ABSTRACT
Sleep and its disorders are still the least understood phenomena in biology. This systematic review aims to identify four personality psychobiological models related to sleep disorders and sexual dysfunction. The literature search for English-language records was conducted on the Web of Science, Scopus, PubMed, ProQuest and Cochrane databases from January 1990 to April 2019. According to the number of studies available related to research variables and heterogeneous designs, the pooled effect size was calculated separately for both case-control and correlation studies using the random-effects method because of the heterogeneity of the studies and use of the I2 measure. The numbers of studies on parasomnias, Restless Legs syndrome, sleep breathing disorders, daytime sleepiness and sexual dysfunction were minimal. However, the present meta-analysis on four studies on insomnia showed that the cases and controls in harm avoidance (HA) are significantly different (p=0.001). Another meta-analysis on 11 studies showed that a significant correlation exists between poor sleep quality assessed using the Pittsburgh Sleep Quality index with negative affect (NA) (z=10.437, p<0.001) and positive affect (PA) (z=-6.826, p<0.001). In addition to the few studies on sleep and sexual disorders, the greater challenge is the wide range of these disorders, making it difficult to perform a comprehensive meta-analysis and draw a conclusion. Generally, NA/PA and HA may play critical roles in sleep quality and insomnia, respectively. These findings and limitations are debatable.
Keywords: Character, meta-analysis, personality, sexual dysfunction, sleep disorders, temperament
Introduction
Sleep disorders and sexual dysfunction are among the most common psychiatric morbidities in various populations around the world (1). Sleep disorders that are common in 0.4-48 percent of the United States population included insomnia, sleep breathing disorders (Central Sleep Apnea syndrome, obstructive sleep apnea, and snoring), central hypersomnolence disorders (hypersomnia and narcolepsy), parasomnias, sleep-related movement disorders, and circadian rhythm sleep-wake disorders (2). Meanwhile, sexual dysfunction includes sexual dysfunction (desire, arousal, orgasm, painful intercourse and vaginismus, and premature ejaculation) and paraphilias. Sexual dysfunction is seen in a significant proportion of men and women, depending on the type of disorder (3).
Attempts to identify predisposing risk factors for sleep disorders and sexual dysfunction have highlighted the role of some demographic and clinical variables (4-8). However, the unstable nature of these risk factors has not been able to help differential diagnoses of the wide range of these disorders. Therefore, it is assumed that the differential diagnosis of sleep disorders and sexual dysfunction requires screening of some relatively stable risk factors such as personality traits (9,10). Recently, some studies have examined the personality correlations of these disorders based on factor analysis approaches that have found contradictory findings (9,10). However, given that sleep and its disorders are still one of the least understood phenomena in biology (11), in the last two decades, the study of personality correlations of these disorders based on psychobiological approaches has accelerated (12-15).
In models based on personality psychobiological theories, temperament (and/or character) abnormalities are the main cause of psychiatric disorders (16,17). The main psychobiological theories of personality include Cloninger’s temperament and character model (18,19), the positive and negative affect/temperament model proposed by Watson et al. (20), Akiskal’s affective temperaments model (21), and the affective and emotional composite temperament proposed by Lara et al. (22). So far, several studies have examined the relationship between the structures of these theories and sleep disorders and sexual dysfunction (11-14,23-28). However, the findings are very scattered and the integrated interpretation of them is challenging. For example, two studies (29,30) have found dissimilar findings on the relationship between harm avoidance (HA) in Cloninger’s theory and insomnia disorder. Other studies have also reported contradictory results on the relationship between positive affect and sleepiness (31,32). On the other hand, some studies have suggested inconsistent results about the relationship between HA and sexual satisfaction (33,34). Thus, the combined findings of previous studies in the form of a comprehensive review could help to the differential diagnosis of sleep disorders and sexual dysfunction using personality psychobiological theories. So, the present review was conducted to identify four personality psychobiological models by Cloninger (18,19), Watson et al. (20), Akiskal et al. (21), and Lara et al. (22) related to sleep disorders and sexual dysfunction.
Materials and Methods
Databases, Search Strategies, and Selection Criteria
The present systematic review and meta-analysis process, including the determination of data sources and databases, search strategies, quality assessment of studies, and data extraction, follows the PRISMA 27-item checklist guidelines. Web of Science, Scopus, PubMed, ProQuest, and Cochrane were searched for English-language records published during January 1990 and April 2019. In the search strategy, first, the keywords related to the purpose of the study were extracted based on previous studies and reviews. Then, the systematic search of keywords in the title/abstract of the records was done by two members of the research team (AH & SK). The list of keywords included [“Temperament” OR “temperament and character inventory (TCI)” OR “TCI-R” OR “tridimensional personality questionnaire (TPQ)” OR “novelty seeking” OR “harm avoidance” OR “reward dependence” OR “persistence” OR “self-directedness” OR “cooperativeness” OR “self-transcendence” OR “affective temperament” OR “TEMPS-A” OR “positive and negative affect schedule (PANAS)” OR “positive affect” OR “negative affect” OR “affective emotional composite temperament” OR “AFECTS”] AND [“sleep-wake disorders” OR “insomnia disorder” OR “hypersomnolence disorder” OR “narcolepsy” OR “obstructive sleep apnea hypopnea” OR “central sleep apnea” OR “sleep-related hypoventilation” OR “circadian rhythm sleep-wake disorders” OR “parasomnias” OR “sleepwalking” OR “sleep terrors” OR “nightmare disorder” OR “harm avoidance” OR “restless legs syndrome” OR “sleepiness” OR “sleep quality”] (OR) [“sexual disorders” OR “sexual dysfunctions” OR “delayed ejaculation” OR “erectile disorder” OR “orgasmic disorder” OR “arousal disorder” OR “penetration disorder” OR “hypoactive sexual desire disorder” OR “premature ejaculation”].
The inclusion criteria for systematic review were: English-language original research containing adults aged 18 and over, publication in academic journals, and coverage of at least two keywords simultaneously. Exclusion criteria included studies containing qualitative methodology, unpublished dissertations and articles, studies with unrelated/unsuitable design, interventions without assessment in the baseline, reviews and meta-analysis, use of non-standard tools without validity and unstructured interviews in evaluating variables, not using TCI, TCI-R, TPQ, TEMPS, PANAS, AFECTS to measure temperament, conference abstracts or without full text, lack of access to the full text of the article, and low-quality reports according to the strengthening the Reporting of observational studies in epidemiology (STROBE). Also, in proportion to the type of sleep disorder or sexual dysfunction, correlation studies without a reported partial correlation coefficient or case-control studies without a reported mean and standard deviation were excluded from the meta-analysis. Studies containing duplicate samples were also excluded (Figure 1).
Quality Assessment of Studies and Data Extraction
The quality of the studies was assessed using a 22-item STROBE checklist used for cross-sectional, case-control, and longitudinal studies. The process of evaluating the quality of studies was performed independently by two members of the research team (AH & SK). The disagreement between the two assessors was resolved using the opinion of a third party (KR). Then, to extract the data, a comprehensive table was designed to record the extracted data. After selecting the appropriate quality articles, the data and information were each entered into the results registration form/table. The process of combining the extracted data was performed by the tabulation method and an accurate description of the findings of each study. The table of results was organized based on the first author and the year of study, country, design, sample size and age mean (or age range) of samples, tools, statistical analysis, related findings, and limitations and level of evidence.
Statistical Analysis
Two meta-analyses were performed to calculate (i) difference in the TCI subscales (novelty seeking=NS, harm avoidance=HA, reward dependence=RD, persistence=P, self-directedness=SD, cooperativeness=C, and self-transcendence=ST) between cases with insomnia and normal controls- 4 studies; and (ii) the correlation coefficients between each of the subscales of PANAS (positive affect=PA, negative affect=NA) and sleep quality assessed by the Pittsburgh Sleep Quality index-11 studies. Case-control studies were combined based on sample size, mean, and standard deviation of the variables in the cases and healthy controls. Pooled effects sizes for standard mean differences (cases vs. controls) are presented with 95% confidence intervals (CI) using a combined forest plot. Cohen’s d-values were used for measuring the effect size. Differences between cases and controls were compared using the Z-standard score. Correlation studies were combined based on sample sizes and partial correlation coefficients between variables. The unit of analysis was an individual study and effect sizes in these meta-analyses were obtained by the Fisher z-transformation of correlation coefficients. Estimates of the correlation coefficients in individual studies were graphically presented for each study based on the subscales of TCI or PANAS in pooled forest plots.
Heterogeneity was considered in a large number of studies with I2 higher than 50% (81% of studies related to sleep and 83% studies related to sexual dysfunction). Thus, pooled estimates of standard means difference was calculated for the subscales of TCI and types of sleep disorders using the random-effects method. Also, pooled estimates of correlation coefficients were calculated for the subscales of PANAS and sleep disorders or the subscales of TCI and sexual dysfunction using the random-effects method. We studied the heterogeneity of the study samples using the I2 statistics for a 95% CI. A p-value <0.05 for I2 higher than 50% was considered a significant heterogeneity test. Egger’s test was used to possible publication bias. All hypotheses were tested at the level of <0.05 and performed using the comprehensive meta-analysis (CMA.2) software.
Results
Studies Included in the Systematic Review
A systematic literature search yielded 469 papers related to sleep disorders and 99 papers related to sexual dysfunction. Finally, 43 (15 in the meta-analysis) papers were entered into the final sample related to sleep disorders, and 10 papers were entered into the final sample related to sexual dysfunction. The process of selecting and excluding studies is shown in Figure 1. Most studies have examined the relationship between the TCI/PANAS and sleep disorders or TCI and sexual dysfunction. Very few studies have been found on the relationship between the TEMPS/AFECTS and sleep disorders or sexual dysfunction. The summary of the methods and results of the studies can be seen in Tables 1 and 2.
Differences Between Cases with Sleep Disorders and Controls in the Subscales of TCI
Table 3 summarizes the systematic review results of TCI and sleep disorders. As can be seen in this table, the number of studies in the field of parasomnias, restless legs syndrome (RLS), and sleep breathing disorders are very limited. Differences between cases with insomnia disorder and healthy controls in the subscales of TCI can be seen in Figure 2. The effect sizes in the pooled forest plot with 95% CI, mean and range are presented. Also, the Z-standard score and its statistical significance are presented for those subscales that have significant differences. As can be seen, HA is significantly higher in patients with insomnia (z=3.346, p=0.001). In other subscales of TCI, there was no difference between the cases and controls (p>0.05). The publication bias was not seen for any of the variables (p>0.05 for Egger’s statistic).
Correlations Between the Subscales of PANAS and Sleep Disorders
As can be seen in Table 3, according to the results of all four available studies, NA is the strongest correlates of daytime sleepiness. Also, the relationship between NA and PA with PSQI was confirmed in 87% and 73.9% of studies, respectively. Figure 3 presents the correlations between the subscales of PANAS and sleep quality assessed by PSQI. As can be seen, there is a positive significant relationship between NA and poor sleep quality assessed by PSQI (z=10.437, p<0.001). Besides, PA and poor sleep quality assessed by PSQI are negatively related (z=-6.826, p<0.001). The publication bias was not seen in the relationship between any of the variables (p>0.05 for Egger’s statistic).
Correlations Between the Subscales of TCI and Sexual Dysfunctions
Table 3 shows the correlation between the subscales of TCI and sexual dysfunction. As can be seen, according to the results of all three available studies, HA and SD are the strongest correlates of sexual dissatisfaction/dysfunction. Probably, there is a positive significant relationship between HA and sexual dissatisfaction/dysfunction. Also, SD is likely negatively related to sexual dissatisfaction/dysfunction. Because of the limited number of available studies, the relationship between the other subscales of the TCI and sexual dysfunctions is ambiguous. The publication bias was not seen in the relationship between any of the variables (p>0.05 for Egger’s statistic).
Discussion
The study aimed to systematically review and meta-analysis the psychobiological theories of personality in sleep disorders and sexual dysfunction. The present study found that, in general, relatively few studies have examined the relationship between personality psychobiological models, particularly TEMPS and AFECTS, and sleep disorders and sexual dysfunction. However, the relationship between insomnia and TCI/TPQ, as well as the relationship between sleep quality and PANAS, has been considered in more studies. In the following, the relationship between each of these models and sleep and sexual disorders is discussed separately.
Insomnia: The results of the present meta-analysis showed that the HA in patients suffering from insomnia is higher than healthy controls. Although few studies met inclusion criteria for the meta-analysis (12,30,41,46), other studies in the present review have also pointed to the relationship between these two components (29,51). People with high HA have traits such as fear, embarrassment, and pessimistic worries and are easily tired, which can play a role in insomnia (12,41). High HA is strongly associated with anxiety and depressive disorders (17,72) and the symptoms of anxiety and depression can lead to insomnia (73). In fact, insomnia may be a marker for undiagnosed anxiety and depression (41). In addition to the above, the role of genetics and neurobiological mechanisms can also be discussed. Some components of HA, including fatigability (HA4) and anticipatory worry (HA1), are associated with reduced REM sleep duration and REM delay and these REM sleep disorders are biological markers of depression in patients with insomnia (41).
Sleep quality: Considering the very few studies related to sleep quality using TCI, in the present meta-analysis, the relationship between PA/NA evaluated using PANAS with this component was investigated (24,38-40,42,53,55,56,63). The current results showed that there is a positive relationship between NA and poor sleep quality. Also, PA has a negative relationship with poor sleep quality. PA and NA are the most important correlations of neuroticism traits in the five-factor model (74). PA/NA and neuroticism are highly correlated with symptoms of anxiety and depression (74,75). Other studies have suggested a link between NA components including fear, distress, guilt, anxiety, sadness, frustration, fatigue, embarrassment, and hostility with sleep problems (76-78). Therefore, it seems that poor sleep quality, which indicates poor quantity and quality of sleep, is affected by high NA and low PA. This is likely to lead to daytime sleepiness. Some studies have suggested a link between poor sleep quality and daytime sleepiness (24,38,79).
Other sleep disorders: There were very few studies on parasomnias, RLS, sleep breathing disorders, and daytime sleepiness. Only two studies covered RLS (13,48), both of which indicated a positive relationship between HA and RLS. Therefore, it is uncertain whether high HA is an important component in RLS. Two studies also have noted a significant relationship between low SD and RLS/parasomnias (13,52). Two studies reported a significant relationship between NS and parasomnias/sleep breathing disorders (52,54). On the other hand, several studies have suggested a significant relationship between NA/PA and parasomnias/daytime sleepiness (24,31,32,38,55). Despite the reported relationships between psychobiological models and sleep disorders in the recent section, achieving more accurate results requires meta-analysis on a large number of studies.
Sexual dysfunction: Few studies in this area, as well as the wide range of types of sexual dysfunction, made it difficult to conduct an integrated analysis. Previous studies have sporadically examined the role of personality components in sexual dysfunction/dissatisfaction (33,34,70), sexual distress (33), sexual experiences and arousal (23,34), and premature ejaculation (67). Only three studies covered sexual dysfunction/dissatisfaction, all of which indicated a positive relationship between this component and HA (33,34,70). Also, all three studies reported a significant relationship between sexual dysfunction/dissatisfaction and SD. Therefore, based on the results of the current review, there is an uncertain possibility that HA and SD are two effective components in dysfunction/sexual dissatisfaction. Despite the complexity of the issue, higher levels of HA indicate anxious tendencies, and may this component is a causal factor for sexual dysfunction through a type of serotonergic deficiency (33,34). On the other hand, Cloninger introduces the character as conscious consciousness programs that consist of three individual (SD), social (C), and global (ST) cognitive sets. Unlike the temperaments that are considered as habitual systems and unconscious automatic reactions, characters are the mental and cognitive dimensions that provide a conceptual insight for the individual (80). Disorder in the characters, especially SD, can lead to the development and persistence of mental disorders (81,82).
Study Limitations
Current systematic review and meta-analysis in the field of the relationship between personality psychobiological theories and sleep disorders/sexual dysfunction is a pioneering study. However, there were some limitations. The lack of numerous studies related to TEMPS and AFECTS has prevented us from doing meta-analyses and any summary and scientific conclusion. This limitation should not lead to the neglect of the suggested models by Akiskal et al. (21) and Lara et al. (22) in future studies. Therefore, if researchers have access to studies related to these models, future reviews and meta-analyses can cover them. A small number of studies related to TCI/TPQ and insomnia and examined small sample sizes can be effective in reducing the validity of the results. The wide range of types of sexual dysfunction and related factors such as dissatisfaction, distress, embarrassment, disgust, and sexual arousal also made it difficult to draw an accurate comprehensive conclusion. This heterogeneity between variables and the lack of case-control studies focused exclusively on one special disorder can also be a serious challenge to generalizing the findings to a variety of sexual disorders. If more studies are available in the next decade, analyzes can also take into account age and sex differences. Participants in the studies that were entered into the meta-analysis of sleep quality assessed PSQI were a combination of the clinical and general population. Although PSQI can be used as a dimensional instrument (without a cut-off point), the use of case-control studies containing established poor sleepers can provide more accurate results. Another limitation is related to psychiatric comorbidities with sleep disorders, which were not rule-out in the present meta-analyses. Therefore, repetition of the present study in the coming years, if there are enough case-control studies for meta-analysis, can provide more accurate and valid findings for specialists and clinicians.
Conclusions
Studies of sleep disorders and sexual dysfunctions based on personality psychobiological models, especially case-control studies, are very few. Although, the bigger challenge is the wide range of these disorders (insomnia, sleep breathing disorders and its sub-components, central hypersomnolence disorders, parasomnias, sleep-related movement disorders, circadian rhythm sleep-wake disorders, poor sleep quality, and sexual disorders related to the desire, arousal, orgasm, painful intercourse and vaginismus, premature ejaculation, and paraphilias) that make it difficult to draw a comprehensive meta-analysis and conclusion. The affective and emotional temperaments assessed by TEMPS-A and AFECTS have largely been ignored. Considering the importance of biomarkers in the pharmacotherapy of sleep/sexual disorders, future studies are encouraged to examine the relationship between the two models and the diagnosis of these disorders.
The models measured using TCI and PANAS, especially concerning sleep disorders, have produced relatively more studies. According to the findings of the present meta-analysis, NA/PA and HA may play a key role in sleep quality and insomnia, respectively. At lower levels of confidence from the current review, HA and SD along with NA may be associated with other sleep disorders. Also, HA and SD are likely to be effective components in sexual dissatisfaction/dysfunction.
Although there was no publication bias related to the sleep/sexual disorders in the meta-analyses for insomnia and sleep quality, low evidence levels were seen in approximately 90% of sleep and 80% sexual dysfunction studies. Small sample size, non-randomized sampling, and internet-telephone based assessment were among the main limitations of the studies included in this review. Therefore, although the temperament traits assessed using the TCI and PANAS can potentially be important, to obtain findings with more acceptable validity, a future meta-analysis may provide valuable findings if access to a greater number of case-control studies and a focus on age and sex differences.
Acknowledgments
This article is part of a Ph.D. thesis of Mr. Saeid Komasi entitled “Review and reconsideration of Hierarchical Taxonomy of Psychopathology (HiTOP) model with an emphasis on psychobiological models in the explanation of somatoform disorders”, financial support and sponsored by the Neurosciences Research Center and the Vice-Chancellor for Research & Technology of Kurdistan University of Medical Sciences, Sanandaj, Iran (ID: 98724).
Ethics
The project was approved and received a code of ethics by the Kurdistan University of Medical Sciences, Sanandaj, Iran (IR.MUK.REC.1398.169).
Peer-review: Internally peer-reviewed.
Authorship Contributions
Concept: F.R., A.H., K.R., S.K., Design: F.R., A.H., K.R., S.K., Data Collection or Processing: F.R., A.H., K.R., S.K., Analysis or Interpretation: F.R., A.H., K.R., S.K., Literature Search: F.R., A.H., K.R., S.K., Writing: F.R., A.H., K.R., S.K.
Conflict of Interest: The authors are not aware of any conflict of interest.
Financial Disclosure: The project was funded by the Kurdistan University of Medical Sciences, Sanandaj, Iran (ID: 98724).