Significance

Previous studies have shown that social isolation is significantly associated with morbidity and mortality. One of the most pressing issues in Japan is child-rearing isolation of mothers with infants. Child-rearing isolation is considered a risk factor for abuse and disruption of healthy parent-child relationships. However, no studies have examined the associated factors of child-rearing isolation among mothers while adjusting for confounding factors. Our results provide evidence that child-rearing isolation is a risk factor for maltreatment. In addition, we found new evidence that maternal child-rearing isolation is significantly associated with age at birth, lack of spousal participation in parenting, and urban living.

Introduction

Social isolation has been identified as a risk factor for morbidity and mortality for a number of diseases (Elovainio et al., 2017; Leigh-Hunt et al., 2017). Therefore, preventing social isolation is one of the key challenges that public health faces.

Social isolation can occur in any age group; however, child-rearing isolation of mothers with infants is a pressing issue in Japan. In fact, child-rearing isolation has become a major issue to be addressed in three maternal and child health policies in Japan: countermeasures against adverse effects of the relationship between parents and children (Maternal and Child Health Division, Children and Families Bureau, Ministry of Health, Labour and Welfare, 2000), child abuse (Ministry of Health, Labour and Welfare, 2005), and birth rate decline (Cabinet Office, 2004). In previous studies, social isolation has been defined as the subjective feeling of one’s life lacking social connections or social support (Holt-Lunstad et al., 2015) and a situation in which one lacks anyone to consult regarding important matters (Brashears, 2011). Specifically, maternal social isolation has been defined as mothers lacking anyone to speak or consult with regarding child-rearing concerns when they struggle with child-rearing (Honda et al., 2019).

The lack of social support has been shown to be associated with maternal stress and depression (Mulvaney and Kendrick, 2005a) and not adopting safe practices for the prevention of childhood injury (Mulvaney and Kendrick, 2005b). Moreover, child-rearing isolation represents a risk factor for child abuse (U.S. Department of Health and Human Services, 2019), which has a deleterious impact on children’s development. In Japan, socially isolated mothers with 6-month-old infants spent JPY 4,186 (Japanese yen; around USD 35) more per month on child-rearing costs compared to non-isolated mothers (Honda et al., 2019), which suggested that parenting costs are higher due to lack of social support for mothers. Therefore, from a public health perspective, understanding the current situation regarding child-rearing isolation and its related factors is critical. However, to date, few studies in Japan have examined child-rearing isolation nationwide.

Therefore, this study aims to examine the relationship between child-rearing isolation and socio-demographic variables, including maternal age, child age, employment, socioeconomic status, spousal support, living municipality, and maternal psychosocial wellbeing using data from all prefectures in the Final Survey of Healthy Parents and Children 21 (Yamagata et al., 2014).

Methods

Setting and participants

Healthy Parents and Children 21 is a national campaign to promote maternal and child health in Japan (Osawa et al., 2019), for which a final survey was conducted to reveal new issues in 2014. This study used the data from the final survey as a secondary analysis of an existing dataset.

Participants were residents of 472 municipalities that had been identified as targets of the Final Survey of Healthy Parents and Children 21 and were also the parents of children scheduled for child health checkups during the survey period. Municipalities were divided into quartiles by population, and participants were randomly selected from each group. In total, 89,404 print questionnaires were mailed to the identified number of participants in each municipality. Surveys were collected from 75,622 parents whose children were receiving health checkups during the survey period: 3–4-month checkup, n = 20,729; 18-month checkup, n = 27,922; 3-year checkup, n = 26,971 (response rate = 84.6%). The surveys were mailed in February 2013 and were collected between April and August 2013. Surveys with missing variables were excluded. Therefore, the final sample size for this study was 69,337 mothers (Fig. 1).

Fig. 1
figure 1

Flowchart of study participants

Full size image

Sampling and data Collection

The Maternal and Child Health Section of each municipality requested that parents of young children who were scheduled for health checkups complete the “Parents and Children Health Survey Questionnaire,” which was collected at the children’s checkups. This questionnaire was created to clarify how much the mental and physical health of children and parents had improved for the identified issues surrounding maternal and child health in Healthy Parents and Children 21.

Included variables

The dependent variable in this study was maternal child-rearing isolation. Child-rearing isolation was measured by the question “Who is the person you [the mother] consult about your child-rearing in daily life?” Mothers were asked to choose from 11 options (multiple answers allowed), including my husband, the child’s grandparents, neighbor(s), friend(s), primary doctor, a public health nurse or midwife, nursery or preschool teachers, telephone consultation, the Internet, other, and nobody. We identified the mothers who were “isolated in child-rearing” as mothers who responded nobody to the aforementioned question. Mothers who selected any of the other ten options were defined as being “not isolated in child-rearing.” The use of the Internet, especially social networking sites, is associated with having social networks outside of one’s family and is related to the formation of weak ties at the neighborhood level (Hampton et al., 2011); therefore, we considered the Internet a potential consultant in this study.

The explanatory variables were selected based on existing studies regarding factors associated with child-rearing isolation in mothers with infants (Honda et al., 2019). To further explore the demographic and psychological factors associated with maternal isolation, we added the “anxiety about child-rearing” and “living municipality” items. The selected variables included the demographic characteristics of the mother, the age of the child, the environmental factors associated with child-rearing, and child-rearing anxiety. The selected explanatory variables included maternal age at childbirth (< 25, 25–29, 30–34, 35–39, and > 39 years), child’s age (3–4 months, 18 months, and 3 years), mother’s employment status (employed or unemployed), socioeconomic status (average or above average vs. difficult), husband’s participation in child-rearing (yes vs. almost none), municipality of residence (city, city designated by government, town/village, and special ward: Tokyo ward), time to relax with the child (yes vs. no), child-rearing confidence (confident vs. not confident), and concerns regarding abusing the child (yes vs. no).

Statistical analysis

A multivariate logistic regression analysis by means of a forced entry method was performed using complete data with no missing variables. Further, for sensitivity analysis, we accounted for missing data with multiple imputation by chained equations (MICE) for two hundred imputed datasets (White et al., 2011). The imputation model included all of the study variables. In each case, the odds ratio and 95% confidence intervals were calculated, and the statistical significance level was set at p < .05. Stata ver. 13 (StataCorp LP, College Station, Texas, USA) was used as the software program for all analyses.

Ethical considerations

As an ethical consideration, the questionnaires were completed anonymously with no forms of identifying information. As this study is a secondary analysis of an existing dataset with no access to personal identifiers, the requirement for informed consent was waived. Data analysis and publication of the results were approved by the Ethics Committee at XXX (receipt number: XXX).

Results

Demographic characteristics

Table 1 shows the participants’ characteristics. Of the 69,337 mothers who provided complete data, 160 (0.2%) reported child-rearing isolation. Among the mothers reporting child-rearing isolation, 26.9% (n = 43) were 35 to 39 years old, compared to only 20.4% (n = 14,094) of the non-isolated mothers. Moreover, 5.6% (n = 9) of the isolated mothers were 40 years old, compared to 3.3% (n = 2,264) of the non-isolated mothers.

Table 1 Participants’ Characteristics
Full size table

The majority of mothers who reported child-rearing isolation had 3-year-old toddlers (n = 85, 53.1%), while having 3-year-old toddlers was reported by only 35.1% of the non-isolated mothers (n = 24,256). Moreover, more than half of the isolated mothers were unemployed (n = 91, 56.9%), whereas in non-isolated mothers, the results were more evenly distributed, with slightly more mothers being employed (n = 35,945, 52.0%) than unemployed (n = 33,232, 48.0%). This finding demonstrates a reverse trend from the overall sample, as 48.1% (n = 33, 323) of all mothers were unemployed, while 56.9% (n = 91) of the isolated mothers were unemployed.

The overwhelming majority (n = 141, 88.1%) of isolated mothers lived in an urban environment (i.e., city, city designated by government ordinance, or special ward), which is analogous to their proportion of the total sample size, as they represented 85.6% (n = 59,187) of the non-isolated mothers. Mothers living in the special wards in Tokyo were over-represented among the socially isolated mothers, as they represented 1.8% (n = 1,219) of the non-isolated mothers but 6.3% (n = 10) of the isolated mothers.

Although 90.4% (n = 62,563) of the non-isolated mothers reported that the children’s fathers helped with child-rearing, only approximately half of the isolated mothers reported receiving assistance from the fathers (n = 82, 51.3%). In addition, 76.3% (n = 122) of the isolated mothers reported not having time to relax with their children, compared to only 29.3% (n = 20,273) of the non-isolated mothers. Further, 88.8% (n = 142) of the socially isolated mothers reported lacking confidence about their child-rearing, compared to only 70.9% (n = 49,027) of the non-isolated mothers. Moreover, 58.8% (n = 94) of the socially isolated mothers reported being concerned about abusing their children, compared to 24.3% (n = 16,812) of the non-isolated mothers.

Mothers who reported experiencing financial difficulties comprised only 31.3% (n = 21,685) of the non-isolated mothers but 56.3% (n = 90) of the isolated mothers; thus, mothers with financial concerns were overrepresented in the socially isolated group.

Logistic regression analysis results

A multiple logistic regression analysis was conducted. A significant overall regression equation was found, χ2(15) = 354.84, p < .001, with a McFadden pseudo R 2 = 0.16. Table 2 shows the results of the logistic regression analysis (univariate/multivariate).

Table 2 Results of Logistic Regression Analysis with “Isolation of Child-Rearing” as the Objective Variable
Full size table

The results of the multivariate logistic regression analysis showed that the mothers whose age at childbirth was 35 to 39 years old had significantly higher odds of reporting child-rearing isolation compared to mothers whose age at childbirth was 30 to 34 years old (OR = 1.6, 95% CI[1.0, 2.4], p = .036). Mothers who were unemployed (OR = 1.7, 95% CI[1.3, 2.4], p = .001), who reported financial difficulties (OR = 1.8, 95% CI[1.3, 2.5], p < .001), and whose husbands did not offer significant support in the child-rearing (OR = 5.7, 95% CI[4.2, 7.9], p < .001) showed significantly higher odds for experiencing child-rearing isolation compared to mothers in the reference groups (Table 2). Regarding the municipalities, mothers living in special wards (i.e., the wards of the Tokyo Metropolis) had significantly higher odds of reporting child-rearing isolation than mothers living in other cities (OR = 4.2, 95% CI[2.2, 8.3], p < .001). Mothers who reported having concerns regarding abusing their children (OR = 2.1, 95% CI[1.5, 2.9], p < .001) and having no time to relax with their children (OR = 4.5, 95% CI[3.1–6.7], p < .001) had higher odds of reporting social isolation compared to mothers who were not concerned about abusing their children and reported having time to relax with them. The differences reached statistical significance in most variables between mothers who were and mothers who were not included in this analysis. Details are presented in Appendix (1) However, in the sensitivity analysis, the results of the analysis of the complete data and the results of the MICE were similar. Details are presented in Appendix (2) Frequencies of complete and imputed variables are reported in Appendix 3 (imputed data sample size 75,622).

Discussion

We identified a number of variables that were significantly associated with child-rearing isolation based on the results of the multivariate logistic regression analysis (Table 2).

Association with maternal age at Childbirth

Mothers whose age at childbirth was 35 to 39 years old had significantly higher odds for child-rearing isolation compared to mothers aged 30–34 at childbirth, suggesting that being older at the time of childbirth is associated with child-rearing isolation. Previous research examined adolescent mothers’ isolation (Kim et al., 2017); however, few studies have investigated the association between advanced maternal age and the child-rearing environment. According to a report by the Organisation for Economic Cooperation and Development (OECD; OECD Family Database, 2019), most women giving birth to their first child are age 30 or older in OECD counties, and the average age of women at childbirth has increased by 2 to 5 years from 1970 to 2017. As advanced maternal age at childbirth will likely increase in the future, risk factors, including the child-rearing environment and pregnancy outcomes, associated with it should be explored in future research.

Association with maternal employment

Compared to working mothers, non-working mothers had significantly higher odds of reporting child-rearing isolation. In a previous study, the absence of opportunities for social interaction in the workplace was listed as one of the major structural causes of social isolation (Stewart et al., 2009), and researchers have identified paid jobs as a critical area of social contact and interaction (Gordon et al., 2006). Thus, unemployed mothers are presumed to have relatively fewer opportunities for social interaction and fewer social support resources compared to employed mothers.

Non-working mothers spend a greater percentage of their time with their children and have relatively fewer social interactions with others compared to working mothers, which might lead to child-rearing isolation. Although maternal employment serves as a factor to prevent child-rearing isolation, research has suggested that mothers must involuntarily quit their jobs in some situations to assume their roles as mothers (Mitsubishi UFJ Research and Consulting, 2015); thus, improvements in the working environment that promote mothers with infants returning to work while raising their children are desirable.

Socioeconomic status

Mothers who reported having financial difficulties had significantly higher odds of child-rearing isolation compared to mothers who reported having average or above-average finances. Previous studies have shown that people with a low income are at increased risk of social isolation (Gallie et al., 2003). As economic disadvantages such as poverty and unemployment are associated with shame and stigma, individuals with these challenges might restrict or avoid social contact (Lindsay, 2010).

Stewart et al. (2009) reported that individuals with low household incomes had extremely limited social support resources compared to people reporting high incomes. Structural factors contributing to social isolation for people living with a low income include lacking the financial means to access public facilities, lacking social interaction due to insufficient education and employment opportunities, avoiding family and colleagues due to poverty, and engaging in self-isolation when fearing the threat of criticism (Stewart et al., 2009). Isolated mothers spend more money on child-rearing than non-isolated mothers (Honda et al., 2019), which may lead to financial difficulties. Overall, these results may suggest that financial difficulties can result in child-rearing isolation, and child-rearing isolation leads to increased child-rearing cost, which creates a vicious cycle.

Association with Husband’s child-rearing

In a study conducted in the United States, the number of individuals who reported having a network of people who are not relatives whom they can talk with regarding important matters was one-third smaller than the number of people who reported such circumstances in 1985 (McPherson et al., 2006). In Japan, mothers consult primarily with their husbands about child-rearing on a daily basis (i.e., emotional support source; Yamazaki et al., 2018). Thus, given the current trends, the social ties with people who are not relatives have decreased, and intimate networks primarily comprise spouses. A similar change has been observed among mothers with infants, suggesting that when no spousal social support is available, mothers can easily become isolated. Therefore, lacking social support from husbands is a factor directly linked to child-rearing isolation in mothers.

UNICEF has indicated that although Japan’s parental leave system is substantial, the percentage of fathers taking paternity leave is low (UNICEF, 2019a). A report by the International Network on Leave Policies and Research (2019) reported that the percentage of fathers taking paternal leave between 2015 and 2018 varied across countries. The reasons for not taking paternal leave for Japanese men include a limited workforce and an unsupportive atmosphere in the workplace, which accounted for a high percentage of men who were hesitant to take time off (Cabinet Office, 2018). Much evidence has demonstrated that paternity leave increases fathers’ involvement in child-rearing, which is beneficial for both infants and mothers (UNICEF, 2019b), suggesting that paternity leave should be provided and supported by employers.

Associations with Place of Residence

Mothers living in Tokyo special wards had significantly higher odds of child-rearing isolation compared to mothers living in other places. Due to the unipolar concentration of Japan’s politics and economy, in Tokyo’s special wards, newly developed residential areas and high-rise apartments have increased. Previous research has been reported that mothers with infants consider establishing friendships and social support networks a “time-consuming process” in newer residential areas (Strange et al., 2014). Mothers living in a metropolis are more likely to increase their risk for child-rearing isolation until a support network is built in the metropolis.

The proportion of the world’s population living in urban areas is increasing. Currently, half of the world’s population lives in urban cities, and this percentage is expected to increase to 70% by 2050 (Kennedy and Adolphs, 2011). The urban environment has been found to increase the risk of mental disorders (Krabbendam and Van, 2005; Lederbogen et al., 2011; Pedersen and Mortensen, 2001) and the risks for hypertension, overweight, and diabetes (Eckert and Kohler, 2014). Although urban life might increase various health-related risks, little data has linked urban features such as isolation to population health. National and local governments should take into account the risk of isolation of mothers living in an urban environment in their maternal and child policies.

Association with child-rearing anxiety

Variables related to anxiety about child-rearing were significantly associated with child-rearing isolation. Child-rearing isolation represents a risk factor for child abuse (U.S. Department of Health and Human Services, 2019), and the parenting environment of having no one to talk to about the child has been reported to be significantly associated with child abuse (Mochizuki et al., 2014). Our research complements findings from previous studies, which have reported an association between child-rearing isolation and child abuse.

Increased anxiety has been shown to lower the desire for social contact (Sarnoff and Zimbardo, 1961). Thus, child-rearing anxiety, such as concerns regarding abusing the child, might lead mothers to avoid social contact. In that sense, a mother who has concerns about her own parenting may be isolated with no one to talk to, which may result in child abuse. Additionally, social isolation has been reported to be associated with increased mental health risks, including mood disorders (Chou et al., 2011). These mental health issues might mediate the association between child-rearing isolation and anxiety. Early detection and the prevention of child-rearing isolation might alleviate mother’s anxiety about child-rearing and prevent child abuse and mental health challenges. Therefore, developing child-rearing support systems that can prevent isolation of child-rearing mothers should be a priority.

Limitations and possibilities of this study

Social isolation is considered to be a lack of social support, but no consensus on this definition currently exists; therefore, our results are dependent on the definition used in this study. Missing data reduced the sample size in a multivariate logistic regression analysis, although the results were similar when using missing data imputed with MICE. The possibility of residual confounding cannot be completely ruled out in observational studies such as ours. This research followed a cross-sectional design; thus, drawing causal inferences is not possible, as the directionality between child-rearing isolation and the associated variables may be reversed.

This study’s main strength is that it clarified the actual situation of mothers’ child-rearing isolation in Japan and examined the associated factors using a large, nationally representative dataset. A large sample size substantially reduced the risk of random error and allowed us to have adequate power for analysis. In addition, while the child-rearing isolation percentage is low, child-rearing isolation still has detrimental effects on child development and maternal health, making this study extremely significant from an epidemiological standpoint.

Future perspectives

Mothers who are isolated in child-rearing may have many risk factors, such as poverty, child-rearing anxiety, lack of support from the spouse, and, particularly, child abuse; thus, identifying sources of support to eliminate child-rearing isolation is necessary. Therefore, the degree of fathers’ involvement in child-rearing is thought to contribute substantially to child-rearing isolation in mothers. Improvements in the work environment to allow both mothers and fathers to continue to work while raising their children are needed. Public health policies addressing these issues might reduce adverse health effect for mothers, children, and fathers. Therefore, to clarify the association between isolation and health outcomes, longitudinal and intervention studies are needed in the future.

Conclusions

We examined the characteristics of mothers reporting child-rearing isolation and the associated factors using data from a large-scale, nationwide survey in Japan. Child-rearing isolation was found to be associated with advanced age at childbirth, unemployment, financial difficulties, lack of the husband’s participation in child-rearing, anxiety about child-rearing, concerns about abusing one’s own child, no time to relax with one’s child, and living in Tokyo’s special wards.