Autism Spectrum Disorders, Anxiety, and Religion: The Role of Personality Traits

Abstract

In mental health care, religion and spirituality can both support and hinder the therapeutic process. This is related to the way people see God or the divine, known as ‘God representations’. Previous research suggests that God representations of persons with autism spectrum disorders (ASD) tend to be more negative compared with others. The current study, conducted among 103 participants, shows that after adjusting for religious saliency, having an ASD diagnosis had no independent power to predict God representations. However, certain personality traits, being associated with ASD, did. Specifically, low self-directedness and low reward dependence were associated with more negative God representations. ASD usually is a diagnosis for life, and personality traits do not easily change. Scientific and clinical implications are discussed.

1. Introduction

Religiousness and spirituality (R/S) affect the well-being and functioning of psychiatric patients to a large extent (e.g., Hackney and Sanders 2003; Koenig 2009). As supporting or impeding factors during their illness, they play a role in mental health treatment (e.g., Borras et al. 2007; Pardini et al. 2000). Positive spirituality, with its overtones of support, comfort, and hope, is accompanied by an optimistic life orientation, higher resilience, and lower levels of anxiety (Koenig 2009). Hence, it may serve as a supporting factor in therapeutic interventions. In contrast, negative spirituality, being coloured by anxiety and isolation, hinders the therapeutic process (Koenig 2009). For professionals in mental health care, it is useful to know which patients can benefit from religion or spirituality during an illness and the recovery process, and for which groups of patients this is not the case. In the current study, religion and religiousness were seen as the social, institutional, and cultural context of spirituality (Pargament 2007).

Empirical studies have suggested that R/S may be associated with autism spectrum disorders (ASD) in multiple ways. Some studies have indicated that individuals with ASD are more likely to identify as atheists or agnostics, and to report a lower level of belief in God, possibly due to difficulties in mentalizing (Caldwell-Harris et al. 2011; Ekblad and Oviedo 2017; Norenzayan et al. 2012; Reddish et al. 2016). Additionally, studies found that individuals with stronger autistic traits tended to score lower on measurements of spirituality and report more struggles related to R/S (Crespi et al. 2019; Schaap-Jonker et al. 2012, 2013). However, other studies have suggested that R/S can play a beneficial role in the lives of some individuals with ASD. Religion has been linked to a greater sense of self, a stronger feeling of belonging, and increased social support in various populations, including individuals with ASD (Biggs and Carter 2016; Crawford Sullivan and Aramini 2019). Some individuals with ASD have also described religious experiences in ways that are meaningful to them, including supportive sensory-based supernatural experiences and socially and emotionally comprehensible imaginative interactions with religious agents (Visuri 2018, 2020). The variations in research findings suggest that the relationship between ASD and R/S is complex, and it may depend on individual differences in cognitive processing, personal backgrounds, and social environments.

A key aspect of religiousness and spirituality is the way in which people relate to God/the divine1, known as ‘God representations’. Both religious and nonreligious patients, as well as those who identify as spiritual, have ways in which they relate to God or the divine. God representations refer to God as being subjectively experienced by the individual. Answers to the question ‘Who is God to you?’ comprise both the cognitive and affective meaning of God to the individual. The content of God representations can be operationalised in two dimensions: (1) the feelings the individual experiences in their relationship with God, and (2) the perceptions of God’s actions. As a mental characteristic, God representations do not only provide insight into the individual’s religious experiences and religious behaviour, but also show how the individual relates to others and the world (Davis et al. 2013).

In our previous study on God representations and ASD, we found that God representations of individuals with ASD have fewer positive traits than God representations of other psychiatric patients (Schaap-Jonker et al., 2013). Additionally, in comparison to a non-clinical norm group, the God representations of those with ASD were characterised by fewer positive feelings and more anxiety and anger towards God, as well as with perceptions of God as being less supportive and more ruling/punishing and passive. Those with ASD who scored higher on religious saliency (which reflected that religion played an important role in their daily lives and had intrinsic value for them) reported more positive feelings towards God, more supportive and ruling/punishing perceptions, less anger towards God, and less passive perceptions of God’s actions.

Concerning the God representation ‘feelings of anxiety towards God’ in relation to autism, some specific findings deserve attention. An association was found between the number of reported autistic traits, especially in the social domain, and the level of anxiety towards God. Furthermore, religious denomination also played a role in this context. Strict Calvinists (belonging to orthodox denominations) with ASD experienced more anxiety towards God in comparison to others (Schaap-Jonker et al., 2013). When analysed in detail, in this regard, two types of anxiety seemed to play a role, namely ‘anxiety driven by uncertainty’ and ‘anxiety driven by guilt’. Individuals with ASD who experienced difficulties especially in the social domain were more likely to feel anxiety towards God driven by uncertainty. Additionally, Strict Calvinists with ASD reported more anxiety towards God that seemed to be related to guilt (Schaap-Jonker et al. 2012). The first association may indicate that difficulties in the social and interpersonal domain extend into the religious/spiritual domain. A possible explanation for the second finding may be found in religious culture and religious beliefs; Calvinists tend to see God predominantly as a judge (Schaap-Jonker et al. 2013).

These results raise several questions. Are anxious and uncertain traits in the representation of God specific to patients with ASD, or are they also present among patients with other diagnoses, for example, with anxiety disorders? Perhaps the association between ASD and an anxious and uncertain God representation reflects a more general attitude towards life in general, and (social) relationships in particular, which is characterized by anxiety and uncertainty. Additionally, one may question to what extent religious culture is related to aspects of God representations (Schaap-Jonker 2018). Is anxiety driven by guilt specific to orthodox Protestants for example? It may be useful to repeat and expand our former study among other samples of patients and non-patients with various degrees of religiousness.

At the same time, another aspect is also worthwhile considering. It may be possible that associations between R/S and ASD are mediated by personality traits. Several independent studies have shown that adults with ASD present with a specific pattern of personality characteristics that differs from the normal population. In the Temperament and Character Inventory (TCI) (Svrakic et al. 2002), this personality profile is typified by high harm avoidance (HA) and low reward dependence (RD), low self-directedness (SD) and cooperativeness (CO), and high self-transcendence (ST) (Sizoo et al. 2009; cf. Anckersäter et al. 2006). In a study among non-clinical churchgoers, four of these personality characteristics were associated with predominantly negative God representations. High HA was associated with more negative (i.e., anxious and angry) and less positive feelings towards God, as well as with fewer perceptions of God’s actions as supportive (Eurelings-Bontekoe et al. 2005). Persons low in RD also had fewer positive feelings and fewer supportive perceptions in relation to God. Furthermore, low SD and low CO were related to more negative and less positive feelings towards God and to perceptions of God’s actions as less supportive and more passive. Conversely, the authors found that high ST was associated with more positive feelings towards and supportive perceptions of God’s actions. However, after controlling for psychological distress, the relationships between personality traits and negative feelings disappeared. Although the sample of the study of Eurelings-Bontekoe et al. (2005) was a non-clinical one, we may conclude that God representations of individuals with ASD more or less correspond to God representations of others with the same personality profile.

Taken together, the question is to what extent God representations of individuals with ASD are primarily determined by the nature of their pathology (type of diagnosis), personality profile, psychological distress, or level of religiousness. More insight into the dynamic relationships between religiousness, autistic traits, and personality may guide the therapist’s or spiritual counselor’s interpretation and evaluation of the patient’s religiousness and function as a point of departure for specific clinical interventions. A better understanding of the relationship between God representation and ASD will enable the therapist or spiritual counselor to more effectively tune into the beliefs and cognitions of patients with ASD, increasing the trust needed for a good therapeutic relationship.

The aim of the current study, therefore, is to gain more insight into the God representation of individuals with ASD as compared to individuals with anxiety disorders or without diagnoses. We address the following primary research question: To what extent are autism spectrum disorders and anxiety disorders associated with two types of anxiety (anxiety because of uncertainty and anxiety because of guilt) towards God? In order to gain insight into God representations, diagnoses, and personality traits, we additionally address a secondary research question: To what extent are differences in God representations predicted by diagnoses (autism spectrum disorders and anxiety disorders), degree of religiousness, psychological distress, or personality traits? These theoretical insights may contribute to a better understanding of how R/S can be integrated into clinical practice.

2. Method

The current study had a cross-sectional research design and was conducted within the mental health care organization Dimence, in the Netherlands. Before the start of the study, the Medical Ethics Review Committee (METC) of VUmc determined that the study did not fall under the scope of the Dutch Medical Research Involving Human Subjects Act (WMO). This meant that formal approval from the METC was not required. The principles of the Declaration of Helsinki (version October 2008) were maintained during the study. The data used were anonymized and stored according to the legal provisions at Dimence.

2.1. Sample Selection and Recruitment

Practitioners approached patients with ASD or anxiety disorder (AD) directly to ask if they were interested in participating in the study. However, participation was completely voluntary, and patients were informed that declining to participate would have no impact on their treatment or relationship with the therapist. Diagnoses were made and communicated by experienced mental health professionals. Exclusion criteria were: mental retardation and a high risk of crisis. Furthermore, the researchers recruited a community sample of non-clinical persons. Possible participants received a letter with information about the aim and methods of the study. Before participating, all the respondents signed an informed consent form. Participants completed a paper questionnaire in their own time, which took about 40 min. They were asked to return the questionnaires and informed consent forms. The control group was approached by the researchers applying the same procedure. A total of 103 participants were included in the study, with ASD (n = 42), AD (n = 20) or nonclinical status (n = 41).

2.2. Measures

The survey included general demographic questions, a questionnaire concerning God representations (QGR), and two concerning the level of religiousness (DR and RS), as well as two additional questionnaires to investigate personality traits (VTCI-NL) and general distress (BSI). We wanted to explore both types of anxiety, as well as mediation through personality traits. Including the BSI provided a good measure to prevent distortion of results due to overall distress.

Questionnaire God Representations (QGR). This questionnaire contains 33 items and covers 2 dimensions: feelings someone has in their relationship with God, and the perception of God’s actions. The dimension ‘feelings towards God’, has three subscales: Positive Feelings (POS, e.g., thankfulness, love, Cronbach’s α = 0.97), Anxiety (ANX, Cronbach’s α = 0.87), and Anger (ANG, Cronbach’s α = 0.86) towards God. The dimension ‘God’s actions’, also has three subscales: Supportive Actions (SUP, Cronbach’s α = 0.98), Ruling and/or Punishing Actions (RULP, Cronbach’s α = 0.88), and Passivity (PAS, Cronbach’s α = 0.87). Answers were rated on a 5-point scale, ranging from 1 (not at all descriptive, ‘that never applies to me’) to 5 (completely descriptive, ‘that always applies to me’). Psychometric qualities of the QGR are adequate (Schaap-Jonker et al. 2016; Sharp et al. 2019). Normative data are available for psychiatric outpatients and the general population and for respondents of various religious denominations (Schaap-Jonker and Eurelings-Bontekoe 2009). In addition to the 33 items, 5 experimental anxiety items were added to the questionnaire in order to construct 2 subscales, ‘anxiety driven by uncertainty’ and ‘anxiety driven by guilt’, trying to replicate the results of a former study (Schaap-Jonker et al. 2013). These five items were: anxiety of making the wrong choices, anxiety of being ignored, anxiety of being abandoned, doubt, and anxiety of failing.

Degree of religiousness (DR). One item measured the extent to which participants see themselves as religious, using a 5-point scale with score 1 (‘secular’), score 2 (‘less than average religious’), score 3 (‘religious’), score 4 (‘more than average religious’), and score 5 (‘orthodox religious’). Furthermore, a questionnaire on religious saliency (RS) was administered. This questionnaire examines how important faith or a religious worldview is to people. The questions were derived from questions conceived by Eisinga et al. (2002) and consisted of three items: ‘religion is important to me’, ‘religion plays a role by making decisions’, and ‘I could not live without faith’. Answers were rated on a 5-point scale, ranging from score 1 (‘that never applies to me’) to score 5 (‘that always applies to me’), resulting in a range of 3–15 (Cronbach’s α = 0.95).

‘Verkorte Temperament and Character Inventory –NL’ (VTCI-NL). This is an abbreviated Dutch version of the 240-item Temperament and Character Inventory. The VTCI-NL questionnaire measures personality dimensions and comprises 105 items with answer options of ‘correct’ or ‘incorrect’ (Cloninger et al. 1994; Duijsens et al. 1997). The questionnaire has seven subscales with ranges of 0–15, reflecting four so-called temperament dimensions: novelty seeking (Cronbach’s α = 0.70), harm avoidance (Cronbach’s α = 0.92), reward dependence (Cronbach’s α = 0.69), and persistence (Cronbach’s α = 0.72); and three so-called character dimensions: self-directedness (Cronbach’s α = 0.88), cooperativeness (Cronbach’s α = 0.78), and self-transcendence (Cronbach’s α = 0.91).

Brief Symptom Inventory (BSI). In this questionnaire, participants rate the extent to which they have been bothered in the past week by various symptoms (Derogatis 1993; De Beurs and Zitman 2006). The BSI consists of 53 items (α = 0.96) divided into 9 scales. Items were rated on a 5-point scale from 0 (not affected by the complaint) to 4 (very affected by the complaint). In the current study, we used an 18-item subscale, consisting of 3 subscales: somatic complaints, anxiety, and depression. This scale was considered a ‘general distress’ measure (range 0–72, Cronbach’s α = 0.94).

2.3. Statistics

First, descriptive statistics were performed on the variables used. Differences between the diagnosis groups were tested by means of χ² tests, ANOVA analyses, and a Kruskal-Wallis test for nonparametric data. Furthermore, Pearson’s r was calculated for religious saliency and degree of religiousness.

Second, a principal component analysis with varimax rotation and Kaiser normalization was applied to the old and new anxiety items, measured by the QGR, to investigate the two types of anxiety towards God (anxiety driven by guilt and anxiety driven by uncertainty). The Kaiser-Meyer-Olkin test showed that the sample was appropriate (0.81), and Bartlett’s test of sphericity was significant (0.000). The items were forced into two components, and items that had cross-loadings of >0.50 were removed from the analysis. Furthermore, items with factor loadings > 0.40 on both components with a difference less than 0.20 were excluded. The PCA was performed using only data from the respondents with ASD or AD in order to replicate the findings by Schaap-Jonker et al. (2013).

Third, the two new scales of anxiety towards God were hierarchically regressed on sex and age (block 1), diagnosis (ASD, AD, nonclinical, block 2), general distress (BSI, block 3), and religious saliency (RS, block 4). Beforehand, we checked the outcome variables for normality by means of the Kolmogorov-Smirnov test. When this test was significant (p < 0.05), indicating that the distribution of the sample was significantly different from a normal distribution, we checked the assumption of the quality of the covariance matrices by conducting a Box’s test. This was also applied to the analyses described below. This test was non-significant (p > 0.05) for all outcome variables, which implied that the covariance matrices were equal.

Fourth, six additional hierarchical linear regression analyses were performed with all subscales of the QGR as outcome variables. Predictive factors included in the models were: diagnosis, RS, BSI, and seven subscales of personality characteristics (VTCI-NL). Block 1 included the diagnostic groups, in block 2 religious saliency was added, in block 3 general distress, and block 4 additionally included personality traits. A variance inflation factor <4 was maintained, and the software used was IBM SPSS Statistics, version 26.

3. Results

Participant characteristics are presented in

Table 1. Participant characteristics.

		Total Population (n = 103)	ASD (n = 42)	AD (n = 20)	Non-Clinical (n = 41)	χ2	F	p
Sex (% male)		50	67	40	34	9.5		0.009
Age (M, SD)		34.4 (12.9)	33.4 (11.5)	35.1 (11.3)	35.0 (15.1)		0.2	0.828
Partner (%)	Yes	52	41	50	63	5.8		0.443
	No	41	50	40	32
	No longer	7	7	10	5
Education (%)	Low (±8 years)	12	12	10	7	14.1		0.296
	Medium (±12 years)	40	43	55	29
	High (±16 years)	49	41	35	63
Degree of religiousness (%)	Secular	36	43	20	37	18.1		0.021
	Less than average religious	21	31	30	7
	Average religious	20	17	30	20
	More than average religious	13	7	15	17
	Orthodox religious	10	2	5	20
Religious saliency (range 3–15 M, SD)		8.5 (4.5)	6.6 (3.5)	9.1 (4.6)	10.1 (4.8)		7.1	0.001
BSI (range 0–72 Median, Interquartile Range)		11 (23)	18 (20)	22.5 (22.8)	4 (6.5)			0.000
VTCI-NL
	Novelty seeking (range 0–15 M, SD)	5.0 (3.0)	4.5 (2.8)	4.2 (2.1)	6.0 (3.4)		3.2	0.044
	Harm avoidance (range 0–15 M, SD)	10.1 (4.7)	13.2 (2.0)	12.1 (2.4)	5.8 (4.4)		55.6	0.000
	Reward dependence (range 0–15 M, SD)	8.5 (3.4)	7.4 (3.6)	10.1 (2.6)	9.1 (3.1)		4.5	0.012
	Persistence (range 0–15 M, SD)	8.9 (3.3)	8.5 (2.8)	10.6 (3.1)	8.5 (3.7)		3.1	0.050
	Self-directedness (range 0–15 M, SD)	9.9 (4.3)	7.9 (3.8)	7.9 (3.8)	13.0 (2.9)		20.9	0.000
	Cooperativeness (range 0–15 M, SD)	12.6 (2.7)	11.3 (2.9)	13.9 (1.3)	13.3 (2.3)		9.5	0.000
	Self-transcendence (range 0–15 M, SD)	4.5 (3.8)	3.9 (3.8)	6.0 (4.3)	4.5 (3.5)		1.9	0.156
QGR
Feelings	Positive (range 8–40 M, SD)	22.4 (11.3)	18.7 (9.6)	26.2 (8.9)	24.4 (12.9)		4.1	0.020
	Anxiety (range 5–25 M, SD)	9.1 (4.7)	9.1 (5.0)	11.6 (5.5)	8.0 (3.6)		4.0	0.021
	Anger (range 3–15 M, SD)	5.0 (2.9)	5.2 (3.2)	6.0 (3.0)	4.5 (2.3)		2.1	0.133
Perceptions	Supportive (range 8–40 M, SD)	22.7 (12.5)	18.3 (10.0)	26.3 (11.2)	25.5 (14.1)		4.5	0.013
	Ruling/punishing (range 4–20 M, SD)	9.3 (5.3)	7.9 (4.8)	10.6 (5.2)	10.2 (5.9)		2.6	0.078
	Passivity (range 2–8 M, SD)	4.3 (2.8)	5.1 (3.2)	5.2 (2.6)	3.2 (1.9)		6.8	0.002

. Most of the variables used in the analyses differed significantly between the diagnosis groups. The age of the participants ranged between 17 and 63. Participants differed concerning religious background and religious behavior, which were measured in various ways. Concerning church attendance: 43% of the research population never went to church, 24% attended church regularly, and 18% attended church weekly. Furthermore, 34% never prayed, 27% prayed less than once a week, 10% prayed once a week, 10% prayed several times a week, and 30% prayed at least daily. Religious saliency was highly correlated with the degree of religiousness (Pearson’s r = 0.81, p < 0.01). Patients considering themselves ‘secular’ scored M 4.7 (SD 3.1) on religious saliency; ‘less than average religious’ scored M 6.4 (SD 2.4); ‘average religious’ scored M 11.4 (SD 2.0); ‘more than average religious’ scored M 13.5 (SD 2.8), and ‘orthodox religious’ scored M 14.0 (SD 1.1). Therefore, we used the RS saliency questionnaire as a measure of the degree of religiousness in our regression analyses (see below).

3.1. Principal Component Analysis

The principal component analysis on the anxiety items of the QGR with a fixed two-factor solution explained 74.7% of variance (

Table 2. Principal component analysis for two types of anxiety towards God.

Item	Component
	1	2
Anxiety of being punished	0.868
Guilt	0.855
Anxiety of failing	0.834	0.404
Doubt		0.848
Uncertainty	0.434	0.725
Anxiety of being ignored		0.719
Eigenvalues	2.435	2.045
% of variance	40.590	34.087
Cronbach’s α	0.88	0.74

). The first factor comprised the items ‘guilt’, ‘anxiety of failing’, and ‘anxiety of being punished’, and was labeled as ‘anxiety driven by guilt’. This factor explained 40.6% of the variance and had an eigenvalue of 2.44, (Cronbach’s α = 0.88). The second factor, explaining 34.1% of variance and with an eigenvalue of 2.05, was labeled as ‘anxiety driven by uncertainty’. This factor consisted of the items ‘doubt’, ‘uncertainty’, and ‘anxiety of being ignored’ (Cronbach’s α = 0.74).

3.2. Two Types of Anxiety Related to Religiousness and Diagnosis

The first two multiple hierarchical regression analyses on the anxiety scales of the QGR revealed that autism spectrum disorders and anxiety disorders were not significantly related to both types of anxiety when adjusted for general distress and religious saliency (see the final models presented in

Table 3. Linear regression analyses for two types of anxiety.

	Anxiety Driven by Guilt						Anxiety Driven by Uncertainty
	B	β	SE	t	p	B	β	SE	t	p
Sex	−0.08	−0.01	0.59	−0.13	0.895	−1.11	−0.15	0.69	−1.62	0.110
Age	−0.02	−0.08	0.02	−0.90	0.369	0.02	0.07	0.03	0.72	0.472
Diagnoses
ASD	0.52	0.08	0.76	0.70	0.496	−0.39	−0.05	0.88	−0.44	0.663
Anxiety disorder	−1.40	−0.22	0.85	−1.66	0.101	1.32	0.15	0.98	1.34	0.183
General distress	0.04	0.18	0.02	1.83	0.071	0.10	0.41	0.03	3.94	0.000
Religious saliency	0.33 1	0.47	0.07	4.74	0.000	0.14	0.18	0.08	1.80	0.076
Adjusted R2					0.26					0.22
Df					6, 98					6, 97
F					5.4					5.6

¹ Significant results are presented in bold.

). Anxiety driven by guilt was predominantly related to religious saliency, and associations with anxiety disorders in block 2 (β = 0.22, t = 2.0, p = 0.053) and general distress in block 3 (β = 0.21, t = 1.8, p = 0.069) could be seen as a statistical trend. Anxiety driven by uncertainty was predominantly associated with general distress (Table 3). In block 2, a significant association with anxiety disorders also was present (β = 0.30, t = 2.9, p = 0.005), but the subsequent blocks showed that this association was mediated by a higher level of general distress.

3.3. Personality Profiles and God Representations

The first blocks of the hierarchical regression analyses on the six different scales of the QGR revealed that participants with ASD had less positive feelings towards God (β = −0.28, t = −2.3, p = 0.024) and less perceptions of God’s actions as supporting (β = −0.31, t = −2.6, p = 0.010) and/or ruling/punishing (β = −0.26, t = −2.2, p = 0.034) compared with others. However, these associations were explained by a lower score for religious saliency for this group, as the associations disappeared when religious saliency was added to block 2. In addition, patients with ASD who had a higher score of religious saliency had more feelings of anxiety towards God compared with others (β = 0.29, t = 2.4, p = 0.019), but this association was mediated by general distress, which in turn was mediated by personality traits. Furthermore, patients with ASD and patients with anxiety disorders experienced God more as passive compared with nonclinical individuals (ASD β = 0.31, t = 2.4, p = 0.018; AD β = 0.24, t = 2.0, p = 0.045), which again was mediated by general distress and personality traits, respectively. The significant associations between general distress and feelings of anxiety towards God (β = 0.27, t = 2.4, p = 0.018), feelings of anger towards God (β = 0.24, t = 2.1, p = 0.039), and experiencing God as passive (β = 0.37, t = 3.2, p = 0.002) turned to insignificance when personality traits were added to the models in the last blocks. However, the association between general distress and positive feelings towards God turned to significance when personality traits were added (

Table 4. Linear regression analyses for the subscales of the questionnaire on God representations.

Feelings	Positive					Anxiety					Anger
	B	β	SE	t	p	B	β	SE	t	p	B	β	SE	t	p
Diagnoses
ASD	0.37	0.02	2.19	0.17	0.865	−0.79	−0.09	1.52	−0.52	0.602	−0.79	−0.15	0.93	−0.84	0.402
Anxiety disorder	0.56	0.02	2.43	0.23	0.818	−0.51	−0.04	1.68	−0.30	0.764	0.41	0.06	1.04	0.39	0.697
Religious saliency	1.78 1	0.70	0.19	9.34	0.000	0.49	0.49	0.13	3.77	0.000	0.22	0.37	0.08	2.73	0.008
General distress	−0.17	−0.22	0.06	−2.71	0.009	0.01	0.04	0.04	0.29	0.770	0.00	−0.00	0.03	−0.01	0.989
Personality traits
Cooperativeness	0.08	0.02	0.31	0.26	0.793	0.19	0.12	0.22	0.89	0.376	−0.25	−0.26	0.13	−1.86	0.067
Harm avoidance	0.26	0.11	0.26	1.00	0.320	0.22	0.22	0.18	1.22	0.225	0.11	0.19	0.11	1.03	0.304
Novelty seeking	0.06	0.02	0.26	0.24	0.808	0.04	0.03	0.18	0.23	0.823	0.10	0.12	0.11	0.91	0.367
Reward dependence	0.48	0.14	0.23	2.10	0.039	−0.11	−0.08	0.16	−0.72	0.477	−0.15	−0.18	0.10	−1.54	0.128
Persistence	−0.41	−0.12	0.21	−1.94	0.057	−0.11	−0.08	0.14	−0.73	0.466	−0.05	−0.05	0.09	−0.51	0.611
Self-directedness	−0.36	−0.14	0.25	−1.43	0.158	−0.41	−0.40	0.18	−2.35	0.022	−0.10	−0.17	0.11	−0.94	0.349
Self-transcendence	0.78	0.26	0.23	3.39	0.001	−0.22	−0.19	0.16	−1.40	0.166	−0.01	−0.01	0.10	−0.11	0.915
Adjusted R2					0.76					0.28					0.23
Df					11, 78					11, 78					11, 79
F					23.7					3.7					3.1
Actions	Supporting					Ruling/Punishing					Passive
	B	β	SE	t	p	B	β	SE	t	p	B	β	SE	t	p
Diagnoses
ASD	1.03	0.04	2.28	0.45	0.652	0.70	0.07	1.33	0.53	0.598	0.44	0.08	0.91	0.48	0.632
Anxiety disorder	0.98	0.03	2.61	0.38	0.707	0.90	0.06	1.48	0.61	0.545	0.97	0.13	1.04	0.93	0.357
Religious saliency	2.21	0.81	0.20	11.0	0.000	1.06	0.90	0.12	9.08	0.000	−0.14	−0.23	0.08	−1.72	0.089
General distress	−0.15	−0.18	0.07	−2.21	0.030	0.02	0.06	0.04	0.55	0.587	0.02	0.13	0.03	0.92	0.363
Personality traits
Cooperativeness	0.42	0.10	0.33	1.30	0.199	0.18	0.10	0.19	0.95	0.345	−0.17	−0.18	0.13	−1.34	0.186
Harm avoidance	0.17	0.06	0.27	0.63	0.534	0.04	0.03	0.16	0.25	0.806	−0.14	−0.24	0.11	−1.32	0.193
Novelty seeking	0.02	0.01	0.27	0.07	0.947	0.11	0.07	0.16	0.72	0.477	0.02	0.03	0.11	0.22	0.822
Reward dependence	0.35	0.10	0.24	1.50	0.137	0.01	0.01	0.14	0.09	0.931	0.08	0.10	0.09	0.89	0.379
Persistence	−0.45	−0.12	0.22	−2.03	0.046	−0.35	−0.22	0.13	−2.74	0.008	0.08	0.09	0.09	0.86	0.395
Self-directedness	−0.37	−0.13	0.26	−1.41	0.164	−0.04	−0.03	0.15	−0.24	0.812	−0.26	−0.42	0.11	−2.42	0.018
Self-transcendence	0.35	0.11	0.24	1.47	0.145	−0.24	−0.17	0.14	−1.70	0.094	0.16	0.22	0.10	1.68	0.098
Adjusted R2					0.78					0.59					0.27
Df					11, 77					11, 77					11, 78
F					25.5					10.9					3.6

¹ Significant results are presented in bold.

), indicating that personality traits play a moderating role.

The final models showed that religious saliency was the main significant predictor of all types of God representations, except for seeing God as passive (see the final models presented in Table 4). The associations between religious saliency and seeing God’s actions as supportive, ruling/punishing, and having positive feelings towards God were strong, and the associations of religious saliency with feelings of anxiety and anger towards God were moderate. After adjustment for religious saliency, diagnosis had no independent power to predict a particular aspect of God representations. However, several personality traits showed associations with certain aspects: higher self-transcendence and—to a lesser extent—higher reward dependence were associated with more positive feelings towards God; higher self-directedness with fewer feelings of anxiety towards God and seeing God’s actions as less passive; and higher persistence with seeing God’s actions as less supportive and less ruling/punishing. In addition, a higher level of general distress was related to less positive feelings towards God and perceptions of God’s actions as less supportive.

4. Discussion

The current study offers new insights into the field of religion/spirituality (R/S) and autism spectrum disorders (ASD). It shows that God representations are mainly determined by psychological traits and that patients with ASD do not necessarily have more negative God representations than others. This is an additional and partly contrasting insight in comparison with our former study on ASD and God representations, which showed that (more negative) God representations were associated with specific autistic traits (Schaap-Jonker et al. 2012, 2013). The difference between these results might be explained by the differences in the sample: our former study concerned a heterogeneous Christian population, whereas the current study included a more diverse sample with regard to outlook on life.

In our former study, patients with ASD had more anxiety towards God compared with a normal population, and it appeared that this concerned anxiety was due to uncertainty, being distinguished from anxiety due to guilt (Schaap-Jonkeret al. 2012, 2013). The current study, containing extended questions concerning anxiety towards God, confirms that these types of anxiety differ from each other. Anxiety due to uncertainty was predominantly associated with general distress, whereas anxiety due to guilt was mainly related to religious saliency. Both diagnoses, ASD and anxiety disorders, had no independent power to predict these types of anxiety. At the same time, it may be noticed that there is likely to be some overlap of these diagnoses with general distress (Rosen et al. 2018; Iani et al. 2019). In any case, the results confirm the suggestions that anxiety can have different sources: guilt, more religiously colored, and uncertainty, more psychologically colored (Schaap-Jonker et al. 2012, 2013).

Next to anxiety towards God, other aspects of God representations were also not predominantly related to ASD or anxiety disorders. God representations were predominantly determined by religious saliency, which is in line with the earlier study (Schaap-Jonker et al. 2013). The more people see themselves as religious, the more they experience positive feelings towards God, and the more they report God’s actions as supportive. At the same time, these persons also have more feelings of anxiety and anger towards God and see God’s actions more as ruling/punishing, which may not be surprising, because when God exists for someone, there may be feelings and perceptions that are absent for people who do not believe in God. Positive and negative experiences of R/S often occur together (van Nieuw Amerongen-Meeuse et al. 2019).

In general, the current study confirms that patients with ASD are less often religious compared with others (Norenzayan et al. 2012). Possibly, patients with ASD have problems in mentalizing, for example, in engagement with supernatural powers. The same problems could relate to the fact that religious patients with ASD report more anxiety towards God compared with a normal population (Schaap-Jonker et al. 2012, 2013). The current study, however, adds to this knowledge that—also with regard to anxiety towards God—both psychological and religious factors play a part, more than the diagnosis of ASD itself.

A new insight of the current study is that certain specific personality traits more strongly predict God representations than diagnoses. The more people have a low score for self-directedness, the more they experience anxiety towards God and report his actions as ‘passive’. Individuals with low reward dependence experience fewer positive feelings towards God. Persons with low self-transcendence have fewer positive feelings towards God than others. Finally, patients with high persistence experience less support, but also less ruling/punishing actions from God. This means that variations in God representations among people, to a certain extent, go hand in hand with their personality traits. These results are partially in line with the results of Eurelings-Bontekoe et al. (2005) in their study among nonclinical churchgoers. It may not be the diagnostic label that determines ‘positive’ or ‘negative’ religiousness, but the psychological factors behind diagnoses. Low self-directedness and low reward dependence in the current study show associations with more negative God representations, and are known to characterize patients with ASD (Anckersäter et al. 2006; Sizoo et al. 2009; Svrakic et al. 2002). These factors, therefore, may be related to psychopathology, as well as to negative God representations. However, high harm avoidance and low cooperativeness, which are more present in this group, showed no associations, and high self-transcendence was even related to more positive feelings towards God.

Negative God representations can be a source of or contribute to ‘religious distress’. Religious distress, as a broader concept of the struggles with faith and with God, has shown to be related to higher rates of depression and suicidality (Exline et al. 2000). It might be relevant to find out whether religious distress would also be related more to personality traits than to specific psychiatric diagnoses. Understanding this distinction may have implications for appropriate interventions targeting religious distress.

While the primary aim of the study was a theoretical one, we believe that understanding the relationship between religiousness, autistic traits, and personality can provide valuable insights for practitioners working with individuals with ASD. The current study may offer therapists and spiritual counselors a better understanding of how to approach the religious and spiritual needs of their patients. Previous studies have suggested that religious distress diminishes in line with recovery processes. However, ASD is generally considered a lifelong diagnosis, and personality traits are relatively stable over time. This means that for individuals with ASD, religious distress may not follow the same patterns of change as in other populations. Consequently, interventions should focus on cognitive reframing and religious coping strategies rather than assuming a natural decrease in distress over time.

Mental health professionals considering the role of religion and spirituality in their patients’ lives are recommended to realize that autism spectrum disorder is often associated with negative God representations (Schaap-Jonker et al. 2012, 2013) and religious distress. At the same time, our study highlights the importance of looking beyond diagnoses; personality traits appear to play a more significant role in negative God representations. Therefore, we emphasize the need for an individualized approach that tailors interventions to specific personal challenges and religious/spiritual care needs.

5. Limitations and Recommendations

A limitation that should be noted is the small sample size of the current study. This is related to the fact that patients were recruited within one mental health institution, where not everyone had the desire to fill out a questionnaire about religion. Furthermore, the ASD sample exclusively consists of individuals who were receiving treatment at a mental health institution, which makes the results not generalizable to all individuals with autism. In addition, the group participants with anxiety disorders were relatively small. The small sample size may have influenced (a lack of) significance levels, and it would be relevant to duplicate the current study in a larger sample. Future research could also focus on how God representations relate to specific religious denominations or subcultures, beliefs, and narratives. In the current study, patients with ASD, for example, scored lower on religious saliency. It would be relevant to examine comparable associations among ASD patients with high religious saliency. Another limitation is the fact that the measurement was restricted to self-report questionnaires and, therefore, could be biased because of social desirability. In this study, participants were asked to complete the questionnaires at home and return them when finished. As a result, participants were not under time pressure, which likely reduced the risk of anxiety or passivity influencing their responses. However, we did not control for whether participants completed the questionnaires in one sitting or across multiple sessions.

6. Conclusions

Anxiety towards God can be distinguished in anxiety caused by guilt and anxiety due to uncertainty. Mental health professionals are recommended to take this into account when investigating the presence and nature of anxiety. The nature of anxiety in individuals with ASD may be influenced by a variety of factors, including religious beliefs (such as guilt) or psychological factors (such as uncertainty), with these influences potentially interacting in complex ways. It is important to note that religion is not inherently linked to guilt, nor is it necessarily separate from psychological factors like uncertainty. Instead, the relationship between religion and anxiety can vary depending on individual experiences, belief systems, and the broader social and psychological context. Furthermore, in general, low scores on self-directedness, reward dependence, self-transcendence, and persistence are associated with less positive God representations, i.e., with more negative religiousness. Mental health professionals are recommended to take this into account. It would be relevant to investigate how negative religious factors contribute to psychopathology, but also how R/S for patients in these categories could be turned into a helping and supportive factor in their lives.