The relationship between perfectionistic self-presentation and reactions to impairment and disability following spinal cord injury

Abstract

Univariate and multivariate relationships between perfectionistic self-presentation and reactions to impairment and disability following spinal cord injury were examined. A total of 144 adults with spinal cord injury (M = 48.18 years old, SD = 15.96) completed self-report measures. Analyses revealed that, after controlling for time since injury and gender, perfectionistic self-presentation predicted six of eight reactions, shock, depression and internalised anger particularly strongly. In addition, at multivariate level, perfectionistic self-presentation was positively related to non-adaptive reactions and negatively related to adaptive reactions. The findings suggest that perfectionistic self-presentation may contribute to poorer psychosocial adaptation to spinal cord injury.

Keywords

chronic illness correlation disability health psychology psychological distress quantitative methods

The World Health Organization (2013) has estimated that between 250,000 and 500,000 people suffer a spinal cord injury (SCI) every year. SCI entails physical damage to the spinal cord and paralysis. As a result of paralysis, individuals can experience poor or complete loss of motor control, impaired sensory awareness and impaired homeostatic regulation. These permanent impairments require individuals to adapt to profound changes to their lives. This includes obvious altered physical capabilities but also changed social, sexual and vocational roles (Middleton et al., 2003). Individuals display considerable variability in the degree to which they are able to adapt to SCI. It has been found, for example, that up to 30 per cent of people who suffer SCI report clinically significant levels of depression (Craig et al., 2009). In this study, we sought to better understand factors that may undermine adaptation to SCI by examining the relationship between perfectionistic self-presentation (PSP) and reactions to impairment and disability following the injury.

Psychosocial adaptation and SCI

There is considerable debate regarding the conceptualisation of the psychosocial adaptation that follows SCI and other chronic illnesses and disabilities (CID) (Livneh and Wilson, 2003). There is, however, broad agreement that adaptation is best characterised as a process whereby a person moves from a state of conflict between current and desired self-concept towards a new state of acceptance (Livneh and Parker, 2005). Livneh and Antonak (1997) define psychosocial adaptation as ‘… an evolving, dynamic, general process through which the individual gradually approaches an optimal state of person-environment congruence’ (pp.8). In Livneh’s (2001) corresponding model of psychosocial adaptation, he differentiates between three parts: (1) events that can be attributed directly or indirectly to how the CID started and biological, psychological and contextual variables active at the onset of the CID; (2) the relationship between CID triggered reactions (or reactions to impairment and disability) and contextual variables (disability, sociodemographic, personality and external factors); and (3) the degree to which a person successfully adapts to CID in terms of real-world functioning. This model provides a broad framework in which the complex process of adaptation to CID can be understood.

In this study, we focus on reactions to impairment and disability. Reactions to impairment and disability refer to overlapping response phases during the adaptation process (Livneh and Wilson, 2003). Livneh’s model (Livneh and Antonak, 1990, 1997) includes eight reactions. Listed in decreasing proximity to the onset of CID, these reactions are shock (an emergency reaction to the CID onset), anxiety (a panic-stricken reaction upon initial realisation of the crisis), denial (the first active defence mobilisation against early realisation of the situation via wishful and unrealistic ideas about recovery), depression (the realisation of true loss and future limitations), internalised anger (self-directed resentment and bitterness often associated with feelings of self-blame), externalised hostility (retaliation against functional limitations directed at other persons, objects or aspects of the environment via aggressive acts and verbalisations), acknowledgement (a cognitive acceptance of the conditions permanency and the future implications) and adjustment (both an affective internalisation, or emotional acceptance, of new functional limitations and a sociobehavioural reintegration into new life situations). The last two reactions are considered adaptive and the other six non-adaptive with the caveat that denial may be independent from both sets of reactions (Livneh and Antonak, 1990, 1997).

Adaptation to SCI has received substantial attention in research. The work of Craig, Kennedy, Krause and Tate, among others, is particularly noteworthy (see Chevalier et al., 2009; Craig et al., 2009; Post and Van Leeuwen, 2012; for reviews). Their research has provided extensive information on the experiences of those with SCI and suggests that most people are able to adapt well to their changed lives. However, the findings of this research also suggest that individual trajectories can vary in the adaptation process and a significant minority of individuals experience mental health difficulties following SCI. In addition, successful adaptation to SCI is influenced by various factors and includes contextual disability-related variables (e.g. time since injury), sociodemographic variables (e.g. gender), situational appraisals and coping processes (e.g. perceptions of threat vs challenge) and personality factors (e.g. neuroticism and extraversion). This research also includes impressive longitudinal studies that have tracked those with SCI over decades of their lives and studies that indicate psychosocial adaptation to SCI can be improved via intervention (see Mehta et al., 2011).

Against this backdrop, there have been a smaller number of studies that have directly examined reactions to impairment and disability following SCI. The findings of this research have illustrated how reactions to impairment and disability discriminate between those experiencing a SCI in terms of its onset (recent onset vs long-term injury) as well as between different groups (civilians vs veterans) (e.g. Livneh and Antonak, 1990; Livneh and Martz, 2003). In addition, there is also evidence that reactions to impairment and disability are related to individual differences in a sense of coherence (belief that the world is comprehensible, manageable and meaningful), a future time orientation (degree of general concern, engagement and involvement in the future) and hope (a sense of successful agency and pathways to meaningful goals) (Livneh and Martz, 2014; Lustig, 2005; Martz, 2004). Research therefore provides some support for Livneh’s (2001) model and the relationship between reactions to impairment and disability and personality characteristics, in particular.

PSP and CID

One personality factor that may be important in terms of reactions to impairment and disability is perfectionism. Perfectionism is broadly defined as a commitment to excessively high standards and overly critical self-evaluation (Frost et al., 1990). It is thought to be a personality characteristic that can manifest as a trait, cognitive style and presentational style (Flett et al., 1998; Hewitt et al., 2003; Hewitt and Flett, 1991). In terms of the latter, PSP is a style of impression management that includes the facets of perfectionistic self-promotion (seeking opportunities to demonstrate one’s perfection), nondisplay of imperfection (minimising the public display of mistakes, flaws and shortcomings) and nondisclosure of imperfection (minimising admission of mistakes, flaws and shortcomings). PSP is thought to stem from a desire to protect and boost low and fragile self-esteem (Hewitt et al., 2003). In accord, research has found that PSP is related to a range of maladaptive outcomes including a lower sense of being important to others (Flett et al., 2012), depression (Flett et al., 2014) and suicide ideation (Roxborough et al., 2012).

Researchers have recently begun to examine the role of perfectionism in health. This includes examining whether perfectionism predicts the experiences of those with various CID (e.g. Flett et al., 2011; Molnar et al., 2012; Shanmugasegaram et al., 2014). This research has almost exclusively focused on trait perfectionism. However, more recently, PSP has begun to receive attention. Of the two studies to date, the first examined the relationship between PSP, other personality characteristics (trait perfectionism, optimism and neuroticism), health-related coping and the physical and psychosocial impact of CID among individuals with irritable bowel syndrome (Flett et al., 2011). The second examined the relationship between PSP, other personality characteristics (trait perfectionism and type D personality) and health-related coping among cardiac rehabilitation patients (Shanmugasegaram et al., 2014). In both studies, PSP was related to coping strategies considered less conducive to psychosocial adaptation (e.g. emotional preoccupation). In addition, in Flett et al.’s study, PSP was also related to greater negative psychosocial impact of illness and remained so after controlling for the physical impact of the illness, neuroticism and optimism. As such, initial indication is that PSP is associated with how people cope with and adapt to CID.

Examining PSP and SCI offers a further opportunity to explore the nature of PSP within context of CID. Flett et al. (2011) have argued that perfectionism may exacerbate the impact of CID for a number of reasons. Perfectionism entails negative cognitive biases that exacerbate stressful situations (stress reactivity) and also makes it difficult to adjust achievement behaviour following illness (stress generation). In both of these regards, PSP would be expected to be related to less adaptive and more non-adaptive reactions to SCI. The irrational desire to portray life as ideal and to be ‘perfectly healthy’, discussed by Flett et al. in context of irritable bowel syndrome, is especially relevant to SCI. This is because the visible nature of SCI (e.g. use of a wheelchair), ongoing health complications and requirement of self-care make depicting an image of being ‘perfectly healthy’ impossible. In addition, PSP is thought to be underpinned by a highly negative self-view. Therefore, in a similar manner to how trait perfectionism is related to a pessimistic attributional style (i.e. the tendency to attribute negative events to internal, stable and global factors; Chang and Sanna, 2001), it is possible that SCI may be viewed as affirmation of some inherent lack of importance, worth or value, and worsen reactions to SCI. Finally, personal and professional goals will be fundamentally challenged following SCI and those exhibiting PSP are likely to lack the flexibility to reassess and adjust these goals and may even have particular difficultly accepting that this is required.

Method

Participants

A total of 144 adults (aged 18 years or older) with SCI participated in the study. The participants were mostly male (75%), middle-aged (M = 48.18 years, SD = 15.96, range of 18–78), White British (88%),¹ married (48%) and paraplegic (57%). The total time since injury ranged from 1 to 660 months (M = 148.00, SD = 165.97). Participants were recruited from hospital (n = 48) and a community (n = 96) settings in the United Kingdom. Recruitment in a hospital setting was coordinated with medical care professionals who distributed information sheets. If patients expressed an interest in participation, they then met the first author to discuss participation. Recruitment in a community setting was via targeted advertisement with wheelchair user groups or SCI groups (e.g. mouth painting classes, online support forums and SCI charities). All participants provided informed consent prior to taking part. Thereafter, participants completed a multi-section questionnaire. Questionnaires were completed either independently or with the assistance of a researcher (first author) where necessary. Ethical approval was obtained from both the institution of the lead researcher and the National Health Service.

Instruments

PSP

PSP was measured using the Perfectionistic Self-Presentation Scale (PSPS; Hewitt et al., 2003). The PSPS comprises 27 items and 3 subscales, namely, perfectionistic self-promotion (10 items; e.g., ‘If I seem perfect, others will see me more positively’), nondisplay of imperfection (10 items; e.g., ‘Errors are much worse if they are made in public rather than in private’) and nondisclosure of imperfection (7 items; e.g., ‘I should always keep my problems to myself’). These are measured on a 7-point scale ranging from 1 (Strongly Disagree) to 7 (Strongly Agree). Evidence has been provided by Hewitt et al. (2003) to support the validity and reliability of the instrument including factor structure and internal reliability. The PSPS has also been used in research among individuals with CID (Flett et al., 2011; Shanmugasegaram et al., 2014).

Reactions to impairment and disability

Reactions to impairment and disability were measured using the Reaction to Impairment and Disability Inventory (RIDI) (Livneh and Antonak, 1990, 2008). The RIDI consists of 60 items and 8 subscales: shock (7 items; e.g., ‘my mind goes blank’), anxiety (8 items; e.g., ‘I find myself trembling without any reason’), denial (7 items; e.g., ‘if I become a better person my problems will be cured’), depression (8 items; e.g., ‘I am a failure as a person’), internalised anger (8 items; e.g, ‘I am embarrassed about my impairment’), externalised hostility (7 items; e.g., ‘nobody is going to tell me what to do’), acknowledgement (7 items; e.g., ‘I have been through a crisis and feel that I understand things better’) and adjustment (8 items; e.g., ‘I am satisfied with my present abilities despite my disability’). The instructions of the instrument were altered so to focus respondents on SCI. Participants respond by identifying the frequency with which they have experienced each item on a 4-point scale from 1 (Never) to 4 (Often). Evidence has been provided by Livneh and colleagues (Livneh and Antonak, 1990, 2005; Livneh et al., 2006) to support the validity and reliability of the instrument including factor structure and internal reliability. The use of the RIDI is also common in research examining adaptation to SCI (e.g. Livneh and Martz, 2014; Lustig, 2005; Martz, 2004).

Results

Preliminary analysis

Prior to conducting the primary analyses, data were assessed in terms of missing data and screened for univariate and multivariate outliers using procedures described by Tabachnick and Fidell (2007). Due to large amounts of missing data (>5%), one participant was removed. There were 120 remaining cases with complete data and 23 cases with incomplete cases. Incomplete cases displayed a very low amount of missing items (M = 1.57 items, SD = 0.90, range = 1–4). In addition, Little’s missing completely at random (MCAR) test provided evidence that data were missing completely at random: χ² (1793) = 1866.27, p > .05. Missing values were subsequently replaced using the mean of the non-missing items from the subscale in each individual case (see Graham et al., 2003).

Three participants were removed due to being identified as univariate outliers (z score ±3.29, p < .001, two-tailed). No multivariate cases were removed: Mahalanobis distance used for assessment was χ² (11) = 31.264, p < .001. Subsequently, the final sample consisted of 140 participants (47 inpatients, 93 outpatients) who were predominantly male (74%), White British (89.3%), paraplegic (56.4%) with an age range of 18–78 years (M = 48.03, SD = 15.92), married (47.9%) and a time since onset of injury from 1 to 660 months (M = 147.78, SD = 167.71). Following removal of outliers, a number of variables were also transformed so that they more closely approximated a normal distribution (shock, anxiety, denial, depression, and externalised hostility, all LOG10 transformations). The transformed variables were almost perfectly correlated with the corresponding original variables (r = .98–.99). In addition, bivariate correlations and multiple regression analyses included standard errors and hypothesis tests based on 95 per cent bias-corrected and accelerated (BCa) bootstrap estimates (1000 resamples).

Finally, each instrument was assessed for its internal reliability (Cronbach’s α). These are displayed in Table 1. Acknowledgement and denial displayed lower levels of internal reliability (<.70). Rather than not analyse these subscales further, we retained them for exploratory purposes. The findings concerning these particular subscales should therefore be interpreted cautiously.

Table 1.

Descriptive statistics, internal reliabilities and bivariate correlations for demographic variables, perfectionistic self-presentation and reactions to impairment and disability inventory.

Variable	M	SD	α	1	2	3	4	5	6	7	8	9	10	11	12
1. Perfectionistic self-promotion	38.54	9.98	.82	–
2. Nondisplay of imperfection	36.98	10.97	.86	.63**	–
3. Nondisclosure of imperfection	23.48	7.84	.76	.50**	.61**	–
4. Shock	11.63	4.20	.78	.30**	.51**	.37**	–
5. Anxiety	13.36	4.90	.83	.22*	.34**	.26**	.76**	–
6. Denial	9.80	3.11	.62	.10	.03	.18*	.07	.03	–
7. Depression	14.17	4.98	.83	.30**	.52**	.43**	.68**	.64**	−.01	–
8. Internalised anger	13.83	4.49	.73	.28**	.49**	.44**	.61**	.51**	.02	.73**	–
9. Externalised hostility	11.71	3.51	.69	.12	.28**	.23**	.50**	.51**	.01	.55**	.49**	–
10. Acknowledgement	20.98	3.61	.59	.09	−.13	−.06	.09	.00	.11	−.13	.00	−.13	–
11. Adjustment	25.36	4.66	.76	−.17**	−.35**	−.28**	−.41**	−.37**	.15	−.55**	−.52**	−.21**	.35**	–
12. Gender	–	–	–	.05	.05	−.08	.23*	.24**	−.13	.16	.14	.07	.11	−.12
13. Time since injury	147.78	167.71	–	−.22*	−.20*	−.18*	−.20*	−.09	−.12	−.10	−.17	.09	−.16	.26**	−.10

Time since injury is measured in months. Gender coded as 1 male and 2 female. Means/SD total scores are untransformed. Bivariate correlations include transformed variables.

p < .05, two-tailed; **p < .01, two-tailed.

Descriptive statistics and bivariate correlations

Table 1 displays descriptive statistics and bivariate correlations. The participants reported moderate levels of PSP, low levels of non-adaptive reactions and high levels of adaptive reactions. We note that there were no discernible differences in perfectionistic self-promotion and nondisclosure of imperfection scores in our sample compared to previous student, community and clinical samples (see Hewitt et al., 2003). However, nondisplay of imperfection was slightly lower in our sample than these previous samples. In terms of the RIDI, scores were generally similar to other samples of veterans and civilians with spinal cord injuries with slightly lower shock and denial evident in our sample (see Livneh and Antonak, 2008). Examination of the bivariate correlations revealed that perfectionistic self-promotion displayed medium significant positive relationships with four non-adaptive reactions (shock, anxiety, depression and internalised anger), was unrelated to two others (denial and externalised anger) and had a small-to-medium significant negative relationship with one adaptive reaction (adjustment). Nondisplay of imperfection and nondisclosure of imperfection displayed a very similar pattern of relationships to perfectionistic self-promotion though in the case of nondisplay of imperfection these were typically larger (i.e. medium-to-large). In addition, these two facets were also related to an additional non-adaptive reaction (externalised hostility).

Hierarchical multiple regression analyses

To examine the univariate predictive ability of PSP for each reaction, a series of hierarchical multiple regression analyses were conducted. In each analysis, step 1 was composed of gender and time since injury. These two variables were included so to control for the gender differences and variability in the time since onset of injury in the sample (no other control variables were included). In step 2, the three facets of PSP were added. Overall variance explained in each step was examined, as was incremental predictive ability from steps 1 and 2. The results of the analyses are reported in Table 2.

Table 2.

Hierarchical multiple regression analyses of perfectionistic self-presentation and reactions to impairment and disability.

Model	B	SE	Lower CI	Higher CI	β
Shock
Step 1: F (2, 125) = 5.61, p < .01, R² = .08
Gender	0.08	.04	0.00	0.15	.21*
Time since injury	−0.00	.00	0.00	−0.00	−.18*
Step 2: F (5, 122) = 11.54, p < .001, R² = .32, ΔR² = .24, p < .001
Gender	0.08	.03	0.00	0.14	.21*
Time since injury	−0.00	.00	0.00	0.01	−.08
Perfectionistic self-promotion	−0.01	.01	−0.04	0.01	−.08
Nondisplay of imperfection	0.06	.01	0.03	0.08	.46***
Nondisclosure of imperfection	0.18	.01	−0.01	0.45	.13
Anxiety
Step 1: F (2, 125) = 4.21, p < .05, R² = .06
Gender	0.08	.03	0.02	0.15	.24*
Time since injury	−0.00	.00	0.00	0.00	−.07
Step 2: F (5, 122) = 5.34, p < .001, R² = .18, ΔR² = .11, p < .01
Gender	0.85	.98	−0.19	0.15	.24**
Time since injury	0.00	.03	0.02	0.00	.01
Perfectionistic self-promotion	−0.00	.01	0.00	0.02	−.02
Nondisplay of imperfection	0.03	.01	0.00	0.06	.27*
Nondisclosure of imperfection	0.02	.01	−0.01	0.05	.13
Denial
Step 1: F (2, 125) = 2.15, p > .05, R² = .03
Gender	−0.04	.02	−0.85	0.01	−.14
Time since injury	−0.00	.00	0.00	0.00	−.13
Step 2: F (5, 122) = 1.83, p > .05, R² = .07, ΔR² = .04, p > .05
Gender	−0.03	.02	−0.07	0.00	−.12
Time since injury	−0.00	.00	0.00	0.00	−.11
Perfectionistic self-promotion	0.01	.02	−0.02	0.05	.10
Nondisplay of imperfection	−0.02	.01	−0.05	0.01	−.18
Nondisclosure of imperfection	0.02	.02	−0.01	0.05	.21
Depression
Step 1: F (2, 125) = 2.22, p > .05, R² = .03
Gender	0.06	.03	−0.01	0.13	.16
Time since injury	−0.00	.00	0.00	0.00	−.09
Step 2: F (5, 122) = 11.68, p < .05, R² = .32, ΔR² = .29, p < .05
Gender	0.06	.03	0.00	0.12	−.17*
Time since injury	0.00	.00	0.00	0.00	.02
Perfectionistic self-promotion	−0.01	.02	−0.05	0.02	−.10
Nondisplay of imperfection	0.05	.01	0.03	0.08	.44***
Nondisclosure of imperfection	0.03	.02	0.00	0.06	.23*
Internalised anger
Step 1: F (2, 125) = 2.82, p > .05, R² = .04
Gender	0.17	.11	−0.01	0.41	.12
Time since injury	−0.00	.00	0.00	0.00	−.16
Step 2: F (5, 122) = 10.27, p < .05, R² = .30, ΔR² = .25, p < .001
Gender	0.19	.01	0.00	0.40	−.14
Time since injury	0.00	.00	0.00	0.00	.02
Perfectionistic self-promotion	−0.05	.05	−0.14	0.05	−.11
Nondisplay of imperfection	0.17	.05	0.08	0.26	.38**
Nondisclosure of imperfection	0.13	.06	0.01	0.25	.25*
External hostility
Step 1: F (2, 125) = 0.89, p > .05, R² = .01
Gender	0.02	.03	−0.03	0.08	.07
Time since injury	0.00	.00	−0.00	0.00	.10
Step 2: F (5, 122) = 3.52, p < .05, R² = .13, ΔR² = .11, p < .01
Gender	0.03	.03	−0.03	0.08	.09
Time since injury	0.00	.00	0.00	0.00	.16
Perfectionistic self-promotion	−0.02	.02	−0.05	0.01	−.15
Nondisplay of imperfection	0.03	.01	0.00	0.05	.30*
Nondisclosure of imperfection	0.02	.02	−0.01	0.05	.16
Acknowledgement
Step 1: F (2, 125) = 2.11, p > .05, R² = .03
Gender	0.12	.11	−0.11	0.35	.09
Time since injury	0.00	.00	−0.00	0.00	−.15
Step 2: F (5, 122) = 2.42, p < .05, R² = .09, ΔR² = .06, p = .06
Gender	0.12	.11	−0.11	0.35	.10
Time since injury	0.00	.00	−0.00	0.00	−.16
Perfectionistic self-promotion	0.11	.06	0.07	0.23	.23
Nondisplay of imperfection	−0.13	.05	−0.25	−0.03	−.30*
Nondisclosure of imperfection	−0.00	.05	−0.11	0.10	−.01
Adjustment
Step 1: F (2, 125) = 5.06, p < .01, R² = .08
Gender	−0.13	.11	−0.33	0.07	−.10
Time since injury	0.00	.00	−0.00	0.00	.25
Step 2: F (5, 122) = 5.50, p < .001, R² = .18, ΔR² = .11, p < .01
Gender	−0.14	.10	−0.33	0.05	−.10
Time since injury	0.00	.00	0.00	0.00	.19*
Perfectionistic self-promotion	0.07	.08	−0.06	0.22	.14
Nondisplay of imperfection	−0.15	.05	−0.24	−0.05	−.32**
Nondisclosure of imperfection	−0.07	.06	−0.19	0.04	−.13

SE: standard error; CI: confidence interval.

SE and CI based on bias-corrected and accelerated (BCa) bootstrap estimates.

p < .05, **p < .01, ***p < .001, two-tailed.

The analyses revealed that the overall models explained 7 to 32 per cent of variance in the eight reactions. Step 1 accounted for 1 to 8 per cent of variance. Step 2 accounted for a further 4 to 29 per cent of variance. On six occasions, the incremental predictive ability in step 2 was statistically significant: shock, anxiety, depression, internalised anger, externalised hostility and adjustment (also marginally for acknowledgement, p = .06). In these models, in terms of individual predictors of each reaction, nondisplay of imperfection was a medium-to-large significant positive unique predictor of all six reactions. Nondisclosure of imperfection was a small-to-medium significant positive unique predictor of two reactions (depression and internalised anger). Finally, perfectionistic self-promotion was not a unique predictor of any reactions.

Canonical correlation analyses

To examine the multivariate relationships between PSP and reactions to CID, two canonical correlation analyses were conducted. These were conducted following the guidelines provided by Tabachnick and Fidell (2007). In the analyses, facets of PSP were included as a set of manifest variables (to load on a new latent variable) and reactions (non-adaptive and adaptive) were used as another set of manifest variables (to load on two other new latent variables). The multivariate relationship is assessed by examining the relationship between manifest variables and the latent variable (structure coefficients, r_s, above .30 were considered meaningful) and the relationship between the two latent variables (a canonical function and corresponding canonical correlation, R_C, with R_C above .30 considered meaningful). The results of the analyses are reported in Table 3.

Table 3.

Canonical correlation between perfectionistic self-presentation, non-adaptive and adaptive reactions to impairment and disability.

Variables	Non-adaptive reactions			Adaptive reactions
	Function			Function
	R_s	R_S² (%)		R_s	R_S² (%)
Perfectionistic self-promotion	−.56	.31	Perfectionistic self-promotion	−.52	.27
Nondisplay of imperfection	−.97	.94	Nondisplay of imperfection	−.95	.90
Nondisclosure of imperfection	−.78	.61	Nondisclosure of imperfection	−.81	.66
Adequacy		.62	Adequacy		.60
Redundancy		.23	Redundancy		.11
Shock	−.84	.71	Acknowledgement	.42	.18
Anxiety	−.56	.31	Adjustment	.99	.98
Depression	−.88	.77
Internalised anger	−.88	.77
Externalised hostility	−.55	.30
Adequacy		.57	Adequacy		.59
Redundancy		.21	Redundancy		.11
Canonical correlation R_C		.60			.43
R _C ²		.36			.18

PSP and non-adaptive reactions

Multivariate test of significance revealed a significant model: Wilks’ λ .60, F (15,364) = 4.95, p < .001. One meaningful canonical function and corresponding correlation emerged (R_C = .60). The r_s values indicated that each facet of PSP loaded meaningfully (>.30) on the latent variable (r_s = −.56 to −.97). Based on these loadings, this was considered to be reflective of (opposing) PSP. The r_s values indicated that the all five reactions also loaded meaningfully on the other latent variable (r_s = −.55 and −.88). Based on these loadings, this was considered to be reflective of (opposing) non-adaptive reactions. Latent (opposing) PSP explained an average of 62 per cent of variance in the manifest facets and latent (opposing) non-adaptive reactions explained an average of 57 per cent of variance in the five manifest reactions. Overall, the canonical correlation suggests that PSP is postively related to non-adaptive reactions.

PSP and adaptive reactions

Multivariate test of significance revealed a significant model: Wilks’ λ 0.79, F (6, 270) = 5.47, p < .001. One meaningful canonical function and corresponding correlation emerged (R_C = .43). The r_s values indicated that each facet of PSP loaded meaningfully (>.30) on the latent variable (r_s = −.52 to −.95). Based on these loadings, this was considered to be reflective of (opposing) PSP. The r_s values indicated that the two reactions also loaded meaningfully on the other latent variable (r_s = .42 and .99). Based on these loadings, this was considered to be reflective of adaptive reactions. Latent (opposing) PSP explained an average of 60 per cent of variance in the manifest facets and latent adaptive reactions explained an average of 59 per cent of variance in the two manifest reactions. Overall, the canonical correlation suggests that PSP is negatively related to adaptive reactions.

Discussion

The aim of this study was to examine the relationship between PSP and reactions to impairment and disability following SCI. It was hypothesised that facets of PSP would positively predict non-adaptive reactions (shock, anxiety, denial, depression, internalised anger and externalised hostility) and negatively predict adaptive reactions (acknowledgement and adjustment). These relationships were expected to be evident at univariate and multivariate levels. In support of the hypotheses, at univariate level, facets of PSP were significant positive predictors of shock, anxiety, depression, internalised anger and externalised hostility and a significant negative predictor of adjustment and, marginally, acknowledgement. However, contrary to the hypotheses, facets of PSP did not predict denial. Again in support of the hypotheses, at multivariate level PSP was positively related to non-adaptive reactions to impairment and disability and negatively related to adaptive reactions to impairment and disability.

The findings are consistent with previous work that has found PSP to be related to poorer psychosocial adaptation to CID (Flett et al., 2011; Shanmugasegaram et al., 2014). At a univariate level, it was evident that differentiating between the three facets of PSP is important in this relationship and that, even though highly correlated, not doing so may mean important unique effects might be missed. Here, nondisplay of imperfection was consistently the strongest predictor, or the only predictor, of each individual reaction. This was not the case in the two previous studies to examine PSP and psychosocial adaptation to CID (namely, irritable bowel syndrome and cardiac rehabilitation). However, this has been the case in studies examining other outcomes (e.g. self-esteem, anxiety and depression; Besser et al., 2010; Hewitt et al., 2003). This particular finding may therefore be because nondisplay of imperfection is especially important in context of SCI or, alternatively, this finding may reflect a general trend towards this facet of PSP being more problematic. Given that nondisplay of imperfection is underpinned by the desire for concealment, and that social stigma is an important issue in the lives of those with SCI (Hammell, 2007), in this instance we speculate it is more likely the former.

PSP was an especially large predictor of shock, depression and internalised anger (incremental predictive variance of 30% or higher). We speculate that a pattern of emotion regulation characterised by both expressive suppression (inhibiting the emotion-expressive behaviour; Gross, 1998) and emotion preoccupation (fixating on emotional consequences; Endler et al., 1998) may help explain these findings. The tendency to hide emotions predicts poorer well-being generally and has been found to predict poorer adaptation to CID. This is because hiding emotions is thought to leave emotions unresolved and denies individuals the opportunity of relief via sharing of experiences and bond building with others (De Ridder et al., 2008). We are not aware of any study that has examined the relationship between PSP and expressive suppression. However, trait perfectionism has been found to be positively related to this form of emotion regulation (e.g. Bergman et al., 2007), and it is likely this is the case for PSP. PSP has, however, been found to be positively related to emotional preoccupation in the context of CID and, again, this is associated with poor adaptation to CID (Flett et al., 2011; Shanmugasegaram et al., 2014). One can envisage how the habitual use of a combination of these two forms of emotion regulation may inhibit the proper processing of reactions such as shock and internal anger and how they may allow a sense of depression to develop and endure over time.

The only reaction not predicted by PSP was denial. It is not clear why this was the case. It is possible that levels of PSP have little influence on the experience of denial or that the relationship is more complex such as non-linear (e.g. curvilinear) or subject to moderating factors (e.g. time since injury). These possibilities have some support in existing research (e.g. Greenhouse et al., 2000). Disentangling the PSP-denial relationship is further complicated by the complexity of denial as possibly both an adaptive and non-adaptive reaction. For instance, denial may be useful in the short-term and ward of extreme feeling of anxiety and depression. However, if denial persists after these feelings have subsided, and rehabilitation is compromised as a consequence, denial will be unhelpful and even harmful (De Ridder et al., 2008). Unfortunately, the lower internal reliability and introduction of substantial measurement error ultimately means it is difficult to infer anything from the lack of findings involving this subscale. With this in mind, the current findings are inconclusive in regard to the relationship between PSP and denial.

The univariate findings were complemented by the multivariate findings. Specifically, PSP was related to poorer psychosocial adaptation (higher non-adaptive and lower adaptive reactions). Multivariate relationships have not been examined in previous research examining PSP and psychosocial adaptation to CID (Flett et al., 2011; Shanmugasegaram et al., 2014). We consider this to be an important oversight because of the benefits of multivariate models when examining complex inter-related outcomes. These models are more likely to closely match the reality of the experiences of those with CID. That is, the experience of a mix of multiple reactions that can be loosely arranged as more or less adaptive and non-adaptive and will help or hinder psychosocial adaptation. As illustrated here, facets of PSP uniquely predict a number of individual reactions that are likely to be problematic but, overall, PSP is associated with a collection of reactions that will almost certainly undermine psychosocial adaptation to SCI and other CID.

Clinical implications

One of the main difficulties that face those with SCI is the development of secondary conditions (e.g. cardiac arrhythmias, urinary tract infections and gastroparesis; Cardenas et al., 2004). It is estimated that as many as 95 per cent of people with SCI have a secondary condition and over half will be hospitalised as a consequence (Anson and Shepherd, 1996). Most secondary conditions can be avoided or managed by self-care (e.g. self-administered medication, nutrition and exercise). Those with higher nondisclosure and nondisplay of imperfection are likely to find engaging in proper self-care more difficult as they seek to portray a false image of good health. We suggest that assessing PSP may be useful to help identify individuals for whom this may be a problem. Thereafter, challenging and changing the irrational beliefs underpinning PSP may offer a means of promoting psychosocial adaptation to SCI and also help prevent the extreme psychological difficulties associated with PSP (e.g. depression and suicide ideation). We believe interventions aimed at managing and treating trait perfectionism, including flexible, non-intrusive self-help strategies, provide a useful starting point in this regard (see Lloyd et al., 2015) and will be especially effective if informed by existing programmes aimed at improving adaptation to SCI (see Mehta et al., 2011).

Limitations and other future directions

The findings must be considered in context of the limitations to the study. One limitation is that a cross-sectional design was employed. Therefore, causality or direction of causality cannot be inferred. It may be that reactions to impairment contribute to more or less PSP or the relationship is reciprocal. Future research should examine these possibilities using longitudinal designs. Relatedly, psychosocial adaptation and the reactions observed here are part of process that unfolds over time. Although time since injury was controlled statistically, the study can offer only a static snapshot of this dynamic process. Again, future research using longitudinal designs would be better able to capture this process. Another limitation is the reliance on self-report measures. This may contribute to common method variance and inflate the observed relationships. To avoid this, other-report measures (e.g. views of clinicians or relatives) may be useful in future research and also avoid any response bias or concealment on the part of the participants. Finally, two subscales of the RIDI (denial and acknowledgement) displayed lower levels of internal reliability. As noted earlier, findings involving these particular subscales should be interpreted cautiously and may require especial scrutiny in future work.

Conclusion

Research is beginning to emerge that suggests PSP is related to poorer psychosocial adaptation to CID. We found further evidence of this relationship in context of SCI. Specifically PSP, and nondisplay of imperfection, in particular, was related to less adaptive and more non-adaptive reactions to SCI.

Footnotes

Acknowledgements

The authors would like to acknowledge and thank Dr C. Glass, and Mr F. Jamil, and Mr R. Wajid for assistance with data collection.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors would like to acknowledge and thank the Wellcome Trust for providing funding that supported part of the data collection for this study.

Notes

References

Anson

Shepherd

(1996) Incidence of secondary complications in spinal cord injury. International Journal of Rehabilitation Research 19(1): 55–66.

Besser

Flett

Hewitt

(2010) Perfectionistic self-presentation and trait perfectionism in social problem-solving ability and depressive symptoms. Journal of Applied Social Psychology 40(8): 2121–2154.

Bergman

Nyland

Burns

(2007) Correlates with perfectionism and the utility of a dual process model. Personality and Individual Differences 43: 389–399.

Cardenas

Hoffman

Kirshblum

. (2004) Etiology and incidence of rehospitalization after traumatic spinal cord injury: A multicenter analysis. Archives of Physical Medicine and Rehabilitation 85(11): 1757–1763.

Chang

Sanna

(2001) Negative attributional style as a moderator of the link between perfectionism and depressive symptoms: Preliminary evidence for an integrative model. Journal of Counseling Psychology 48(4): 490–495.

Chevalier

Kennedy

Sherlock

(2009) Spinal cord injury, coping and psychological adjustment: A literature review. Spinal Cord 47(11): 778–782.

Craig

Tran

Middleton

(2009) Psychological morbidity and spinal cord injury: A systematic review. Spinal Cord 47(2): 108–114.

De Ridder

Geenen

Kuijer

. (2008) Psychological adjustment to chronic disease. The Lancet 372(9634): 246–255.

Endler

Parker

Summerfeldt

(1998) Coping with health problems: Developing a reliable and valid multidimensional measure. Psychological Assessment 10: 195–205.

10.

Flett

Baricza

Gupta

. (2011) Perfectionism, psychosocial impact and coping with irritable bowel disease: A study of patients with Crohn’s disease and ulcerative colitis. Journal of Health Psychology 16(4): 561–571.

11.

Flett

Besser

Hewitt

(2014) Perfectionism and interpersonal orientations in depression: An analysis of validation seeking and rejection sensitivity in a community sample of young adults. Psychiatry 77(1): 67–85.

12.

Flett

Galfi-Pechenkov

Molnar

. (2012) Perfectionism, mattering, and depression: A mediational analysis. Personality and Individual Differences 52(7): 828–832.

13.

Flett

Hewitt

Blankstein

. (1998) Psychological distress and the frequency of perfectionistic thinking. Journal of Personality and Social Psychology 75(5): 1363–1381.

14.

Frost

Marten

Lahart

. (1990) The dimensions of perfectionism. Cognitive Therapy and Research 14(5): 449–468.

15.

Greenhouse

Meyer

Johnson

(2000) Coping and medication adherence in bipolar disorder. Journal of Affective Disorders 59(3): 237–241.

16.

Graham

Cumsille

Elek-Fisk

(2003) Methods for handling missing data. In: Schinka

Velicer

(eds) Research Methods in Psychology. New York: Wiley, pp. 87–112.

17.

Gross

(1998) Antecedent- and response-focused emotion regulation: Divergent consequences for experience, expression, and physiology. Journal of Personality and Social Psychology 74: 224–237.

18.

Hammell

(2007) Quality of life after spinal cord injury: A meta-synthesis of qualitative findings. Spinal Cord 45(2): 124–139.

19.

Hewitt

Flett

Sherry

. (2003) The interpersonal expression of perfection: Perfectionistic self-presentation and psychological distress. Journal of Personality and Social Psychology 84(6): 1303–1325.

20.

Hewitt

Flett

(1991) Perfectionism in the self and social contexts: Conceptualization, assessment, and association with psychopathology. Journal of Personality and Social Psychology 60: 456–470.

21.

Livneh

(2001) Psychosocial adaptation to chronic illness and disability: A conceptual framework. Rehabilitation Counseling Bulletin 44(3): 151–160.

22.

Livneh

Antonak

(1990) Reactions to disability: An empirical investigation of their nature and structure. Journal of Applied Rehabilitation Counseling 21(4): 13–21.

23.

Livneh

Antonak

(1997) Psychosocial Adaptation to Chronic Illness and Disability. Gaithersburg, MD: Aspen Publishers.

24.

Livneh

Antonak

(2005) Psychosocial adaptation to chronic illness and disability: A primer for counselors. Journal of Counseling and Development 83(1): 12–20.

25.

Livneh

Antonak

(2008) Reactions to Impairment and Disability Inventory: User’s Manual, Unpublished manuscript.

26.

Livneh

Martz

(2003) Psychosocial adaptation to spinal cord injury as a function of time since injury. International Journal of Rehabilitation Research 26(3): 191–200.

27.

Livneh

Martz

(2014) Coping strategies and resources as predictors of psychosocial adaptation among people with spinal cord injury. Rehabilitation Psychology 59(3): 329–339.

28.

Livneh

Parker

(2005) Psychological adaptation to disability perspectives from chaos and complexity theory. Rehabilitation Counseling Bulletin 49(1): 17–28.

29.

Livneh

Wilson

(2003) Coping strategies as predictors and mediators of disability-related variables and psychosocial adaptation an exploratory investigation. Rehabilitation Counseling Bulletin 46(4): 194–208.

30.

Livneh

Martz

Bodner

(2006) Psychosocial adaptation to chronic illness and disability: A preliminary study of its factorial structure. Journal of Clinical Psychology in Medical Settings 13(3): 250–260.

31.

Lloyd

Schmidt

Khondoker

. (2015) Can psychological interventions reduce perfectionism? A systematic review and meta-analysis. Behavioural and Cognitive Psychotherapy 43(6): 705–731.

32.

Lustig

(2005) The adjustment process for individuals with spinal cord injury: The effect of perceived premorbid sense of coherence. Rehabilitation Counseling Bulletin 48(3): 146–156.

33.

Martz

(2004) Do reactions of adaptation to disability influence the fluctuation of future time orientation among individuals with spinal cord injuries? Rehabilitation Counseling Bulletin 47(2): 86–95.

34.

Mehta

Orenczuk

Hansen

. (2011) An evidence-based review of the effectiveness of cognitive behavioral therapy for psychosocial issues post-spinal cord injury. Rehabilitation Psychology 56(1): 15–25.

35.

Middleton

Tate

Geraghty

(2003) Self-efficacy and spinal cord injury: Psychometric properties of a new scale. Rehabilitation Psychology 48(4): 281–288.

36.

Molnar

Flett

Sadava

. (2012) Perfectionism and health functioning in women with fibromyalgia. Journal of Psychosomatic Research 73(4): 295–300.

37.

Post

MWM

Van Leeuwen

CMC

(2012) Psychosocial issues in spinal cord injury: A review. Spinal Cord 50(5): 382–389.

38.

Roxborough

Hewitt

Kaldas

. (2012) Perfectionistic self-presentation, socially prescribed perfectionism, and suicide in youth: A test of the perfectionism social disconnection model. Suicide and Life-Threatening Behavior 42(2): 217–233.

39.

Shanmugasegaram

Flett

Madan

. (2014) Perfectionism, type D personality, and illness-related coping styles in cardiac rehabilitation patients. Journal of Health Psychology 19(3): 417–426.

40.

Tabachnick

Fidell

(2007) Using Multivariate Statistics (5th edn). Boston, MA: Allyn & Bacon.

41.

World Health Organization (2013) Spinal cord injury, November. Available at: http://www.who.int/mediacentre/factsheets/fs384/en/