Social Dysfunction in Older Age and Relationships with Cognition,Depression,and Apathy: The GreatAGE Study

Abstract

Background:

Most studies focused on only one measure of social dysfunction in older age, without proper validation and distinction across different dimensions including subjectivity, structural, and functional aspects.

Objective:

We sought to validate the Social Dysfunction Rating Scale (SDRS) and its factorial structure, also determining the association of SDRS with cognitive functions, global psychopathology, and social deprivation.

Methods:

The SDRS was administered to 484 Italian community-dwelling elderly, recruited in the GreatAGE study, a population-based study on aging conducted in Castellana Grotte, Bari, Southern Italy. We determined objective and subjective psychometric properties of SDRS against the gold standard evaluation of social dysfunction according to the Semi-structured Clinical Diagnostic Interview for DSM-IV-TR Axis I Disorders (SCID-I) criterion.

Results:

The SDRS showed a moderate accuracy with an optimal cut-off of 26 maximized with higher sensitivity (0.74,95% CI:0.63–0.84) than specificity (0.57,95% CI:0.50–0.64). A five-factor structure was carried out and five dimensions of SDRS were identified (loneliness; social isolation; feeling of contribution/uselessness; lack of leisure activities; anxiety for the health). Education and global cognitive functions were inversely correlated to SDRS, while a direct association with global psychopathology, depression, and apathy was found. The prevalence of higher SDRS scores was major in subjects with current psychiatric disorders versus other subjects.∥Conclusion: The SDRS could be a valid instrument to capture both size and quality of social dysfunction, both in subjects with psychiatric disorders and in normal subjects. Several categories of social dysfunction differed only in the degree of health deprivation, not in social or material deprivation.

Keywords

Cognitive function deprivation lifestyle loneliness old age social dysfunction

INTRODUCTION

Social functioning is defined as “an individual’s ability to perform appropriately everyday social tasks and consequently to maintain an adequate social life” [1]. Social dysfunction depends on individual inability in coping with stressful situational factors and achieving adequate social gratification, and could be defined as the maladaptive way to manage personal, interpersonal, or geographic environment [2]. Individual social roles are strongly dependent on age and may be affected by the presence of psychopathology [3]. The dimension of social dysfunction is key in study of health and diseases in aging populations, especially cognitive decline [4, 5]. Effective social functioning in older population is linked to the proper social interaction with the environment (parents, siblings, and close relatives). It consists of integrated satisfaction of the person’s needs related to his own goals and how he sees himself as achieving them [3]. In terms of environmental exposure, social deprivation could even be more important in social function than material deprivation, manifesting as a lack of social support from the community and family even in places where living standards are higher. This aspect underlines the effects on individual health of the personal state of isolation and anxiety resulting from a lack of social integration (anomy) [6, 7].

In order to develop a better understanding of social functioning in late-life, research should include the development and validation of more direct, objective measures in conjunction with subjective self-report measures [8, 9]. Unfortunately, most epidemiological studies focus on only one measure of social dysfunction, precluding important comparisons. Evidence from the few studies that do include measures of objective as well as subjective aspects of social relationships suggests that the two dimensions are weakly correlated, and that they have independent effects on health-related outcomes [10]. The Social Dysfunction Rating Scale (SDRS) addresses these requirements because it objectively quantifies dysfunctional interaction with the environment [11]. Subjective and objective performance in the self, interpersonal, and performance systems are assessed. It emphasizes role-free assessment, and this is needed to evaluate functional status of adjustment in older age. Dysfunction is viewed as a discontent accompanied by attitudes of negative self-regard. Adjustment is seen as a process of coping, problem solving, and achieving one’s goals.

We know little about the links between objective measures of social integration and individuals’ subjective assessments of their social connections. The objectives of this study were to validate an Italian version of the 21-item SDRS [11, 12], specifically determining its objective and subjective psychometric properties against the gold standard evaluation of social dysfunction in psychiatric disorders according to the Semi-structured Clinical Diagnostic Interview for DSM-IV-TR Axis I Disorders (SCID-I) [13]. Furthermore, we evaluated the relationship of SDRS with cognitive function, global severity psychopathology, and social deprivation in community-dwelling Italian older subjects recruited in the GreatAGE Study, a population-based study on aging.

METHODS

Participants

The present study adhered to the “Standards for Reporting Diagnostic Accuracy Studies” (STARD) guidelines (http://www.stard-statement.org/), the “Strengthening the Reporting of Observational Studies in Epidemiology” (STROBE) guidelines (https://www.strobe-statement.org/), and in accord with the Helsinki Declaration of 1975. The sample was collected within the GreatAGE Study (Fig. 1), a population-based study on aging conducted on subjects older than 65 and residents in Castellana Grotte, a town located near Bari, Puglia in the Southeast of Italy. The final sample frame was the electoral list including at December 31, 2014, 19,675 subjects, with 4,008 people aged 65 years or older. The GreatAGE Study focused on the impact of nutrition, frailty, and age-related sensory impairments as predictors of common neurodegenerative and psychiatric diseases in older age [14]. The present study was a cross-sectional analysis of the ongoing GreatAGE Study. Detailed methodology of the study is reported elsewhere [15]. We cooperated with general practitioners who invited older subjects previously selected to participate in the study with the support of the city census office. The Institutional Review Board of the “National Institute of Gastroenterology “S. De Bellis” approved this study. Written informed consent was obtained from all the community-dwelling older adults included in this study.

Fig.1

Flow diagram of the study population at the different phases of the survey. The GreatAGE Study.

Data collection and measures assessment

Information was collected through semi-structured questionnaires. Demographic characteristics including age, sex, marital status, household composition, residential arrangement, and education were obtained. The information on psychiatric disorders was obtained from psychiatric examinations with a standardized anamnestic interview for dimensional and categorical diagnosis for all subjects involved in the study. This assessment was followed by the SCID-I [13], performed by trained clinicians, and aimed to categorize the presence of possible major mental disorders. Depression and other current (last month) and past (lifetime) psychiatric disorders were diagnosed according to the American Psychiatric Association criteria (DSM-IV-TR) [16]. Social functioning was defined as “an individual’s ability to perform and fulfill normal social roles”. The impairment of social functioning is considered an important characteristic and diagnostic criterion of psychiatric disorders [2]. Social dysfunction was diagnosed according to the SCID-I and psychiatric interview, in three groups: Group 0 = subjects without psychiatric disorders and without any social dysfunction, Group 1 = subjects without a current psychiatric disorder, but with social dysfunction according to the SCID-I criterion for all psychiatric disorders: “Clinically significant psychological suffering or impairment of social functioning, working or other important functional areas”, and Group 2 = subjects with a psychiatric disorder and affected by current occurrence (diagnosed in the last month) with high social dysfunction. The SCID-I diagnosis of social dysfunction was considered the gold standard. The SDRS and the SCID-I interviews were completed by two independent interviewers in the middle or at the end of the interview, i.e., when a considerable degree of self-disclosure from the respondents might be expected. The SDRS is applicable to all adults, but is designed for older age. The SDRS is composed of twenty-one items rated on a 6-point ordered-rating scale. The rater is required to make judgment of the patient’s performance in relation to the patient’s peer group and compared to the general population. The time period assessed is “today” or “the past week” [17]. Recreational and occupational roles are covered under the performance system. The self and interpersonal items include symptom assessments. Response points are global and the questions require a skilled interviewer [18, 19]. Total SDRS score can be viewed as a simple and rapid measures of general social function. Moreover, measures of 5 factors could be identified according to original factorial analysis: Factor 1 - Apathetic-Detachment, Factor 2 – Dissatisfaction, Factor 3 – Hostility, Factor 4- Health-Finance Concern, and Factor 5 - Manipulative- Dependency. This 21-item scale has also a three-subsystem structure (self system, interpersonal system, and performance system) [11, 19]. The 21 items are rated on a 6-point scale ranging from not present to very severe by a mental health professional using a semi-structured interview guide (score range: 21–126) [11, 19].

Social deprivation was studied with the Deprivation in Primary Care Questionnaire (DiPCare-Q), administered to analyze three areas of material, social, and healthcare deprivation [20]. Clinical variables included cognitive function, severity of psychopathology, and neurological examination. We firstly used the Mini-Mental State Examination (MMSE) [21] and the Clinical Dementia Rating scale (CDR), a tool designed to grade subjects from normal cognition and functions through various stages of dementia [22]. Older subjects with CDR = 0–0.5 (“normal” and “questionable dementia”) and without a diagnosis of dementia were included in the study. The Symptoms Checklist 90 Revised Global Severity Index subdomain (GSI-SCL-90R) was considered as an indicator of the degree of current psychopathology [17]. A battery of tools including the Frontal Assessment Battery (global executive function) [24], the Geriatric Depression Scale (GDS; 30-items) [25], and the Starkstein Apathy Scale (SAS) [26] was used as well.

Statistical analysis

Patients’ characteristics were reported as mean±standard deviation (SD), and frequencies and percentages, for continuous and categorical variables respectively. Differences between the three diagnostic subgroups according to SCID-I were assessed using Pearson χ² for categorical variables and Kruskal-Wallis one-way ANOVA followed by Nemenyi post-hoc test for quantitative ones. Mann-Whitney U-test was used to assess differences in SDRS between two groups. Spearman’s rank correlation coefficient was used to assess associations between quantitative variables. The reliability and the internal consistency of the SDRS has been estimated using Cronbach’s alpha coefficient [27]. The receiver operating characteristic (ROC) curve for SDRS was used to detect the optimal cut-off score maximizing both sensitivity and specificity for the discrimination of social dysfunction across three diagnostic categories: subjects with current psychiatric disorder; subjects without a current psychiatric disorder and with social dysfunction according to the SCID-I criterion for all psychiatric disorders; subjects without psychiatric diagnosis and without social dysfunction. The area under the curve (AUC), sensitivity and specificity were calculated along with their 95% bootstrapped confidence interval (CI), considering 2,000 sampling replications. ROC analyses were performed overall and separately for each subgroup according to sex (male or female), age strata (65–71, 72–79, and 80+ years), and educational level (0–5 or 6–18 years). To determine factor dimensions useful in conceptualizing social dysfunction, an Exploratory Factorial Analysis Vari-max rotation with 5 factors was carried out on the 21-item SDRS. Multivariate logistic regression model was estimated for evaluating the influence of specific variables (educational level, sex, SAS, MMSE, GSI-SCL-90R, and CDR global score) on the performances of the SDRS using a valid cut-off. C-index and Hosmer-Lemeshow goodness of fit tests were used to assess the discrimination and the calibration of the logistic model. A p-value of 0.05 was assumed to be the threshold for statistical significance. All analyses were performed using R (v 3.3.1) and Rstudio (v 0.99.903).

RESULTS

The sample included 484 subjects (46.5% males), constituted by a subgroup of people who completed the SDRS. According to the gold standard diagnosis of social dysfunction performed with the SCID-I, the prevalence of social dysfunction in older age amounted at 30.2% overall, subdivided in 20.5% of subjects with current psychiatric disorders (Group 2) and 9.7% of subjects without current psychiatric disorders, but which meet at least the criterion of social dysfunction of the SCID-I interview (Group 1). Therefore, about 70% of subjects did not have any social dysfunction according to the gold standard. Clinical and socio-demographic variables are compared across the three diagnostic subgroups and are reported in Table 1. Compared to older subjects without psychiatric disorders, social dysfunction in both psychiatric disorders and subjects which met SCID-I criterion for social dysfunction only was more frequently present in females (70–72 % versus 46%). Mean GSI- SCL-90R and GDS-30 scores in subjects with psychiatric disorders were significantly higher than those of older subjects without psychiatric disorders and with at least the SCID-I criterion for social dysfunction. No significant differences were found between the three groups in mean age, education and in mean SAS, MMSE and CDR, FAB, and DiPCare-Q global scores. Among the three DiPCare-Q subdomains, only DiPCare-Q health differed in a statistically significant manner (p = 0.040), not DiPCare-Q social or material (p > 0.005).

Table 1

Clinical and socio-demographic characteristics of the study sample subdivided in three social dysfunction profiles according to the Semi-structured Clinical Diagnostic Interview for DSM-IV-TR Axis I Disorders (SCID-I): Group 0 = subjects without psychiatric disorders and without social dysfunction; Group 1 = subjects without a current psychiatric disorder, but with social dysfunction according to the SCID-I criterion for all psychiatric disorders; Group 2 = subjects with a current psychiatric disorder. The GreatAGE Study

Variable	Category	Subjects without psychiatric disorders		Subjects without current psychiatric disorders, with SCID-I social dysfunction criterion		Subjects with current psychiatric disorders		p
N		338		47		99
		Mean±SD and median (min-max) for continuous variables; Frequency (N) for categorical variables
Age (y)		75.75±6.59	75 (65.00–95.00)	76.19±6.62	76 (65.00–88.00)	75.26±6.96	75 (65.00–92.00)	0.607
Gender	Female	0.46 (156)		0.72 (34)		0.7 (69)		< 0.0001
	Male	0.54 (182)		0.28 (13)		0.3 (30)
Education (y)		6.88±4.06	5.00 (0.00–18.00)	6.03±3.12	5.00 (2.00–13.00)	6.68±4.31	5.00 (0.00–18.00)	0.626
SAS		4.37±3.07	4.00 (0.00–14.00)	4.76±4.99	4.00 (0.00–19.00)	6.89±6.55	5.00 (0.00–26.00)	0.096
GDS-30		2.89±3.37	2.00 (0.00–15.00)	3.87±4.72	2.00 (0.00–21.00)	8.17±7.16	6.00 (0.00–29.00)	< 0.0001
FAB		12.34±3.89	13.00 (4.00–18.00)	12.27±3.36	12.00 (4.00–18.00)	12.47±3.85	13.00 (3.00–18.00)	0.924
MMSE		26.26±3.96	27.00 (13.00–30.00)	26.03±4.75	27.5 (11.00–30.00)	25.56±4.64	27(13.00–30.00)	0.685
CDR global score	0	0.94 (238)		1 (26)		0.92 (56)		0.3136
	0.5	0.06 (14)		0 (0)		0.08 (5)
GSI-SCL-90R		0.16±0.17	0.13 (0.00–1.61)	0.25±0.28	0.18 (0.00–1.69)	0.31±0.19	0.28 (0.06–1.13)	< 0.0001
DiPCare-Q		1.32±0.92	1.36 (0.00–3.9)	1.26±0.94	1.09 (0.00–3.44)	1.38±0.99	1.31 (0.00–4.15)	0.791
SDRS		25.63±4.63	24.00 (21.00–47.00)	27.04±5.02	26.00 (21.00–41.00)	30.98±8.02	28.00 (21.00–57.00)	< 0.0001
Gender	Female	26.27±5.23	25.00 (21.00–47.00)	27.15±5.3	26.00 (21.00–41.00)	30.55±7.6	28.00 (21.00–54.00)	< 0.0001
	Male	25.08±3.98	24.00 (21.00–38.00)	26.77±4.38	25.00 (21.00–36.00)	31.97±8.98	29.00 (21.00–57.00)	< 0.0001
Age strata	65–71 y	25.49±4.37	24.00 (21.00–45.00)	26.92±5.35	25.5 (21.00–37.00)	31.83±8.22	29.00 (21.00–57.00)	< 0.0001
	72–79 y	25.87±5.19	25.00 (21.00–47.00)	27.94±5.63	27.5 (21.00–41.00)	28.72±7.18	26.00 (21.00–54.00)	0.007
	>80 y	25.49±4.28	24.00 (21.00–42.00)	26.18±4.17	26.00 (21.00–34.00)	32.42±8.39	30.5 (21.00–50.00)	< 0.0001
Level of education	0–5 y	26.09±5.19	25.00 (21.00–47.00)	28.81±5.39	29.00 (21.00–39.00)	33.85±9.32	31.00 (21.00–57.00)	< 0.0001
	6–18 y	25.04±4.26	24.00 (21.00–45.00)	24.44±2.13	24.00 (21.00–28.00)	31.7±8.14	29.00 (22.00–54.00)	< 0.0001

SD, standard deviation; SAS, Starkstein Apathy Scale; GDS-30, 30-item Geriatric Depression Scale; FAB, Frontal Assessment Battery; MMSE, Mini-Mental State Examination; CDR, Clinical Dementia Rating scale; GSI-SCL-90R, Symptoms Check-list 90 Revised Global Severity Index subdomain; DiPCare-Q, Deprivation in Primary Care Questionnaire; SDRS, Social Dysfunction Rating Scale.

Box-plots of the SDRS scores of the three diagnostic subgroups are shown in Fig. 2. Median SDRS score of subjects with current psychiatric disorders [28 (21–57)] was significantly higher than those of subjects with at least the SCID-I criterion for social dysfunction [26 (21–41)] and normal subjects [24 (21–47)]. SDRS median scores did not differ between normal subjects (Group 0) and the group which met at least the SCID-I criterion for social dysfunction (Group 1) (p = 0.142). Differences in median SDRS scores among the three diagnostic subgroups were explored by gender, age, education, cognitive, and depression status and the other clinical, behavioral, and psychosocial assessments. Spearman’s correlation between SDRS scores (and its original five factors), age, levels of education, MMSE, FAB, SAS, GDS-30, GSI- SCL-90R scores, and DiPCare-Q and its subdomains are shown in the online-only Table 1. A significant inverse relationship was found between SDRS scores and the levels of education and global cognitive function assessed with MMSE (r = – 0.155, p = 0.005) and FAB (r = – 0.157, p = 0.004), while there was a significant direct correlation of SDRS with global levels of psychopathology (r = 0.470, p < 0.0001), depressive symptoms (r = 0.312, p < 0.0001), and apathy (r = 0.196, p < 0.0001). SDRS scores were significantly associated with gender (higher in females), but not with CDR global scores categories. Spearman’s correlations between SDRS scores and DiPCare-Q subscales did not find any significant association, except for DiPCare-Q Health Deprivation subdomain and Factor 1 - Apathetic-Detachment (Supplementary Table 1).

Fig.2

Comparison of the Social Dysfunction Rating Scale scores among the three social dysfunction diagnostic subgroups according to the Semi-structured Clinical Diagnostic Interview for DSM-IV-TR Axis I Disorders (SCID-I): Group 0 = subjects without psychiatric disorders; Group 1 = subjects without a current psychiatric disorder, but with social dysfunction according to the SCID-I criterion for all psychiatric disorders; Group 2 = subjects with a psychiatric disorder in course (current). The GreatAGE Study.

Using factorial analysis with Vari-max rotation, five factors were identified for the 21-item SDRS. Table 3 shows the five factors with their respective loadings. These five factors explained only 38% of the overall variability. The items SDRS 18 and SDRS 19 had the higher loadings (0.75 and 0.91, respectively) for the Factor 1 (“Social Isolation”); the items SDRS 11, SDRS 12, and SDRS 13 had the higher loadings (0.73, 0.65, and 0.72, respectively) for the Factor 2 (“Loneliness”); the items SDRS 14 and SDRS 15 had the higher loadings (0.85 and 0.70, respectively) for the Factor 3 (“Feelings of Contribution/Uselessness”); the items SDRS 16 and SDRS 17 had the higher loadings (0.54 and 0.90, respectively) for the Factor 4 (“Lack of Leisure Activities”); the items SDRS 9 and SDRS 4 had the higher loadings (0.78 and 0.57, respectively) for the Factor 5 (“Anxiety for the Health”).

Table 2

Performance of 21-item Social Dysfunction Rating Scale (SDRS) compared to the gold standard of a diagnosis of social dysfunction in older age performed with the Semi-structured Clinical Diagnostic Interview for DSM-IV-TR Axis I Disorders at optimal cut-off scores. Optimal cut-off scores referred to the value obtained in the receiver operating characteristic curve analysis maximizing jointly sensitivity and specificity for gender, educational levels, and age strata

	SDRS-21 cut-off scores		Sensitivity (95% CI)	Specificity (95% CI)	Accuracy (95% CI)	PPV (95% CI)	NPV (95% CI)
OVERALL		Optimal	26	0.74 (0.63–0.84)	0.57 (0.50–0.64)	0.72 (0.66–0.77)	0.306	0.894
Gender	Female	Optimal	26	0.72 (0.55–0.85)	0.52 (0.41–0.60)	0.67 (0.59–0.74)	0.352	0.837
	Males	Optimal	27	0.73 (0.51–0.86)	0.70 (0.55–0.80)	0.78 (0.69–0.85)	0.271	0.944
Educational level	Lower than 6 y	Optimal	28	0.64 (0.44–0.79)	0.67 (0.47–0.88)	0.75 (0.64–0.83)	0.283	0.899
	From 6 to 18 y	Optimal	27	0.70 (0.51–0.88)	0.71 (0.57–0.87)	0.80 (0.70–0.88)	0.365	0.910
Age strata	65–71 y	Optimal	27	0.73 (0.56–0.87)	0.67 (0.48–0.79)	0.76 (0.67–0.83)	0.428	0.880
	72–79 y	Optimal	25	0.88 (0.62–0.98)	0.47 (0.21–0.57)	0.64 (0.54–0.73)	0.282	0.941
	80–95 y	Optimal	28	0.65 (0.43–0.80)	0.73 (0.45–0.90)	0.75 (0.63–0.85)	0.333	0.912

PPV, positive predictive value; NPV, negative predictive value.

ROC curve for SDRS was performed on the whole sample to discriminate current psychiatric disorders versus other diagnostic categories (AUC: 0.72, 95% CI: 0.664–0.774) (Fig. 3A). The optimal cut-off score of 26 maximized both sensitivity (0.74, 95% CI: 0.63–0.84) and specificity (0.57, 95% CI: 0.50–0.64). The ROC analyses were performed separately for each subgroup (females, males, three classes of age, and two levels of education) (Fig. 3B-D) and the results are shown in Table 2. The association of sociodemographic and clinical variables (educational level, gender, SAS, MMSE, CDR, and SCID-I diagnosis) on the ratings of SDRS has been evaluated at the cut-off of 26. From the multivariate logistic regression model emerged that only the diagnosis of current psychiatric disorder as established with SCID-I was statistically significant associated to SDRS scores (odds ratio (OR) = 4.08; 95% CI: 1.98–8.40).

Fig.3

Receiver operating characteristic (ROC) curve analysis of the sensitivity and specificity of the 21-item Social Dysfunction Rating Scale compared to the Gold Standard Social Dysfunctional Diagnosis performed with Semi-Structured Clinical Diagnostic Interview for DSM-IV-TR Axis I Disorders across the sample separating subjects with current psychiatric disorders (pathology – Group 0) versus other subjects (without current psychiatric disorders) (no pathology – Groups 1 and 2) (A) and subdivided for gender (female and male) (B), for age strata (65–71 years, 72–79 years, and 80–95 years) (C) and for educational level (lower than 6 years and from 7 to 18 years) (D). The GreatAGE Study.

Table 3

Factorial analysis of the 21 items of the Social Dysfunction Rating Scale (SDRS) with their loadings

Description	Items	Factor 1	Factor 2	Factor 3 Loadings	Factor 4	Factor 5
Low self concept	1			0.276		0.215
Goallessness	2	0.415	0.169
Lack of satisfying philosophy or meaning of life	3		0.183	0.307
Self-health concern	4				0.115	0.573
Emotional withdrawal	5		0.149	0.162		– 0.108
Hostility	6	0.151
Manipulation	7			0.116
Over-dependency	8	0.185	0.106	0.158
Anxiety	9					0.780
Suspiciousness	10	0.171				0.362
Lack of satisfying relationships (family)	11	0.133	0.735	0.189
Lack of friends, social contacts	12	0.414	0.651
Expressed need for more friends, social contacts	13	0.143	0.723	0.198	0.175
Lack of work	14	0.211	0.127	0.855	0.115
Lack of satisfaction from work	15	0.151	0.189	0.707	0.273
Lack of leisure time activities	16	0.298	0.201	0.160	0.547
Expressed need for satisfying activities	17				0.907	0.133
Lack of participation in community activities	18	0.751	0.280		0.124
Lack of interest in community affairs	19	0.915	0.135		0.130
Financial insecurity	20	0.193		0.228	0.209
Adaptive rigidity	21			0.143

The prevalence of social dysfunction in the overall sample of 484 subjects, according to the cut-off score of 26 for the SRDS amounted at 49.17%. According to this optimal cut-off, the SDRS found a prevalence of 73.74% of social dysfunction in the group of current psychiatric disorders, a prevalence of 55.32% of social disturbances in the group of lifelong psychiatric disorders, and a prevalence of 41.12% in the group of older subjects without current psychiatric disorders.

DISCUSSION

The present findings showed that the SDRS was a brief and accurate tool to assess social dysfunction in older age. SDRS identified a difference in social dysfunction between subjects affected by current psychiatric disorders and non-affected subjects or subjects affected by only SCID-I criterion of social dysfunction. The SDRS showed a good validity using a SCID-I diagnosis of social dysfunction as the gold standard, with the best cut-off at 26 in a scale ranging from 21 to 126. To the best of our knowledge, this was the first study testing the validation of SDRS among older subjects in a population-based setting. SDRS presented good accuracy and internal consistency, indicating that this instrument may give acceptable results in evaluating social dysfunction in older age.

In older age, our cut-off for the SDRS showed a good sensitivity (73.7%), but relatively low specificity (57.1%). Only 31% of cases of social dysfunction screened with the SDRS were true positive. In the present study, social dysfunction diagnosed with SCID-I in subjects with and without psychiatric disorders (Groups 1 and 2) was common and amounted at 30.2% in the whole sample. In this group, 9.7% of older people did not meet full criteria for psychiatric conditions (Group 1), but only for social dysfunction, so we supposed they had a moderate social dysfunction according to the clinical interview. The total amount of social dysfunction detected with the SDRS optimal cut-off (≥26) (about 50% in the whole sample) was greater than social dysfunction diagnosed with the SCID-I (about 30%). At the optimal cut-off, the SDRS found a prevalence of social dysfunction ranging from 73.74% in the group of current psychiatric disorders to 41.12% in the group of older normal subjects. Therefore, social functioning may be viewed as one of the basic measures of general well-being that could constitute a continuum from normality to pathology [9 , 28]. However, only the diagnosis of current psychiatric disorders was significantly associated to the SDRS cut-off score, without association with other variables (gender, education, apathy, social deprivation, and global cognitive functions).

The five-factor structure reflected five major dimensions in the SDRS, although this simplification and grouping of the 21 items only explained about 38% of the total variability. The value of the social adjustment derived from the grouping of the items in social isolation, loneliness, feelings of contribution/uselessness, lack of leisure activities, and anxiety for the health, rather than the theorized three-subsystem structure of the 21 items (self system, interpersonal system, and performance system) proposed originally in 1969 [11]. The main loading for the SDRS was attributable to eleven items of the scale, related primarily to the external environment (resources) and the feedback that older persons receive in relationship to their behavior [29]. We hypothesized that, among the five factors, the three items most associated with the Factor 2-Loneliness (loss of satisfactory relationships with significant family members, loss of friends and social contacts, and need to have more social contacts) contributed to the assessment of the loneliness both objectively (by the evaluator) and subjectively (as expressed by the subject). The two items associated with the Factor 1-Social Isolation (lack of participation in community activities and lack of interest in community affairs) contributed to the assessment of the social isolation. Loneliness and social isolation measured the aspects of quality and size of social relations, and thus may also have slightly different associations with health outcomes [10].

A novel finding of the present study was that the strength of social deprivation and social dysfunction in health can differ in the adult life’s course [3]. In fact, we showed that psychiatric and non-affected subgroups diagnosed with SCID-I had similar rates of social and material deprivation, and they differed only in the degree of health deprivation as assessed with the DiPCare-Q (data not shown). Only the Health Deprivation subdomain of DiPCare-Q was significantly different among the three subgroups of social dysfunction, maybe because the items concerning mental health and addiction were included in this subdomain. Furthermore, according to the factorial analysis of Linn conducted in 1988 [9], only the Apathetic-Detachment Factor, encompassing lack of participation and interest in community affairs and lack of friends and satisfying relationships (family), could be correlated to the health deprivation, not to social or material deprivation evaluated with DiPCare-Q.

In the present study, the positive association between SDRS scores and GDS-30, GSI- SCL-90R, and SAS suggested that social adjustment may get worse with increasing depressive symptoms, global index of psychopathology, and apathy. In particular, the SDRS factors of apathetic-detachment, hostility, and dissatisfaction correlated to the scores of GDS-30 and GSI- SCL-90R in our population. Individuals may differ in their proneness to experience psychological distress in the form of negative emotions, i.e., depression and anxiety, a personality trait also known as neuroticism, robustly associated with the future occurrence of negative emotions such as depression and anxiety [30]. We supposed that the personality trait mostly investigated with SDRS may be neuroticism encompassing the emotional withdrawal at item 5, the anxiety described at item 9 (encompassed in Factor 5), and the adaptive rigidity in coping to stress at item 21 of the scale. Vulnerable individuals tend to be unable to deal with stress. Anxiety and vulnerability are markers of withdrawal, which accounts for augmented stress reactivity as well as for negative affect [31].

In the present study, we found an inverse correlation between SDRS and global cognitive scores (MMSE), the level of education, and the executive dysfunction (FAB). The relation of negative social interactions with late-life cognitive disorders has not been extensively investigated, but one study found more frequent negative social interaction to be related to an increased risk of developing mild cognitive impairment [32]. This association appeared to be mainly due to the correlation of negative social interactions that may induce lower social engagement and low level of activities stimulating for cognition. Satisfying interactions tend to be cognitively stimulating and thereby to benefit cognition [5].

Several other questionnaires rather than SDRS have asked persons to rate how often they engage in common social activities. Some of their inconsistency may reflect variation in the quality, rather than the quantity, of social interactions [33]. It was not the size of social networks that predicted the subsequent risk of neuropsychiatric disorders, but how satisfying and reciprocal they were. A certain level of well-being in older age would contribute to the presence of intellectual, social, and psychological resources [34] and perhaps to cognitive reserve. So, our hypothesis is that the SDRS with our proposed cut-off may be very useful in detecting social vulnerabilities in older age, such as social frailty, and considering possible interventions to maintain healthy aging. Among frailty models, physical, cognitive, and social frailty phenotypes have been operationalized. Social frailty can be defined as a continuum of being at risk of losing, or having lost, resources that are important for fulfilling one or more basic social needs during the lifespan [28]. However, social frailty has been operationalized with single questions or items from functional and depressive symptom scales or health checklists [28, 35]. The prevalence of social dysfunction in the group of older subjects without psychiatric disorders in the present study (41.12%) was in line with these studies indicating that cumulative prevalence rate of social frailty (pre-frailty plus frailty) was 35.9% and the prevalence rate of social frailty alone was more that 40% in the higher age strata [35].

The strengths of the present study were the large number of subjects and the comprehensive psychiatric, clinical, and neuropsychological assessment performed. However, we also must acknowledge some limitations. In fact, we had no objective measure of social deprivation (i.e., socio-economic status), but only a subjective measure such as the DiPCare-Q. Furthermore, we did not assess the midlife exposure to deprived social circumstances, a measure associated with the increased risk of severe mental disorder [36]. On the other hand, in the present study, the lack of association between material deprivation and mental health in psychiatric and non-affected subjects may be due also to the homogeneous socioeconomic status of a rural, small towns like Castellana Grotte, without sufficient variability to distinguish several classes of social and/or material deprivation. The SDRS could be a valid instrument to capture size (isolation) and quality (loneliness, neuroticism) of social adjustment in older age. The perception of social dysfunction was not associated with material deprivation, and this adds another stratum of complexity in the assessment of health status in older age.

Footnotes

ACKNOWLEDGMENTS

Project realized with the financial support of the Apulia Region - Regione Puglia: DGR 751/2014PRIN 2009 DM 14 luglio 2011 n. 404/ric - prot. 2009E4RM.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Kupferberg

, Bicks

, Hasler

(2016) Social functioning in major depressive disorder. Neurosci Biobehav Rev 69, 313–332.

Birchwood

, Smith

, Cochrane

, Wetton

, Copestake

(1990) The Social Functioning Scale. The development and validation of a new scale of social adjustment for use in family intervention programmes with schizophrenic patients. Br J Psychiatry 157, 853–859.

Hakulinen

, Pulkki-Råback

, Jokela

, Ferrie

, Aalto

, Virtanen

, Kivimäki

, Vahtera

, Elovainio

(2016) Structural and functional aspects of social support as predictors of mental and physical health trajectories: Whitehall II cohort study. J Epidemiol Community Health 70, 710–715.

Matthews

, Gallo

, Taylor

(2010) Are psychosocial factors mediators of socioeconomic status and health connections? A progress report and blueprint for the future. Ann N Y Acad Sci 1186, 146–173.

Wilson

, Bennett

(2017) How does psychosocial behavior contribute to cognitive health in old age? Brain Sci 7, 56.

Bodner

, Bergman

(2016) Loneliness and depressive symptoms among older adults: The moderating role of subjective life expectancy. Psychiatry Res 237, 78–82.

Abbott

, Freeth

(2008) Social capital and health: Starting to make sense of the role of generalized trust and reciprocity. J Health Psychol 13, 874–883.

Weissman

(1975) The assessment of social adjustment. A review of techniques. Arch Gen Psychiatry 32, 357–365.

Linn

(1988) A critical review of scales used to evaluate social and interpersonal adjustment in the community. Psychopharmacol Bull 24, 615–621.

10.

Elovainio

, Hakulinen

, Pulkki-Råback

, Virtanen

, Josefsson

, Jokela

, Vahtera

, Kivimäki

(2017) Contribution of risk factors to excess mortality in isolated and lonely individuals: An analysis of data from the UK Biobank cohort study. Lancet Public Health 2, e260–e266.

11.

Linn

, Sculthorpe

, Evje

, Slater

, Goodman

(1969) A social dysfunction rating scale. J Psychiatr Res 6, 299–306.

12.

Israel

, Kozarevic

, Sartorius

(1988) VALUTARE L’ANZIANO Vol. 1 Manuale di riferimento dei mezzi di studio e di misura delle funzioni mentali Karger

, p. 213, 1984. https://www.karger.com/Article/Pdf/429248.

13.

First

, Spitzer

, Gibbon

, Williams

JBW

(2002) The Structured Clinical Interview for DSM-IV Axis I Research Version (SCID-I/R Version 2.0). New York State Psychiatric Institute, New York.

14.

Panza

, Solfrizzi

, Logroscino

(2015) Age-related hearing impairment-A risk factor and frailty marker for dementia and AD. Nat Rev Neurol 11, 166–175.

15.

Tortelli

, Lozupone

, Guerra

, Barulli

, Imbimbo

, Capozzo

, Grasso

, Tursi

, Di Dio

, Sardone

, Giannelli

, Seripa

, Misciagna

, Panza

, Logroscino

(2017) Midlife metabolic profile and the risk of late-life cognitive decline. J Alzheimers Dis 59, 121–130.

16.

American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev). Washington, DC.

17.

Goodman

, Schulthorpe

, Evje

, Slater

, Linn

(1969) Social dysfunction among psychiatric and nonpsychiatric outpatients. J Am Geriatr Soc 17, 697–700.

18.

Linn

(1972) Perceptions of childhood and social dysfunction in aging. Read before the annual meeting of the American Sociological Association, New Orleans.

19.

Linn

(1988) Social Dysfunction Rating Scale (SDRS). Psychopharmacol Bull 24, 801–802.

20.

Vaucher

, Bischoff

, Diserens

, Herzig

, Meystre-Agustoni

, Panese

, Favrat

, Sass

, Bodenmann

(2012) Detecting and measuring deprivation in primary care: Development, reliability and validity of a self-reported questionnaire: The DiPCare-Q. BMJ Open 2, e000692.

21.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

22.

Morris

, Heyman

, Mohs

, Hughes

, van Belle

, Fillenbaum

, Mellits

, Clark

(1989) The Consortium to Establish a Registry for Alzheimer’s Disease (CERAD). Part I. Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology 39, 1159–1165.

23.

Derogatis

(1983) SCL-90-R Administration, Scoring & Procedures Manual-II. Clinical Psychometric Research, Towson, MD, pp. 14–15.

24.

Dubois

, Slachevsky

, Litvan

, Pillon

(2000) The FAB: A Frontal Assessment Battery at beside. Neurology 55, 1621–1626.

25.

Yesavage

, Brink

, Rose

, Lum

, Huang

, Adey

, Leirer

(1982-1983) Development and validation of a geriatric depression screening scale: A preliminary report. J Psychiatr Res 17, 37–49.

26.

Starkstein

, Mayberg

, Preziosi

, Andrezejewski

, Leiguarda

, Robinson

(1992) Reliability, validity, and clinical correlates of apathy in Parkinson’s disease. J Neuropsychiatry Clin Neurosci 4, 134–139.

27.

Cronbach

(1951) Coefficient alpha and the internal structure of test. Psychometrika 16, 297–334.

28.

Bunt

, Steverink

, Olthof

, van der Schans

, Hobbelen

JSM

(2017) Social frailty in older adults: A scoping review. Eur J Ageing 14, 323–334.

29.

Linn

(1976) 13. Studies in rating the physical, mental, and social dysfunction of the chronically ill aged. Med Care 14(5 Suppl), 119–125.

30.

Kendler

, Kuhn

, Prescott

(2004) The interrelationship of neuroticism, sex, and stressful life events in the prediction of episodes of major depression. Am J Psychiatry 161, 631–636.

31.

Wilson

, Boyle

, James

, Leurgans

, Buchman

, Bennett

(2015) Negative social interactions and risk of mild cognitive impairment in old age. Neuropsychology 29, 561–570.

32.

Hertzog

, Kramer

, Wilson

, Lindenberger

(2008) Enrichment effects on adult cognitive development: Can the functional capacity of older adults be preserved and enhanced? Psychol Sci Public Interest 9, 1–65.

33.

Finch

, Zautra

(1992) Testing latent longitudinal models of social ties and depression among the elderly: A comparison of distribution-free and maximum likelihood estimates with nonnormal data. Psychol Aging 7, 107–118.

34.

Fredrickson

(2001) The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions. Am Psychol 56, 218–226.

35.

Tsutsumimoto

, Doi

, Makizako

, Hotta

, Nakakubo

, Makino

, Suzuki

, Shimada

(2017) Association of social frailty with both cognitive and physical deficits among older people. J Am Med Dir Assoc 18, 603–607.

36.

Smith

, Nicholl

, Cullen

, Martin

, Ul-Haq

, Evans

, Gill

, Roberts

, Gallacher

, Mackay

, Hotopf

, Deary

, Craddock

, Pell

(2013) Prevalence and characteristics of probable major depression and bipolar disorder within UK Biobank: Cross-sectional study of 172,751 participants. PLoS One 8, e75362.