Perceived inadequate care and excessive overprotection during childhood are associated with greater risk of sleep disturbance in adulthood: the Hisayama Study

The data was gathered from participants in the Hisayama Study in 2011. The Hisayama Study is an ongoing, long-term, cohort study done to examine cardiovascular disease and its risk factors in Hisayama, a suburban town adjoining Fukuoka City, a metropolitan area in southwestern Japan, in which annual health checks and surveys have been done since 1961 [28]. The present study was conducted as a cross-sectional sub-study of the Hisayama Study for which participants were recruited in 2011. Of the 2,250 residents aged 40 years or older who participated in the 2011 survey, 860 (38.2 %) consented to participate in this study. After excluding 158 residents without data from the parenting questionnaire available for both parents, 702 (31.2 %, 265 men and 437 women) were enrolled for study.

Sleep disturbance was assessed by the Pittsburgh Sleep Quality Index (PSQI) [29], a self-reported questionnaire that assesses sleep disturbance and sleep quality during the past month that is widely used in clinical and population based settings [30‐32]. It consists of 19 questions rated on a scale of 0–3 and is categorized into seven components: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction. The sum of the scores for these seven components yields a global score. We defined sleep disturbance as a global PSQI score greater than 5, which distinguishes patients from healthy controls with a high sensitivity and high specificity (89.6 % and 86.5 % of sleep disorder patients in the original version, 85.7 and 86.6 % of primary insomnia patients in the Japanese version) [29]. Internal reliability, test-retest reliability, and validity are acceptable in both versions [33].

Perceived parenting styles were measured using the Parental Bonding Instrument (PBI), a self-report questionnaire with 25 items that measures parenting styles in the first 16 years of life, as recalled by the respondents [16]. The PBI is scored separately for the father and mother to evaluate the relationship between the respondent and each parent, as they are subjectively perceived. The respondents are asked to score their parents’ attitudes and behaviors separately, using a 4-point Likert scale. Two subscales of parenting style are measured by the PBI: care and overprotection. The “care” subscale reflects perceived parental warmth, affection, and involvement contrasted with coldness and rejection. The “overprotection” subscale reflects perceived parental psychological over-control and intrusion contrasted with encouragement to psychological autonomy and exploration of the environment.

The parenting styles are divided into four categories (PBI quadrants) by dichotomized care and overprotection scores: “optimal bonding” (high care, low overprotection), “neglectful parenting” (low care, low overprotection), “affectionate constraint” (high care, high overprotection), and “affectionless control” (low care, high overprotection). “Affectionless control,” as proposed by Parker et al. [8] and confirmed by Sato et al. [34], is a maladaptive form of parenting that results in a particular vulnerability to the occurrence of psychopathology. The PBI score reflects the actual parenting attitude and is based on studies using corroborative witnesses and independent observers [35, 36]. The PBI has long-term stability [37] and its subscales have a high level of test-retest reliability and internal consistency [38]. The Japanese version of the PBI has been shown to have adequate validity [39].

A self-administered questionnaire concerning marital status, education, subjective economic status, occupation, current drinking, habitual smoking, habitual exercise, antihypertensive agent use, the use of insulin and oral glucose-lowering agents, existing pain, and past history of disease, including cardiovascular disease, cancer, respiratory disease and digestive disease, was completed by each participant and was checked by trained interviewers at the screening [28]. Body height and weight were measured in light clothing without shoes, and the body mass index (kg/m²) was calculated. Obesity was defined as a body mass index ≥ 25.0 kg/m². Blood pressure was measured three times after the subject had rested for at least five minutes in the sitting position. The mean of the three measurements was used for the present analysis. Hypertension was defined as a systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, and/or current use of antihypertensive agents [40]. Blood glucose was measured by the glucose oxidase method. Diabetes mellitus was defined as a fasting plasma glucose level of ≥ 7.0 mmol/L (126 mg/dL), a 2-h post-loaded or causal glucose level of ≥ 11.1 mmol/L (200 mg/dL), HbA1c (NGSP) ≥ 6.5 %, and/or current use of insulin or oral glucose-lowering agents [41]. “Depressive symptom” as a psychological factor was measured using a self-report questionnaire, the Patient Health Questionnaire-9 (PHQ-9), which has been shown to be a reliable and valid assessment tool for both the diagnosis of depression and the evaluation of depression severity in primary care settings [42]. The PHQ-9 assesses the symptoms of depression over the past two weeks using nine questions rated on a 4-point scale; 0 (not at all) to 3 (nearly every day). The sum of the nine items is calculated to obtain a depression score ranging from 0 to 27. Higher scores indicate increased severity of depression. The validity of the Japanese version of the PHQ-9 has been confirmed [43].

Comparisons of the care and overprotection scores of men and women were performed by Mann-Whitney U-test. We dichotomized care and protection scores according to median scores: low score for care < 28 for fathers and < 31 for mothers and high score for overprotection ≥ 8 for both fathers and mothers. To account for the interaction between the care and overprotection subscales, paternal and maternal bonding was classified into four quadrants: “high care, low overprotection” (optimal bonding), “low care, low overprotection” (neglectful parenting), “high care, high overprotection” (affectionate constraint), and “low care, high overprotection” (affectionless control). Comparisons of characteristics between the high and low parenting subscales (care and overprotection) were performed by a Student t-test for parametric continuous variables, a Mann-Whitney U test for non-parametric ordinal variables, or a Chi-square test for dichotomous variables. Odds ratios (ORs) and the confidence interval (CI) of the parenting styles on the presence of sleep disturbance were estimated by logistic regression analysis, with adjustment for sociodemographic and lifestyle factors (Model 1: age, marital status, educational level, subjective economic level, occupation, current drinking, current smoking, and habitual exercise), physical factors (Model 2: model 1 + obesity, hypertension, diabetes, past history of cardiovascular disease, past history of cancer, past history of respiratory disease, past history of digestive disease, and pain symptom) and a psychological factor (Model 3: model 2 + depressive symptom score). In addition, we performed sensitivity analyses that used the quartiles of the parenting scores in logistic regression analysis. The heterogeneity in the association of age (40–64 as middle age and 65–96 as old age) was assessed by adding the interaction term to the relevant logistic model.

As far as we know there are no studies of the relation between parenting style and the sleep disturbance of adults, but several studies have addressed the relationship between insecure attachment, which is known to be closely related to inadequate parenting (i.e. low care and high overprotection) [25], and sleep disturbance. These previous studies, in accord with ours, showed that insecure attachment is a risk factor for sleep disturbance and lower quality of sleep across the lifespan [23], but they did not directly evaluate parenting styles.

The present study demonstrated that the parenting style of the parent of the same sex (i.e. father for men and mother for women) is more likely to be related to sleep disturbance. Ohtani et al. demonstrated that healthy Japanese volunteers who experienced inadequate parenting from the parent of the same sex have higher interpersonal sensitivity scores, from an analysis of covariance test, than do those who experienced adequate parenting, whereas no such association was observed for inadequate parenting by the parent of the opposite sex. [44, 45]. Interpersonal sensitivity, which is affected by inadequate same sex parenting, may be related to sleep disturbance through a personality factor that induces interpersonal stress.

This study found different influences of the parenting style of parents of the same sex on sleep disturbance. For men, low paternal care was significantly associated with a higher likelihood of sleep disturbance, but this significant association was diminished after adjusting for depressive symptom. Previous studies have reported that low paternal care is associated with the treatment-resistant depression of outpatients with major depression and with the depression of male adolescent delinquents [46, 47]. These findings raise the possibility that the sleep disturbance of men with low paternal care can be mainly attributed to depression. In contrast, there was no evidence of a relation between low maternal care and the sleep disturbance of women.

Maternal overprotection was significantly associated with a higher likelihood of sleep disturbance for women, after adjustment for sociodemographic, lifestyle and physical factors, but further adjustment for depressive symptom attenuated this association. Previous studies suggested that maternal overprotection is related to psychological factors such as depression [20, 48], supporting the hypothesis that maternal overprotection is related to the sleep disturbance of women via depressive symptoms. By contrast, the association between high paternal overprotection and the sleep disturbance of men was significant even after adjustment for depressive symptom. Sadeh et al. reported that infants who fall asleep with significant parental involvement (i.e., while being held, fed, rocked, etc.) are more likely to have an increased number and duration of night waking than infants who fall asleep in their crib with minimal parental assistance. It is assumed that infants who fall asleep with excessive parental overprotection fail to develop their own self-regulation and soothing skills and, therefore, continue to rely on repeated parental intervention during the night. These cognitions and behaviors may be related to sleep habits across the lifespan and contribute to sleep disturbance in adulthood. Men with high paternal overprotection in childhood may have mechanisms other than depression that contribute to sleep disturbance.

The precise reason for the gender related discrepancy in the influence of parenting styles is unclear, but it is possible that men are more susceptible than women to the parenting style during childhood [49]. Further investigations will be necessary to clarify the reasons for the differences.

In this study, high care with high overprotection (affectionate constraint) was associated with sleep disturbance after controlling for physical health factors and depressive symptom, despite previous findings that high care is a protective factor for psychological and physiological diseases. Parker and Lipscombe have shown that the combination of high care and high overprotection induces hypochondria and a tendency to be dependent [50]. Therefore, people who have experienced high care and high overprotection may be likely to be overly sensitive and to complain of sleep disturbances. Further research will be needed to examine sleep disturbance that uses both subjective and objective measures (i.e. polysomnography) to determine the extent to which this association is due to reported or objective differences.

The present study demonstrated that the parenting style combination of low care - high overprotection (affectionless control) in childhood increases the risk of sleep disturbance in adulthood. In European studies of the epidemiology of mental disorders, the authors reported that the parenting style “affectionless control” was associated with anxiety disorder, a risk factor for sleep disturbance that could mediate the association [51, 52]. Selterman and Drigotas have reported that students with insecure attachment, which is known to be closely affected by parenting style (low care and high overprotection), experienced significantly more stress and conflict in their dreams compared with students with secure attachment [53]. Therefore, stressful dreams may contribute to the sleep disturbance of people who experienced low care - high overprotection parenting. The significant association between affectionless control and sleep disturbance disappeared after additionally controlling for physical factors and depressive symptom. Previous studies have reported that low care - high overprotection is associated with physical symptoms (i.e. chronic pain, respiratory disease, and inflammatory bowel disease) and depression [17, 20, 22, 48, 54]. These physical and depressive symptoms may mediate the association.

This study has several limitations. First, the assessment of parenting styles was based on a self-reported, retrospective measure. The PBI is a stable questionnaire that has been shown to have adequate test-retest reliability for a retrospective period of 20 years [37]. Because the PBI score reflects the actual parenting attitude, based on studies using corroborative witnesses and independent observers, it would seem to have an acceptable level of objectivity. However, we cannot rule out the possibility that current mood influenced the subjective appraisal of recalled events about parenting [55]. Additionally, if the participant could not recall environmental information in childhood, Osypuk et al. suggested that associations with health may be biased towards the null [56]. Second, we used only a self-reported questionnaire for measuring sleep disturbance. Although the PSQI questionnaire is widely used for assessing subjective complaints and sleep quality, future studies should add objective measures such as actigraphy. Third, the data are cross-sectional. We thus cannot deny the possibility of reverse causality; that sleep disturbance influenced the rating of the parenting styles. Prospective longitudinal studies are needed to clarify the contribution of inadequate parenting style to the development of sleep disturbance. Fourth, there is the possibility of selection bias because approximately two-thirds of the individuals who participated in the regular Hisayama Study survey did not participate in our research. In an analysis comparing the characteristics of subjects included and excluded, relatively healthy people were suggested to have been more likely to participate in the study than non-healthy people: subjects included in this study were younger and there was a lower frequency of current smokers, hypertensive patients, and a clinical past history of cardiovascular disease, cancer, and digestive disease than subjects excluded (Additional file 2: Table S2). Therefore, the generalizability of our findings to all individuals in the community may be limited. Nevertheless, we believe that our findings provide important information on the influence of inadequate parenting style on sleep disturbance. Finally, the parenting styles could be different by race and culture. Therefore, care must be taken when generalizing our results of the relation between parenting style and sleep disturbance to other countries and cultures, even though they are consistent with those of previous studies about the negative effect of low care and high overprotection on physical or psychological diseases done in other countries [17, 20, 48].