Hostname: page-component-cd9895bd7-lnqnp Total loading time: 0 Render date: 2024-12-24T11:12:47.045Z Has data issue: false hasContentIssue false

Personality assessment in nursing home residents with mental and physical multimorbidity: two informant perspectives

Published online by Cambridge University Press:  25 April 2024

Ankie F. Suntjens*
Affiliation:
Radboud University Medical Center, Research Institute for Medical Innovation, Department of Primary and Community Care, University Knowledge Network for Older Adult Care Nijmegen (UKON), Nijmegen, The Netherlands
Ruslan Leontjevas
Affiliation:
Radboud University Medical Center, Research Institute for Medical Innovation, Department of Primary and Community Care, University Knowledge Network for Older Adult Care Nijmegen (UKON), Nijmegen, The Netherlands Open University, School of Psychology, Heerlen, The Netherlands
Anne M. A. van den Brink
Affiliation:
Radboud University Medical Center, Research Institute for Medical Innovation, Department of Primary and Community Care, University Knowledge Network for Older Adult Care Nijmegen (UKON), Nijmegen, The Netherlands
Richard C. Oude Voshaar
Affiliation:
Department of Psychiatry & Interdisciplinary Center for Psychopathology of Emotion Regulation, University of Groningen & University Medical Center Groningen, Groningen, The Netherlands
Raymond T. C. M. Koopmans
Affiliation:
Radboud University Medical Center, Research Institute for Medical Innovation, Department of Primary and Community Care, University Knowledge Network for Older Adult Care Nijmegen (UKON), Nijmegen, The Netherlands De Waalboog, Joachim en Anna, Center for Specialized Geriatric Care, Nijmegen, The Netherlands
Debby L. Gerritsen
Affiliation:
Radboud University Medical Center, Research Institute for Medical Innovation, Department of Primary and Community Care, University Knowledge Network for Older Adult Care Nijmegen (UKON), Nijmegen, The Netherlands
*
Correspondence should be addressed to: Ankie Suntjens, Radboudumc, Afdeling Eerstelijnsgeneeskunde, Huispostnummer 117, Postbus 9101, 6500 HB Nijmegen, The Netherlands. E-mail: [email protected]

Abstract

Objectives:

In older patients with mental and physical multimorbidity (MPM), personality assessment is highly complex. Our aim was to examine personality traits in this population using the Hetero-Anamnestic Personality questionnaire (HAP), and to compare the premorbid perspective of patients’ relatives (HAP) with the present-time perspective of nursing staff (HAP-t).

Design:

Cross-sectional.

Setting:

Dutch gerontopsychiatric nursing home (GP-NH) units.

Participants:

Totally, 142 GP-NH residents with MPM (excluding dementia).

Measurements:

NH norm data of the HAP were used to identify clinically relevant premorbid traits. Linear mixed models estimated the differences between HAP and HAP-t trait scores (0–10). Agreement was quantified by intraclass correlation coefficients (ICCs). All HAP-HAP-t analyses were corrected for response tendency (RT) scores (−10–10).

Results:

78.4% of the patients had at least one premorbid maladaptive trait, and 62.2% had two or more. Most prevalent were: “disorderly” (30.3%), “unpredictable/impulsive” (29.1%) and “vulnerable” (27.3%) behavior. The RT of relatives appeared significantly more positive than that of nursing staff (+1.8, 95% CI 0.6–2.9, p = 0.002). After RT correction, the traits “vulnerable”, “perfectionist” and “unpredictable/impulsive” behavior scored higher on the HAP than HAP-t (respectively +1.2, 95% CI 0.6–1.7, p < 0.001; +2.1, 95% CI 1.3–2.8, p < 0.001; +0.6, 95% CI 0.1–1.1, p = 0.013), while “rigid” behavior scored lower (−0.7, 95% CI −1.3 to −0.03, p = 0.042). Adjusted ICCs ranged from 0.15 to 0.58.

Conclusions:

Our study shows high percentages of premorbid maladaptive personality traits, which calls for attention on personality assessment in MPM NH residents. Results also indicate that the HAP and HAP-t questionnaires should not be used interchangeably for this patient group in clinical practice.

Type
Original Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2024. Published by Cambridge University Press on behalf of International Psychogeriatric Association

Introduction

Personality disorders (PDs) in older adults are highly relevant and require attention in both research and care (Penders et al., Reference Penders, Peeters, Metsemakers and Van Alphen2020). Also at an advanced age, personality pathology is associated with impaired social functioning (Romirowsky et al., Reference Romirowsky, Zweig, Glick Baker and Sirey2021), poorer treatment outcomes of other physical and psychiatric disorders (Morse et al., Reference Morse, Pilkonis, Houck, Frank and Reynolds2005; Stek et al., Reference Stek, Van Exel, Van Tilburg, Westendorp and Beekman2002; Stevenson et al., Reference Stevenson, Brodaty, Boyce and Byth2011; Veerbeek et al., Reference Veerbeek, Oude Voshaar and Pot2014) and lower perceived quality of life (Botter et al., Reference Botter, Ten Have, Gerritsen, de Graaf, van Dijk, van den Brink and Oude Voshaar2021; Condello et al., Reference Condello, Padoani, Uguzzoni, Caon and De Leo2003). Meanwhile, there are still potential treatment options, as first results show feasibility and positive effects of (mediative) cognitive behavioral therapies on PDs in older patients (Botter et al., Reference Botter, Gerritsen and Oude Voshaar2022; Ekiz et al., Reference Ekiz, Videler and Van Alphen2022; Penders et al., Reference Penders, Peeters, Metsemakers and Van Alphen2020).

One of the major challenges is the diagnostic complexity of PDs in later life. It is known that the current Diagnostic and Statistical Manual of Mental Disorders (DSM-V) (American Psychiatric Association, 2013) aims at a younger social and occupational context, containing several PD criteria that lack validity in older age groups (Balsis et al., Reference Balsis, Gleason, Woods and Oltmanns2007; van Alphen et al., Reference Van Alphen, Rossi, Dierckx and Oude Voshaar2014). This may be one of the reasons for lower PD prevalence rates in older adults, with potential underdiagnosis (Balsis et al., Reference Balsis, Gleason, Woods and Oltmanns2007). Diagnosing PDs at a higher age is even more complex in the (co)existence of cognitive deficits, other psychiatric diseases, physical conditions or polypharmacy, as its symptoms and consequences can be difficult to differentiate (Mordekar and Spence, Reference Mordekar and Spence2008; van Alphen et al., Reference Van Alphen, Oude Voshaar, Bouckaert and Videler2018). Cognitive and communicative impairments could also bias self-report (Eleveld et al., Reference Eleveld, Debast, Rossi, Dierckx and Van Alphen2019; Knauper et al., Reference Knäuper, Carrière, Chamandy, Xu, Schwarz and Rosen2016), in addition to the age-unrelated factors that are associated with personality pathology (e.g., limited self-awareness, distorted self-perceptions or reluctance to disclose problems) (Ganellen, Reference Ganellen2007).

These diagnostic issues are eminently present in the long-term care (LTC) setting, concerning older patients with mental and physical multimorbidity (MPM) (Gibson and Ferrini, Reference Gibson and Ferrini2012). Data on PDs in this growing patient group are scarce (Penders et al., Reference Penders, Peeters, Metsemakers and Van Alphen2020; van den Brink et al., Reference Van Den Brink, Gerritsen, Oude Voshaar and Koopmans2013). However, it is known that LTC (nursing) staff experiences PD-associated behavior as particularly challenging (Collet et al., Reference Collet, De Vugt, Ackermans, Engelen and Schols2019; Gibson and Ferrini, Reference Gibson and Ferrini2012). Also, two independent Dutch studies suggest a high prevalence of (probable) PD comorbidity in MPM LTC patients, both of 44% (Collet et al., Reference Collet, De Vugt, Verhey, Engelen and Schols2018; van den Brink et al., Reference Van Den Brink, Gerritsen, De Valk, Oude Voshaar and Koopmans2017). These findings were not further specified into PD types, when and how the diagnoses were established or how they evolved over time.

The Dutch Hetero-Anamnestic Personality questionnaire (HAP) was specifically developed for personality assessment in older adults, including LTC residents. The HAP is completed by relatives to avoid self-report difficulties, based on age-neutral items, and focused on premorbid personality traits to prevent bias from comorbidities (Barendse et al., Reference Barendse, Rossi and Van Alphen2014; Barendse et al., Reference Barendse, Thissen, Rossi, Oei and Van Alphen2013). Complementary, a “present time” version (HAP-t) was developed, which can be filled out by healthcare professionals of LTC facilities to assess current personality traits (Barendse and Thissen, Reference Barendse and Thissen2019). To date, knowledge on how these two perspectives interrelate is limited. Understanding this relationship seems especially relevant in case of MPM complexity, which would enhance the applicability and interpretation of both questionnaires in clinical practice.

In this study, we used the HAP to gain more insight into the premorbid personality traits of LTC residents with MPM. Secondly, we studied agreement between the informant perspectives of patients’ relatives and nursing staff, i.e., between the HAP and HAP-t. Third, we explored age, sex, somatic and psychiatric diseases and cognitive impairments as potential determinants of the extent to which HAP and HAP-t scores differ.

Methods

We used the cross-sectional data from the MAPPING study, that assessed MPM residents of Dutch gerontopsychiatric (GP) nursing home (NH) units (van den Brink et al., Reference Van Den Brink, Gerritsen, De Valk, Oude Voshaar and Koopmans2017).

Participants

Participants of the MAPPING study were recruited from seventeen NHs with a GP care unit in different parts of the Netherlands. Eligibility of the patients was assessed by their elderly care physician (Koopmans et al., Reference Koopmans, Lavrijsen, Hoek, Went and Schols2010) and residents were included if 1) they needed both physical and psychiatric care, also shown in their medical history and 2) psychiatric or behavioral problems were present for at least two years, without the prospect of substantial recovery. The following exclusion criteria were applied: 1) an established diagnosis of dementia, 2) inability or decline to give informed consent, and 3) too severe physical or mental illness for reliable data collection (van den Brink, Reference Van Den Brink2019).

Data collection

The MAPPING data were collected between April 2012 and September 2015, by means of chart reviews, (brief) neuropsychological testing, and structured interviews and questionnaires of both patients, their relatives and their nursing staff. Data collection was partly longitudinal; a prospective cohort study including patients who were newly admitted, performing baseline measurements (6–10 weeks after admission) (T0) and a follow-up assessment after six months (T1). In addition, cross-sectional data were collected of patients who had been residing on the GP-NH unit for at least six months (Tc). This led to the inclusion of 142 MPM patients (63 longitudinal, 79 cross-sectional) (van den Brink, Reference Van Den Brink2019). For our study, we only used the T1 data of the double assessments, ruling out confounding by admission distress.

Personality assessment

Personality traits were assessed with the HAP, which was filled out by a close relative of the participants (when available). It provides the instruction to consider the patient’s life span before significant illness arose, including mental illness and brain damage. This questionnaire has been validated as a screening instrument for PDs in older adults, with available norm scores for old-age psychiatric patients as well as NH residents (including dementia and somatic patients). Its psychometric properties have shown to be sufficient, with adequate inter-item correlations, inter-rater and test-retest reliability, and demonstrated construct, concurrent and criterion validity (Barendse et al., Reference Barendse, Thissen, Rossi, Oei and Van Alphen2013; Barendse and Thissen, Reference Barendse and Thissen2006). The HAP consists of 62 questions, that can be answered with “yes”, “more or less” or “no”. These items comprise ten different personality traits, derivative of the DSM and Millon PD criteria (Millon, Reference Millon1985), namely: socially avoidant (SOC), uncertain (UNC), socially vulnerable (VUL), somatization (SOM), disorderly (DIS), rigid (RIG), perfectionist (PERF), antagonistic (ANT), self-satisfied (SELF) and unpredictable and impulsive (UNP) behavior. A positive (POS) and negative (NEG) response tendency scale were constructed to correct for possible confounding of sympathy or antipathy feelings in the respondent–patient relationship. Each question scores 0, 1 or 2 points, depending on the level of trait confirmation. The number of items per scale differs from 4 to 9, leading to maximum scale scores ranging from 8 to 18 (Barendse and Thissen, Reference Barendse and Thissen2006).

Furthermore, the HAP-t version of the questionnaire was applied, based on behavioral observations of the last six months by a member of the participant’s nursing staff. This questionnaire contains the same items and outcomes as the HAP, but its questions are formulated in the present instead of past tense, with minor adjustments for the NH setting (e.g., “tasks” instead of “work”) (Barendse and Thissen, Reference Barendse and Thissen2019). Both personality questionnaires were administered once, the HAP at T0 and HAP-t at T1 in the longitudinal cohort.

Secondary outcome measures

Additional measurements were used to explore potential determinants of HAP-HAP-t outcome differences. Age, sex and the medical history were extracted from patients’ medical records. Psychiatric and chronic physical disorders were listed as International Statistical Classification of Diseases and Related Health Problems (ICD-10) codes (van den Brink, Reference Van Den Brink2019). Current psychiatric disorders were assessed in semi-structured patient interviews, by means of a validated shorter version of the Schedules for Clinical Assessment in Neuropsychiatry (mini-SCAN) (Nienhuis et al., Reference Nienhuis, Van De Willige, Rijnders, De Jonge and Wiersma2010). We used the data on whether the DSM criteria were met for mood, anxiety, psychotic and substance abuse disorders. Patients’ cognitive status was tested using the Standardized Mini Mental State Examination (S-MMSE) and Frontal Assessment Battery (FAB). The S-MMSE (standardized with more specific instructions) scores correct responses on 11 small cognitive tests, from 0 to 30 points, with a total score ≤23 indicating cognitive deficits (Molloy et al., Reference Molloy, Alemayehu and Roberts1991). The FAB consists of 6 subtests, exploring frontal executive functioning, with a total score ranging from 0 to 18 and scores ≤12 pointing at frontal impairment (Dubois et al., Reference Dubois, Slachevsky, Litvan and Pillon2000; Slachevsky et al., Reference Slachevsky, Villalpando, Sarazin, Hahn-Barma, Pillon and Dubois2004).

Procedure

All data of the MAPPING study were collected by two experienced elderly care physicians, who were well trained in performing the assessments. Questionnaires and tests were conducted in face-to-face interviews with patients and the nursing staff. The HAP was sent by post to the patient’s relative, after patients gave their informed consent. After two and four weeks, relatives were contacted by phone as a reminder and telephonic participation was offered (van den Brink, Reference Van Den Brink2019).

Data analysis

The MPM sample was characterized using descriptive statistics. Differences between patients with a completed and a missing HAP were tested using independent t-tests (or Mann–Whitney U test in case of non-normality) and chi-square tests (or Fisher’s exact when expected cell counts were <5) on, respectively, continuous and categorical data.

For the HAP and HAP-t outcomes, relative scores were calculated ((scale score/maximum scale score)*10) to facilitate interpretability and intercomparability of the different traits. Missing scale items, with a maximum of 2/9 (22%), were corrected by imputing the individual mean of the answered scale items (Barendse and Thissen, Reference Barendse and Thissen2006). Corrected scale scores of the HAP, adjusted for POS and NEG according to the correction formulas in the questionnaire manual, were interpreted against the available NH norm scores (of somatic and psychogeriatric residents combined). This resulted in six benchmark categories (low to very high), with “high” and “very high” (>85th percentile) indicating clinically relevant maladaptive traits (Barendse and Thissen, Reference Barendse and Thissen2006).

Pairwise differences between HAP and HAP-t scores were visualized in Bland and Altman (BA)-plots (y = HAP minus HAP-t, x = mean of HAP and HAP-t), with display of the mean difference and its 95% limits of agreement (LoA) (±1.96*standard deviation), with 95% confidence intervals (CIs). Next to a sense of agreement, this allows for identification of proportional bias and outliers (Bland and Altman, Reference Bland and Altman1986; Giavarina, Reference Giavarina2015).

Linear mixed models (LMMs) were used to estimate mean differences between HAP and HAP-t. Log-likelihood ratio tests compared the goodness-of-fit of different models, with maximum likelihood estimations. For parameter estimates, the restricted maximum likelihood (REML) method was used (Snijders and Bosker, Reference Snijders and Bosker2011). The LMM regression coefficients reflect an estimation of the mean HAP-HAP-t gap (Δ). Intraclass correlation coefficients (ICCs) were calculated as a measure of HAP-HAP-t agreement, with the formula σ2 α/(σ2 α + σ2 ε); σ2 α representing the between-subject and σ2 ε the within-subject variance. Applying LMM for ICC estimates allowed us to adjust for covariates (Nakagawa and Schielzeth, Reference Nakagawa and Schielzeth2010; Pleil et al., Reference Pleil, Wallace, Stiegel and Funk2018).

First, for each personality trait, a basic model was designed: with the type of questionnaire nested in patients. HAP-t was assigned as reference category. Random slopes were added and tested on model improvement, with different covariance structures (Snijders and Bosker, Reference Snijders and Bosker2011). Since missing data may be not missing at random, a dummy variable for a missing questionnaire was created at the patient level, and added as fixed effect to the LMM (model 1) (Bennett, Reference Bennett2001; Son et al., Reference Son, Friedmann and Thomas2012).

Second, we extended the models by taking possible differences in response tendency (RT) between patients’ relatives (HAP) and professional caregivers (HAP-t) into account. For both questionnaires, RT was calculated as: POS − NEG, potentially ranging from −10 to +10 (Barendse and Thissen, Reference Barendse and Thissen2019). Next, RT was group-mean-centered: computed as the deviation of the mean of both RTs (for HAP and HAP-t) in each patient. This transformed RT was added as fixed effect to model 1, correcting the HAP-HAP-t outcome differences for within-patient RT differences (model 2) (Bell et al., Reference Bell, Jones and Fairbrother2018).

Third, we explored the influence of potential moderating variables on the HAP-HAP-t gap, including age, sex, number of somatic and psychiatric diagnoses in the medical history, and current psychiatric diagnoses (mini-SCAN) and cognitive status (MMSE, FAB). The effect of each potential determinant was analyzed in separate models. The covariables and their interaction with the type of questionnaire were added as fixed effects to model 2 (corrected for the missing data pattern and RT). In case of a significant effect of the missing pattern in model 1, this dummy variable was also added in an interaction term with the potential moderator (Son et al., Reference Son, Friedmann and Thomas2012). For each determinant, the variance explained on the questionnaire level (between HAP and HAP-t), i.e., the proportion change in variance (PCV), was determined by comparing the estimated within-subject variance (σ2 ε) to that of model 2 (σ2 ε2 − σ2 ε3)/σ2 ε2) (Nakagawa and Schielzeth, Reference Nakagawa and Schielzeth2013).

P-values <0.05 were considered significant. HAP-HAP-t differences >1.4 were defined as clinically relevant, being below average consensus (corrected for RT) (Barendse and Thissen, Reference Barendse and Thissen2019). Post hoc analysis showed that our sample size was sufficient for all traits, with a power of 0.85–0.95 (α = 0.05), to identify this 1.4 difference. ICC values <0.5 were indicated as poor, 0.5–0.75 as moderate, 0.75–0.9 as good, and >0.90 as excellent agreement (Koo and Li, Reference Koo and Li2016). All statistical analyses were performed using SPSS 27.0 software.

Ethical considerations

Data collection of the MAPPING study was approved by the Medical Research Ethics Committee Arnhem-Nijmegen, which also declared that it did not fall within the remit of the Medical Research Involving Human Subjects Act. The database was re-used in a pseudonymized version, without access to patients’ personal information. All of this was covered by participant informed consent.

Results

Patient characteristics

Of the 142 participants, the HAP was completed by a close relative in 111 patients (78.2%) and the HAP-t by a nursing staff member for all patients (100%). Patient characteristics are presented in Table 1, including the comparison of cases with and without a filled out HAP. These groups did not significantly differ in demographic and MPM features. However, patients with a missing HAP were slightly younger, less educated and more frequently unmarried. Clinically, this group had more (previous and current) PDs and other psychiatric diagnoses and showed little less cognitive impairments. Overall, 62 patients (43.7%) already had a PD diagnosis in their medical history, and for 48 patients (33.8%) this was the primary reason for NH admission.

Table 1. Demographic and clinical characteristics: comparing participants with a completed versus a missing HAP questionnaire

Note: N = number, SD = standard deviation, mini-SCAN (shorter version of the Schedules for Clinical Assessment in Neuropsychiatry); NPI-NH (Neuropsychiatric Inventory Nursing Home version, 0–144); S-MMSE (Standardized Mini Mental State Examination, 0–30); FAB (Frontal Assessment Battery, 0–18); a) missing data: n = 2 (1.4%) in “present” group; b) missing data: n = 5 (3.5%), 3 in “missing” group; c) missing data: n = 7 (4.9%), 3 in “missing” group; 1) independent t-test (2-tailed); 2) chi-square test; 3) Mann–Whitney U test; 4) Fisher’s exact test.

HAP personality profiles

Table 2 shows the results on the ten different personality traits of the HAP. Based on the mean relative scores, the highest scoring traits were: “rigid” (5.64), “perfectionist” (5.13) and “vulnerable” (4.98) behavior. Meanwhile, the most prevalent maladaptive traits (norm-referenced as “high” or “very high”) were: “disorderly” (30.3%), “unpredictable/impulsive” (29.1%) and “vulnerable” (27.3%) behavior (Figure 1). The proportion of patients with at least one maladaptive trait was 78.4% (87/111). In 62.2% of the patients (69/111), two or more maladaptive traits were seen. The number of maladaptive traits per patient had a median of 2, with an interquartile range of 1–4.

Table 2. HAP results: trait scores and frequencies of maladaptive traits

Note: N = number; SD = standard deviation; relative score = ((raw scale score/maximum scale score)*10), maximum scale score = number of scale items*2; norm-referenced = compared to norm scores of somatic and psychogeriatric nursing home patients, as provided by the HAP questionnaire manual; “high” or “very high” = percentile score >85th.

Figure 1. Stacked histogram: norm referenced HAP results.

Note : norm referenced = compared to norm scores of somatic and psychogeriatric nursing home residents, as provided by the HAP questionnaire manual; “very high” = percentile score ≥96th; “high” = percentile score 86th–95th; “above average” = percentile score 66th–85th; “average” = percentile score 36th–65th; “below average” = percentile score 16th–35th; “low” = percentile score ≤15th.

Of the patients with a PD diagnosis in their medical history, 82.6% (38/46) had at least one maladaptive trait on the HAP versus 75.4% (49/65) of the patients without a previous diagnosis (p = 0.362). Of the patients with and without a PD as primary reason for admission, the HAP showed one or more maladaptive traits in 88.6% (31/35) and 73.7% (56/76) (p = 0.077), respectively.

Plotted differences of HAP and HAP-t

The BA-plots of the paired HAP and HAP-t results (n = 111) show similar graphics for all ten personality traits: a “rhombus shape”, with relatively flat regression lines close to the mean difference (little proportional bias) and few outliers. However, the identified differences and their limits of agreement are substantially scattered, indicating little absolute agreement or consistency between the two questionnaires. The width of the 95% LoA ranged from 12.16 to 16.12 points. Figure 2 shows the BA-plots of the personality traits with the largest (PERF) and smallest (UNC) mean difference of 2.18 and 0.25 points, respectively, with −5.36 to 9.73 and −5.83 to 6.33 as 95% LoA.

Figure 2. Bland and Altman plots: HAP-HAP-t differences for PERF and UNC.

Note: relative scores (0–10); x = (HAP + HAP-t)/2, y = HAP – HAP-t; LoA (limits of agreement) = mean difference ± (1.96*standard deviation); 95% CI (95% confidence intervals) = ± (standard error*t value for degrees of freedom).

Mean differences and agreement of HAP and HAP-t

The LMM results are shown in Table 3. Adding random slopes did not significantly improve the models, therefore only random intercepts were used.

Table 3. Linear mixed model results: estimated differences and ICCs of HAP and HAP-t traits

Note: using restricted maximum likelihood, model 1 = random intercept, with a missing-HAP dummy, model 2 = model 1 + group-mean-centered response tendency.

Δ = estimated gap between HAP and HAP-t relative scores (0–10), with HAP-t as a reference; 95% CI = 95% confidence interval; P-value: for the difference between HAP and HAP-t; *significant difference, p-value <0.05; ICC = intraclass correlation coefficient, calculated with the variance estimate outcomes.

RT (response tendency) = positive – negative scale score (−10−10); SOC (socially avoidant); UNC (uncertain); VUL (vulnerable); SOM (somatizing); DIS (disorderly); RIG (rigid); PERF (perfectionist); ANT (antagonistic); SELF (self-satisfied); UNP (unpredictable and impulsive).

The RT of close relatives (HAP) was on average positive (mean +1.72, SD 4.69), where the RT of professional caregivers (HAP-t) was slightly negative (mean −0.06, SD 4.50). This difference appeared statistically significant (95% CI 0.63–2.86, p = 0.002). Correction for the within-patient RT differences (model 2) led to a significant change of outcomes and better model fit in nine out of ten personality traits (excluding “perfectionist” behavior). These models showed higher scores of the HAP compared to the HAP-t for “vulnerable” (+1.19, p < 0.001), “perfectionist” (+2.07, p < 0.001) and “unpredictable/impulsive” (+0.63, p = 0.013) behavior. Lower HAP than HAP-t scores were shown for “rigid” behavior (−0.67, p = 0.042). For the other six personality traits, no significant differences between HAP and HAP-t were found. Correction for RT differences improved the HAP-HAP-t agreement, with ICCs ranging from 0.12 to 0.41 in model 1, versus 0.15 to 0.58 in model 2.

Moderating variables

Age was only significantly associated with the HAP-HAP-t gap of “socially avoidant” behavior (b = 0.05, 95% CI 0.00−0.10), while sex showed no association for any of the ten personality traits. The number of somatic diagnoses was associated with the gaps of “socially avoidant” and “uncertain” behavior, with b = 0.22 (95% CI 0.03−0.42) and b = 0.23 (95% CI 0.04−0.42), respectively. The number of psychiatric diagnoses was not associated with the HAP-HAP-t gap for any of the personality traits, neither when based on the medical history nor when assessed with the mini-SCAN. Nonetheless, examining specific current psychiatric diagnoses (mini-SCAN) showed associations of mood disorders with the gap for “antagonistic” behavior (b = −0.92, 95% CI −1.82 to −0.02), anxiety disorders with the gap for “unpredictable” behavior (b = 1.21, 95% CI 0.09–2.32) and substance abuse with the gap for “disorderly” behavior (b = −3.73, 95% CI −6.69 to −0.76). Cognitive performance was associated with the HAP-HAP-t gap for “perfectionist” behavior (MMSE: b = −0.17, 95% CI −0.31 to −0.04; FAB: b = −0.21, 95% CI −0.36 to −0.05). The MMSE score was also a determinant of the gap for “disorderly” behavior (b = 0.13, 95% CI 0.01–0.25). These significant moderating effects individually explained 1.8% to 9.8% of the variance between HAP and HAP-t (PCV), with the largest effect of cognitive impairments on the “perfectionist” gap. All results can be viewed in Table S1, published as supplementary material (online attached to the electronic version of this paper at https://www.cambridge.org/core/journals/international-psychogeriatrics).

Discussion

This study aimed to examine maladaptive personality traits – and the informant perspectives of close relatives and nursing staff – in MPM NH residents, using the HAP and HAP-t questionnaire.

Results showed that almost four in five patients had a premorbid maladaptive personality trait, with a vast majority having two or more, according to ratings of close relatives. This indicates that the prevalence of personality pathology in MPM NH residents is (very) high. When comparing the HAP and HAP-t questionnaires, it was found that patients’ relatives tended to answer the questions more positively than members of the nursing staff. Accounting for this different rating tendency, the traits of “vulnerable”, “perfectionist” and “unpredictable/impulsive” behavior scored higher on the HAP (premorbid), while “rigid” behavior scored higher on the HAP-t (present). Nonetheless, only “perfectionist” behavior showed a difference that could be considered clinically relevant, which was partly explained by cognitive decline. While these mean differences between HAP and HAP-t can be seen as minor, BA-plots showed substantial paired differences for all traits. This was confirmed by ICCs that were poor to moderate at best, even after RT correction. These findings imply little agreement of the two questionnaires within individuals. Overall, age, sex and physical and mental comorbidities appeared to minimally explain the HAP-HAP-t differences.

Interpretation

First, our HAP results showed that the highest trait scores were not equivalent to those with the highest norm-referenced labels. This underlines the importance of looking at behavior in a sociocultural context, as stated in the DSM PD criteria (American Psychiatric Association, 2013). In this light, our results can be interpreted as high rates of maladaptive personality traits in the history of GP-NH patients in comparison to NH patients on a somatic or psychogeriatric ward. The HAP provides no benchmark data of “healthy” controls (Barendse and Thissen, Reference Barendse and Thissen2006). Since it could be argued that NH patients in general have more personality pathology than community-dwelling older adults (i.e., impaired social functioning increases the risk of LTC admission) (Jamieson et al., Reference Jamieson, Abey-Nesbit, Bergler, Keeling, Schluter, Scrase and Lacey2019), our results might even be an underestimation relative to the age-matched general population. This would be in line with previous meta-analyses (Friborg et al., Reference Friborg, Martinsen, Martinussen, Kaiser, Overgård and Rosenvinge2014; Friborg et al., Reference Friborg, Martinussen, Kaiser, Overgard and Rosenvinge2013), showing that PDs are much more common in patients with other psychiatric diseases (mood and anxiety disorders) compared to the overall population.

Meanwhile, the risk of underdiagnosis of PDs in older patients seems to be reflected in our results, with 78% having (a) HAP maladaptive trait(s) versus 44% with a reported PD in their medical history, and no significantly less maladaptive traits in patients without a previous diagnosis. It should, however, be mentioned that the HAP is developed as a screening instrument. It does not establish a PD diagnosis, but only gives an indication, interpreting the whole 10-trait profile (Barendse and Thissen, Reference Barendse and Thissen2006). The frequency of full-criteria PDs is therefore likely to be lower than that of maladaptive traits. Nevertheless, only 16% of the patients had one maladaptive trait, while 62% had two or more, which remains highly aberrant. Also, a HAP maladaptive trait is considered clinically relevant in itself (Barendse and Thissen, Reference Barendse and Thissen2006), and accounting for traits and “subthreshold” PDs in older patients is recommended in literature (Botter et al., Reference Botter, Ten Have, Gerritsen, de Graaf, van Dijk, van den Brink and Oude Voshaar2021; Oltmanns and Balsis, Reference Oltmanns and Balsis2011).

Another interesting finding was the generally more positive rating tendency of close relatives compared to nursing staff. To our knowledge, our study is the first to compare those two perspectives. The tendency of informants to rate personality in an overly positive, socially desirable manner has been described before. This mostly concerned family and friends, who – like in our study – were appointed or approved by the subjects themselves, creating a selection bias known as the “letter of recommendation” effect (Leising et al., Reference Leising, Erbs and Fritz2010). Our results suggest that this is less applicable to ratings of professional caregivers. The level of relationship intimacy might play a role in this. Previous research indicates that a more intimate relationship (i.e., partnership) leads to a higher concordance between informant- and self-ratings of personality (Eleveld et al., Reference Eleveld, Debast, Rossi, Dierckx and Van Alphen2019). For the HAP, no significant effect of the relationship type (e.g., spouse, child, other relative) on the inter-rater reliability was identified (Barendse and Thissen, Reference Barendse and Thissen2019). The HAP showed ICCs of 0.67 to 0.85 for the different personality traits, and respectively 0.84 and 0.65 for POS and NEG (Barendse et al., Reference Barendse, Thissen, Rossi, Oei and Van Alphen2013). These are much higher ICCs than found in our study, suggesting that the perspectives of relatives and nurses differ much more than those of relatives amongst each other.

After RT correction, differences between HAP and HAP-t scores could be interpreted as personality changes over time. Alterations of personality in time and its stability into old age have been studied with the Five Factor-model (FFM) (Debast et al., Reference Debast, van Alphen, Rossi, Tummers, Bolwerk, Derksen and Rosowsky2014). Looking at correlations between the HAP and FFM (in a NH setting) (Barendse and Thissen, Reference Barendse and Thissen2006), conclusions in the review of Debast et al. show notable overlap with our results. The decrease of “vulnerable”, “perfectionist” and “unpredictable/impulsive” behavior correlates with a general decrease of neuroticism with age. The increase of “rigid” behavior with a decrease in openness and extraversion. The mean HAP-HAP-t differences that we found might thus be partially explained by aging itself. This was confirmed by the relatively small effects in our moderation analyses, by which we aimed to disentangle behavioral changes due to illness from “real” personality changes in time.

Lastly, the “perfectionist” trait appeared to differ from the other traits in several ways. It showed the largest and only clinically relevant HAP-HAP-t difference, solely not being confounded by RT variation. Also, the decrease of “perfectionist” behavior in our results did not match the general increase of FFM consciousness with age (Debast et al., Reference Debast, van Alphen, Rossi, Tummers, Bolwerk, Derksen and Rosowsky2014). This discrepancy seems partly explained by cognitive decline (moderating effects of MMSE and FAB) in our study population, which makes clinical sense. Additionally, we hypothesize that the “perfectionist” trait might lose criterion validity in translation of the HAP to the present time (HAP-t), since three out of four questions are related to “tasks” (Barendse and Thissen, Reference Barendse and Thissen2019), which can be limited in a NH setting.

Strengths and limitations

Our study has several strengths, such as a thorough statistical approach. HAP and HAP-t trait scores were compared by considering means on the group-level, and by analyzing agreement within individuals (BA-plots, ICCs) (Watson and Petrie, Reference Watson and Petrie2010). Both mean differences and ICCs were corrected for the missing HAP questionnaires and RT differences between HAP and HAP-t. This makes our results directly translatable to clinical practice. In addition, the HAP seems a suitable questionnaire for our study population, overcoming several age-related assessment difficulties. Personality results were displayed as trait spectra (Figure 1) instead of dichotomous PD outcomes, as preferred in older populations (Penders et al., Reference Penders, Peeters, Metsemakers and Van Alphen2020).

Next, some limitations of the study should be mentioned. First, we only norm-referenced the results of the HAP and not the HAP-t. We chose to focus on the premorbid traits, since chronic psychiatric diseases can bias the HAP-t interpretation (Barendse and Thissen, Reference Barendse and Thissen2019). Furthermore, different norm populations are available for the HAP-t, not including somatic and psychogeriatric NH residents (Barendse and Thissen, Reference Barendse and Thissen2019), making the categorized results of HAP and HAP-t not directly comparable. By solely comparing the numerical scores, it remains unclear whether the HAP-HAP-t differences that we found are adaptive (”normal”) or maladaptive, in the changing context of aging, illness and NH admission. Second, the perspective of MPM patients themselves is missing. The HAP was specifically designed for LTC patients with brain damage and/or severe mental illness, for whom self-report is not considered reliable (Barendse and Thissen, Reference Barendse and Thissen2006). Looking at the characteristics of the MAPPING participants, the choice for this informant questionnaire seems valid. Informant reporting also appears better than self-report in the assessment of externalizing personality traits and interpersonal functioning, and as a predictor of adaptability and health (Eleveld et al., Reference Eleveld, Debast, Rossi, Dierckx and Van Alphen2019). However, self-report better reflects intrapsychic characteristics, which form an important aspect of “personality” too. Agreement of self- and informant reports on personality (in older adults) is shown to be low to moderate, so both seem to offer unique information (Eleveld et al., Reference Eleveld, Debast, Rossi, Dierckx and Van Alphen2019). Third, the HAP was administered at different timepoints for the longitudinal and cross-sectional cohorts, respectively 6–10 weeks and ≥6 months after NH admission. Because it explores the premorbid situation, we consider the risk of bias to be low, i.e., the time interval to the premorbid situation is not determined by these timepoints. Recall bias might be present in all cases. Also, reviews of the premorbid situation could be biased by current behavior. The same goes for present-time reviews and awareness of the patient’s history. This, however, is expected to increase HAP-HAP-t agreement, so does not explain the low ICCs in our study. The HAP-t was applied at least six months after NH admission for all patients, corresponding to the recommended minimal observation period (Barendse and Thissen, Reference Barendse and Thissen2019). Fourth, the HAP results are not directly translatable to the presence or absence of a PD, according to the current DSM criteria. While correlations with the DSM have been studied (Barendse et al., Reference Barendse, Rossi and Van Alphen2014; Barendse and Thissen, Reference Barendse and Thissen2019), this creates challenges for the comparability of our results. Fifth, our moderation analyses should be considered exploratory, with relatively small sample sizes and no correction for multiple testing. Nevertheless, the significant effects that were found seem clinically reasonable, and adopting lower p-values would strengthen our conclusion that the tested covariates poorly explain the low ICCs.

Implications and recommendations

Our study sample consists of “gerontopsychiatric” NH residents. In the Netherlands, separate GP-NH units were created for MPM patients who need specialized care due to behavioral problems. This is part of a Dutch development of unique LTC expertise networks (Koopmans et al., Reference Koopmans, Leerink and Festen2022). In addition, our study used a Dutch personality instrument. Yet, we believe that our findings have broader and international relevance, since older patients with MPM are found worldwide and, although cultural norms may differ, personality traits are more universally recognizable.

Several implications can be drawn from our results. 1) The high rates of maladaptive traits stress the need for specific attention on personality pathology in MPM NH patients. Implementation of standard personality assessments is strongly recommended in this group, to prevent underdiagnosis and undertreatment. Preferably, normative HAP(-t) results become available of an age-matched population without mental disorders. Insight into patients’ maladaptive personality traits could provide guidance in dealing with behavioral problems, with a potential win-win for patients and their nursing staff (Penders et al., Reference Penders, Peeters, Metsemakers and Van Alphen2020). Future studies into the effects of maladaptive personality traits on patient well-being and caregiver burden in MPM patients, and intervention effects in this regard, are required. 2) Both researchers and practitioners should be aware that relatives may have a more positive view on patients’ personality than professional caregivers. Further research needs to confirm whether this is a general phenomenon in MPM NH residents, other (older) patient groups and with other personality instruments. 3) Poor agreement implies that the HAP and HAP-t should not be used interchangeably in individual patients. In our study, the HAP was missing in a considerable proportion of cases (22%). Most important reasons for this were the lack of close relatives and participant’s refusal of engaging them (van den Brink et al., Reference Van Den Brink, Gerritsen, de Valk, Mulder, Oude Voshaar and Koopmans2018). In our experience, this is a realistic reflection of the GP practice. It therefore would have been convenient if the HAP-t could be used to replace or predict HAP outcomes. Based on our results, however, this does not seem valid. The HAP-t was also designed to provide unique information, complementary to the HAP (Barendse and Thissen, Reference Barendse and Thissen2019). We looked into this from a psychometric approach. Additional research is needed to establish how (premorbid) personality traits relate to the (challenging) behavior that is observed on GP-NH wards (Collet et al., Reference Collet, De Vugt, Verhey, Engelen and Schols2018; van den Brink et al., Reference Van Den Brink, Gerritsen, De Valk, Oude Voshaar and Koopmans2017).

Conclusion

Our study suggests a (very) high prevalence of premorbid maladaptive personality traits in MPM NH residents, which requires attention in both research and clinical practice. When using informant perspectives for personality assessment in this patient group, a more positive rating tendency of patients’ relatives compared to nursing staff should be taken into account. Despite relatively small mean differences between HAP and HAP-t results, low paired agreement indicates that these questionnaires are not directly interchangeable on an individual level.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/S1041610224000474.

Conflicts of interest

None.

Source of funding

This research was funded by ZonMw (the Netherlands Organisation for Health Research and Development).

Description of authors’ roles

A. Suntjens developed the statistical design of the study, carried out the analyses and wrote the paper. R. Leontjevas was also responsible for the statistical design and analysis, supervised the writing process, and reviewed the paper multiple times. A. van den Brink collected the original MAPPING data, supported the use and interpretation of the data, and reviewed the paper. R. Oude Voshaar, R. Koopmans, and D. Gerritsen formulated the research questions and basic study design, supervised the research process, assisted in the interpretation of the results, and reviewed the paper.

Acknowledgements

None.

References

American Psychiatric Association (APA) (2013). Diagnostic and statistical manual of mental disorders, 5th edn. American Psychiatric Association.Google Scholar
Balsis, S., Gleason, M. E., Woods, C. M., & Oltmanns, T. F. (2007). An item response theory analysis of DSM-IV personality disorder criteria across younger and older age groups. Psychology and Aging, 22(1), 171185. https://doi.org/10.1037/0882-7974.22.1.171 CrossRefGoogle ScholarPubMed
Barendse, H. P., Rossi, G., & Van Alphen, S. P. (2014). Personality disorders in older adults: expert opinion as a first step toward evaluating the criterion validity of an informant questionnaire (HAP). International Psychogeriatrics, 26(1), 173174. https://doi.org/10.1017/S1041610213001312 CrossRefGoogle ScholarPubMed
Barendse, H. P., & Thissen, A. J. (2006). Handleiding van de Hetero-Anamnestische Persoonlijkheidsvragenlijst H.A.P. HAP Uitgeverij Nijmegen.Google Scholar
Barendse, H. P., & Thissen, A. J. (2019). Handleiding van de Hetero-Anamnestische Persoonlijkheidsvragenlijst HAP en HAP-t versie 2.0. HAP Uitgeverij Nijmegen.Google Scholar
Barendse, H. P. J., Thissen, A. J. C., Rossi, G., Oei, T. I., & Van Alphen, S. P. J. (2013). Psychometric properties of an informant personality questionnaire (the HAP) in a sample of older adults in the Netherlands and Belgium. Aging & Mental Health, 17, 623629. https://doi.org/10.1080/13607863.2012.756458 CrossRefGoogle Scholar
Bell, A., Jones, K., & Fairbrother, M. (2018). Understanding and misunderstanding group mean centering: a commentary on Kelley et al.’s dangerous practice. Quality & Quantity, 52, 20312036. https://doi.org/10.1007/s11135-017-0593-5 CrossRefGoogle Scholar
Bennett, D. A. (2001). How can I deal with missing data in my study? Australian and New Zealand Journal of Public Health, 25, 464469.CrossRefGoogle Scholar
Bland, J. M., & Altman, D. G. (1986). Statistical methods for assessing agreement between two methods of clinical measurement. Lancet, 1(8476), 307310.CrossRefGoogle ScholarPubMed
Botter, L., Gerritsen, D. L., & Oude Voshaar, R. C. (2022). Schema therapy in the nursing home setting: a case study of a cognitively impaired patient. Clinical Case Studies, 21, 552570. https://doi.org/10.1177/15346501221091790 CrossRefGoogle Scholar
Botter, L., Ten Have, M., Gerritsen, D., de Graaf, R., van Dijk, S. D. M., van den Brink, R. H. S., Oude Voshaar, R. C. (2021). Impact of borderline personality disorder traits on the association between age and health-related quality of life: a cohort study in the general population. European Psychiatry, 64(1), e33. https://doi.org/10.1192/j.eurpsy.2021.27 CrossRefGoogle ScholarPubMed
Collet, J., De Vugt, M. E., Ackermans, D. F. G. V., Engelen, N.J.J.A, & Schols, J. M. G. A. (2019). Experiences and needs of nursing staff caring for double care demanding patients: a qualitative study. Journal of Gerontology & Geriatric Research, 8(2), Article 1000501. https://doi.org/10.4172/2167-7182.1000501 Google Scholar
Collet, J., De Vugt, M. E., Verhey, F. R. J., Engelen, N., & Schols, J. (2018). Characteristics of double care demanding patients in a mental health care setting and a nursing home setting: results from the SpeCIMeN study. Aging & Mental Health, 22, 3339. https://doi.org/10.1080/13607863.2016.1202891 CrossRefGoogle Scholar
Condello, C., Padoani, W., Uguzzoni, U., Caon, F., & De Leo, D. (2003). Personality disorders and self-perceived quality of life in an elderly psychiatric outpatient population. Psychopathology, 36(2), 7883. https://doi.org/10.1159/000070362 CrossRefGoogle Scholar
Debast, I., van Alphen, S. P. J(Bas), Rossi, G., Tummers, J. H. A., Bolwerk, N., Derksen, J. J. L., Rosowsky, E. (2014). Personality traits and personality disorders in late middle and old age: do they remain stable? A literature review Clinical Gerontologist, 37(3), 253271. https://doi.org/10.1080/07317115.2014.885917 CrossRefGoogle Scholar
Dubois, B., Slachevsky, A., Litvan, I., & Pillon, B. (2000). The FAB: a frontal assessment battery at bedside. Neurology, 55(11), 16211626. https://doi.org/10.1212/wnl.55.11.1621 CrossRefGoogle ScholarPubMed
Ekiz, E., Videler, A. C., & Van Alphen, S. P. J. (2022). Feasibility of the cognitive model for behavioral interventions in older adults with behavioral and psychological symptoms of dementia. Clinical Gerontologist, 45(4), 903914. https://doi.org/10.1080/07317115.2020.1740904 CrossRefGoogle ScholarPubMed
Eleveld, M., Debast, I., Rossi, G. M. P., Dierckx, E., & Van Alphen, S. P. J. (2019). [Concordance and added value of informant- versus self-report in personality assessment: a systematic review]. Tijdschrift voor Gerontologie en Geriatrie, 50(4), 111. https://doi.org/10.36613/tgg.1875-6832/2019.04.01 Google ScholarPubMed
Friborg, O., Martinsen, E. W., Martinussen, M., Kaiser, S., Overgård, K. T., Rosenvinge, J. H. (2014). Comorbidity of personality disorders in mood disorders: a meta-analytic review of 122 studies from 1988 to 2010. Journal of Affective Disorders, 152–154, 111. https://doi.org/10.1016/j.jad.2013.08.023 CrossRefGoogle ScholarPubMed
Friborg, O., Martinussen, M., Kaiser, S., Overgard, K. T., & Rosenvinge, J. H. (2013). Comorbidity of personality disorders in anxiety disorders: a meta-analysis of 30 years of research. Journal of Affective Disorders, 145(2), 143155. https://doi.org/10.1016/j.jad.2012.07.004 CrossRefGoogle ScholarPubMed
Ganellen, R. J. (2007). Assessing normal and abnormal personality functioning: strengths and weaknesses of self-report, observer, and performance-based methods. Journal of Personality Assessment, 89(1), 3040. https://doi.org/10.1080/00223890701356987 CrossRefGoogle ScholarPubMed
Giavarina, D. (2015). Understanding Bland Altman analysis. Biochemical Medicine, 25, 141151. https://doi.org/10.11613/BM.2015.015 CrossRefGoogle ScholarPubMed
Gibson, R., & Ferrini, R. (2012). Difficult resident or personality disorder? A long-term care perspective. Annals of Long-Term Care: Clinical Care and Aging, 20, 2028.Google Scholar
Jamieson, H., Abey-Nesbit, R., Bergler, U., Keeling, S., Schluter, P. J., Scrase, R., Lacey, C. (2019). Evaluating the influence of social factors on aged residential care admission in a national home care assessment database of older adults. Journal of The American Medical Directors Association, 20(11), 14191424. https://doi.org/10.1016/j.jamda.2019.02.005 CrossRefGoogle Scholar
Knäuper, Bärbel, Carrière, K., Chamandy, M., Xu, Z., Schwarz, N., Rosen, N. O. (2016). How aging affects self-reports. European Journal of Ageing, 13(2), 185193. https://doi.org/10.1007/s10433-016-0369-0 CrossRefGoogle ScholarPubMed
Koo, T. K., & Li, M. Y. (2016). A guideline of selecting and reporting intraclass correlation coefficients for reliability research. Journal of Chiropractic Medicine, 15(2), 155163. https://doi.org/10.1016/j.jcm.2016.02.012 CrossRefGoogle ScholarPubMed
Koopmans, R., Leerink, B., & Festen, D.A.M. (2022). Dutch long-term care in transition: a guide for other countries. Journal of The American Medical Directors Association, 23(2), 204206. https://doi.org/10.1016/j.jamda.2021.09.013 CrossRefGoogle ScholarPubMed
Koopmans, R. T., Lavrijsen, J. C., Hoek, J. F., Went, P. B., & Schols, J. M. (2010). Dutch elderly care physician: a new generation of nursing home physician specialists. Journal of The American Geriatrics Society, 58(9), 18071809. https://doi.org/10.1111/j.1532-5415.2010.03043.x CrossRefGoogle ScholarPubMed
Leising, D., Erbs, J., & Fritz, U. (2010). The letter of recommendation effect in informant ratings of personality. Journal of Personality and Social Psychology, 98(4), 668682. https://doi.org/10.1037/a0018771 CrossRefGoogle ScholarPubMed
Millon, T. E., & GS 1985). Personality and its disorders: a biosocial learning approach. John Wiley & Sons.Google Scholar
Molloy, D. W., Alemayehu, E., & Roberts, R. (1991). Reliability of a Standardized Mini-Mental State Examination compared with the traditional Mini-Mental State Examination. American Journal of Psychiatry, 148(1), 102105. https://doi.org/10.1176/ajp.148.1.102 Google ScholarPubMed
Mordekar, A., & Spence, S. A. (2008). Personality disorder in older people: how common is it and what can be done? Advances in Psychiatric Treatment, 14(1), 7177. https://doi.org/10.1192/apt.bp.107.003897 CrossRefGoogle Scholar
Morse, J. Q., Pilkonis, P. A., Houck, P. R., Frank, E., & Reynolds, C. F. (2005). Impact of cluster C personality disorders on outcomes of acute and maintenance treatment in late-life depression. American Journal of Geriatric Psychiatry, 13(9), 808814. https://doi.org/10.1176/appi.ajgp.13.9.808 CrossRefGoogle ScholarPubMed
Nakagawa, S., & Schielzeth, H. (2010). Repeatability for Gaussian and non-Gaussian data: a practical guide for biologists. Biological Reviews: Cambridge Philosophical Society, 85(4), 935956. https://doi.org/10.1111/j.1469-185X.2010.00141.x CrossRefGoogle Scholar
Nakagawa, S., & Schielzeth, H. (2013). A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution, 4(2), 133142. https://doi.org/10.1111/j.2041-210x.2012.00261.x CrossRefGoogle Scholar
Nienhuis, F. J., Van De Willige, G., Rijnders, C. A., De Jonge, P., & Wiersma, D. (2010). Validity of a short clinical interview for psychiatric diagnosis: the mini-SCAN. British Journal of Psychiatry, 196(1), 6468. https://doi.org/10.1192/bjp.bp.109.066563 CrossRefGoogle Scholar
Oltmanns, T. F., & Balsis, S. (2011). Personality disorders in later life: questions about the measurement, course, and impact of disorders. Annual Review of Clinical Psychology, 7(1), 321349. https://doi.org/10.1146/annurev-clinpsy-090310-120435 CrossRefGoogle ScholarPubMed
Penders, K.a P., Peeters, I. G. P., Metsemakers, J. F. M., & Van Alphen, S. P. J. (2020). Personality disorders in older adults: a review of epidemiology, assessment, and treatment. Current Psychiatry Reports, 22(3), 14. https://doi.org/10.1007/s11920-020-1133-x CrossRefGoogle ScholarPubMed
Pleil, J. D., Wallace, M.a G., Stiegel, M. A., & Funk, W. E. (2018). Human biomarker interpretation: the importance of intra-class correlation coefficients (ICC) and their calculations based on mixed models, ANOVA, and variance estimates. Journal of Toxicology and Environmental Health, Part B, 21(3), 161180. https://doi.org/10.1080/10937404.2018.1490128 CrossRefGoogle ScholarPubMed
Romirowsky, A., Zweig, R., Glick Baker, L., & Sirey, J. A. (2021). The relationship between maladaptive personality and social role impairment in depressed older adults in primary care. Clinical Gerontologist, 44(2), 192205. https://doi.org/10.1080/07317115.2018.1536687 CrossRefGoogle ScholarPubMed
Slachevsky, A., Villalpando, J. M., Sarazin, M., Hahn-Barma, V., Pillon, B., Dubois, B. (2004). Frontal assessment battery and differential diagnosis of frontotemporal dementia and Alzheimer disease. Archives of Neurology, 61(7), 11041107. https://doi.org/10.1001/archneur.61.7.1104 CrossRefGoogle ScholarPubMed
Snijders, T.a B., & Bosker, R. J. (2011). Multilevel analysis: an introduction to basic and advanced multilevel modeling. Sage.Google Scholar
Son, H., Friedmann, E., & Thomas, S. A. (2012). Application of pattern mixture models to address missing data in longitudinal data analysis using SPSS. Nursing Research, 61(3), 195203. https://doi.org/10.1097/NNR.0b013e3182541d8c CrossRefGoogle ScholarPubMed
Stek, M. L., Van Exel, E., Van Tilburg, W., Westendorp, R. G., & Beekman, A. T. (2002). The prognosis of depression in old age: outcome six to eight years after clinical treatment. Aging & Mental Health, 6(3), 282285. https://doi.org/10.1080/13607860220142413 CrossRefGoogle ScholarPubMed
Stevenson, J., Brodaty, H., Boyce, P., & Byth, K. (2011). Personality disorder comorbidity and outcome: comparison of three age groups. Australian & New Zealand Journal of Psychiatry, 45(9), 771779. https://doi.org/10.3109/00048674.2011.595685 CrossRefGoogle ScholarPubMed
Van Alphen, S. P., Rossi, G., Dierckx, E., & Oude Voshaar, R. C. (2014). [DSM-5 classification of personality disorders in older persons]. Tijdschrift voor Psychiatrie, 56(12), 816820.Google ScholarPubMed
Van Alphen, S. P. J., Oude Voshaar, R. C., Bouckaert, F., & Videler, A. C. (2018). Handboek persoonlijkheidsstoornissen bij ouderen. De Tijdstroom.Google Scholar
Van Den Brink, A. (2019). Design of the MAPPING study (chapter 4). Thesis ‘Nursing home residents with mental and physical multimorbidity - characteristics, neuropsychiatric symptoms and needs’ (pp. 5772). Radboud University. https://www.ukonnetwerk.nl/media/1142/anne-vd-brink-proefschrift.pdf.Google Scholar
Van Den Brink, A. M., Gerritsen, D. L., Oude Voshaar, R. C., & Koopmans, R. T. (2013). Residents with mental-physical multimorbidity living in long-term care facilities: prevalence and characteristics. A systematic review. International Psychogeriatrics, 25(4), 531548. https://doi.org/10.1017/S1041610212002025 CrossRefGoogle ScholarPubMed
Van Den Brink, A. M. A., Gerritsen, D. L., de Valk, M. M. H., Mulder, A. T., Oude Voshaar, R. C., Koopmans, R. T. C. M. (2018). What do nursing home residents with mental-physical multimorbidity need and who actually knows this? A cross-sectional cohort study. International Journal of Nursing Studies, 81, 8997. https://doi.org/10.1016/j.ijnurstu.2018.02.008 CrossRefGoogle Scholar
Van Den Brink, A. M. A., Gerritsen, D. L., De Valk, M. M. H., Oude Voshaar, R. C., & Koopmans, R. (2017). Characteristics and health conditions of a group of nursing home patients with mental-physical multimorbidity - the MAPPING study. International Psychogeriatrics, 29(6), 10371047. https://doi.org/10.1017/S1041610217000230 CrossRefGoogle ScholarPubMed
Veerbeek, M. A., Oude Voshaar, R. C., & Pot, A. M. (2014). Effectiveness and predictors of outcome in routine out-patient mental health care for older adults. International Psychogeriatrics, 26, 15651574. https://doi.org/10.1017/S1041610214000647 CrossRefGoogle Scholar
Watson, P. F., & Petrie, A. (2010). Method agreement analysis: a review of correct methodology. Theriogenology, 73(9), 11671179. https://doi.org/10.1016/j.theriogenology.2010.01.003 CrossRefGoogle ScholarPubMed
Figure 0

Table 1. Demographic and clinical characteristics: comparing participants with a completed versus a missing HAP questionnaire

Figure 1

Table 2. HAP results: trait scores and frequencies of maladaptive traits

Figure 2

Figure 1. Stacked histogram: norm referenced HAP results.Note: norm referenced = compared to norm scores of somatic and psychogeriatric nursing home residents, as provided by the HAP questionnaire manual; “very high” = percentile score ≥96th; “high” = percentile score 86th–95th; “above average” = percentile score 66th–85th; “average” = percentile score 36th–65th; “below average” = percentile score 16th–35th; “low” = percentile score ≤15th.

Figure 3

Figure 2. Bland and Altman plots: HAP-HAP-t differences for PERF and UNC.Note: relative scores (0–10); x = (HAP + HAP-t)/2, y = HAP – HAP-t; LoA (limits of agreement) = mean difference ± (1.96*standard deviation); 95% CI (95% confidence intervals) = ± (standard error*t value for degrees of freedom).

Figure 4

Table 3. Linear mixed model results: estimated differences and ICCs of HAP and HAP-t traits

Supplementary material: File

Suntjens et al. supplementary material

Suntjens et al. supplementary material
Download Suntjens et al. supplementary material(File)
File 18.6 KB