, and insight. The scores were significantly correlated, r(101) = 0.72, P = 0.001.

Table 2.

Spearman Correlations Between Y-BOCS, Family Accommodation, and Insight.

Variable	1	2
Y-BOCS (1)	–
Family accommodation (2)	0.72**
Insight (3)	0.63**	0.59**

**P ≤ 0.01, Y-BOCS: Yale–Brown Obsessive Compulsive Scale.

Table 3 shows the linear regression analysis indicates that changes in FA and insight can explain 48% and 41%, respectively, of changes in Y-BOCS.

Table 3.

Linear Regression Analysis with Family Accommodation, Insight, and Y-BOCS.

Variable	β	R²	T
Family accommodation Insight	0.69*** 0.64***	0.48 0.41	9.83 8.43

*** P ≤ 0.001,. Dependent variable: Y-BOCS. Y-BOCS: Yale–Brown Obsessive Compulsive Scale

Discussion

Most patients had a moderate–severe level of OCD symptomatology, and severity was significantly correlated with the level of FA. Similar findings have been reported in both adult and pediatric subjects with OCD.^11,13,24 One study had also found a significant correlation that was not as strong as in our study. Another study had found a significant correlation between FA and measures of symptom severity, global functioning, and impairment in work and adjustment. ¹⁰ Another study had reported a significant association between FA and treatment refractoriness. ²⁵ So, these studies support that FA has a moderately positive correlation with OCD symptom severity.

Assessment of insight was done using Y-BOCS’ 11th item, and a maximum number of patients had a good–fair insight about their disease. One-fifth of patients had poor or absent insight, which is within the range of proportions found in previous studies.^26,27 Poor insight was also associated with the current level of OCD symptom severity, which is in accordance with previous studies.^26,28-30

There was a significant correlation between the YBOCS scores and insight and between the FA and insight. A moderately negative correlation existed between the FA scale and poor insight, that is, moving from good to poor insight. Similarly, a moderately negative correlation was found between the YBOCS scores and poor insight, that is, moving from good to poor insight. OCD symptomatology and FA were significantly lower in the good insight group compared to the poor insight group. Studies^17,31,32 in pediatric OCD had revealed an inverse relation of insight with OCD symptom severity and FA. Another study had similarly assessed insight on the 11th item of the Children’s YBOCS and found similar results. ³³ So, insight about illness is negatively associated with the extent of accommodative behaviors of family members and OCD symptom severity.

FA is suggestive of severe symptomatology, and insight might play a role as a mediator. Longitudinal studies with multiple assessments during illness are required to understand the correlates of personality variables of family members and phenomenological characteristics of illness and the interplay of these factors.

Limitations

Though a senior psychiatrist made the diagnosis, it would have been better to use instruments to make the diagnosis and rule out other comorbid psychiatric disorders. Though insight was assessed by Y-BOCS 11th item, it would have been better to include other scales to measure insight.

We included only one family member for each patient. Although we included the primary caregiver, it would be appropriate to include other family members too to get more varied responses and perspectives. This is particularly relevant in the Indian context since patients often stay in extended, nuclear, and joint families. It is possible that the responses of family members could vary depending upon multiple as-yet-unexplored factors.

We used GHQ, which can only screen common mental disorders. Though we took detailed history and the family members were interviewed by a senior psychiatrist, it is still difficult to rule out all psychiatric disorders unless a structured diagnostic interview is used for this purpose.

We relied upon self-report; therefore, the responses might have been given in a socially desirable manner. Possible stressors that might be affecting the scores were not studied. Besides, it was a cross-sectional study; a longitudinal study is required to better understand the relationship of insight with FA and its role as a mediating factor.

Conclusion

In OCD patients, FA has a significant positive correlation with symptom severity. Poor insight is associated with high FA and high OCD symptomatology.