Background
Child maltreatment has been associated with poor emotional, cognitive, and physical health functioning both during childhood and later in life [
1‐
3]. Although child maltreatment exists in every culture and every society, research on child maltreatment focuses primarily on North American and European societies, and much less on Asian societies, especially those societies with limited resources [
4]. In addition, most studies are unable to control for comorbidity of different types of child maltreatment, because they did not investigate all types of maltreatment [
2,
3]. In the current study, we investigated the relation between five types of child maltreatment (sexual abuse, physical abuse, emotional abuse, witnessing parental conflict, and neglect) and child emotional, cognitive, and physical functioning while controlling for comorbidity among different types of child maltreatment in a representative sample in Vietnam.
The effects of different types of child maltreatment on emotional functioning have been studied frequently, mostly in Western high resource countries [
2,
5]. A review of seven meta-analyses on the association between child sexual abuse and mental health showed that child sexual abuse increases the risk of depression, anxiety, post-traumatic stress disorder (PTSD), and impaired self-esteem in adulthood [
6]. In addition, there is meta-analytic evidence for the association between physical abuse, emotional abuse, and neglect and a higher risk on depression and anxiety, and between physical abuse and PTSD and panic disorder [
2]. In a meta-analysis on exposure to domestic violence, a relation was found between experiencing domestic violence, and emotional problems including depression, anxiety, PTSD, and internalizing disorders in general [
5]. Moreover, there is empirical evidence that child maltreatment is linked to poorer interactions with peers [
7].
In addition, there is some, although limited, evidence for the association between child maltreatment and cognitive development. Small-sized studies, mostly with clinical samples or samples referred by Child Protection Services found that emotional abuse [
1], physical abuse, neglect, and witnessing domestic violence [
8] were related to poorer spatial working memory. Furthermore, physical neglect was related to poorer recognition memory [
1]. Physical abuse and sexual abuse were associated with impairments in verbal short-term memory [
9] and working memory for positive emotions [
10]. Controlling for comorbidity of other types of child maltreatment, sexual abuse was uniquely related to deficits in language and memory functioning [
11]. However, a study conducted with a nationally representative sample from the United States found no effect of physical abuse or sexual abuse on memory performance, although the timing of the first experience of sexual abuse did account for differences in memory performance [
12]. Compared with adults without sexual abuse experiences, adults with such an experience during early childhood performed better on a working memory test, while those who experienced sexual abuse only during adolescence performed worse [
12]. Finally, child maltreatment was associated with lower academic performance as a proxy of cognitive functioning [
13,
14].
Furthermore, child maltreatment seems to be related to physical health problems, although again, results are not always pointing in the same direction. A meta-analysis found that child maltreatment was associated with poorer adult health including neurological, musculoskeletal, respiratory, cardiovascular, gastrointestinal, and metabolic problems, more hospital visits, and more general health issues [
3]. However, most of the participants of the studies in this meta-analysis were females, restricting generalizability of the findings to this gender. Another meta-analysis exploring the long term health consequences of child physical abuse, emotional abuse, and neglect showed that the associations with some health problems were weak and inconsistent and the number of studies on some of the other health outcomes was insufficient to be meta-analyzed [
2]. In addition, there is meta-analytic evidence that physical, emotional, and sexual abuse are related to being overweight in adulthood [
15], and studies have shown that neglect and emotional abuse are related to underweight status [
16,
17] although this association is not always confirmed [
18]. Most studies on the physical consequences of child maltreatment assess physical problems in adulthood. It is unclear whether physical health problems as a result of child maltreatment are already visible in childhood. Finally, similar to research on the emotional and cognitive consequences of maltreatment, most studies did not examine the effects of all types of child maltreatment [
2,
3,
15]. Since about half of the maltreated children experience more than one type of child maltreatment [
19,
20], it is essential to control for the other types of maltreatment in order to know how specific types of maltreatment relate to outcomes.
There is little research on the consequences of child maltreatment in Asian societies, especially those societies with limited resources. Vietnam is a middle-income country with a culture highly influenced by Confucianism. In Vietnam, the saying “Spare the rod, spoil the child” describes the spirit of parenting practices. Physical and emotional punishment are considered necessary to raise well-behaved children in Vietnam as well as many other countries in Asia [
21]. Indeed, the prevalence of child maltreatment in Vietnam was found to be high, emotional abuse was most prevalent at 31.8%, followed by neglect (25.0%), physical abuse (19.1%), witnessing parental conflicts (15.3%) and sexual abuse (2.6%) [
22]. These prevalence rates are consistent with similar countries in Asia and Pacific regions [
23,
24], but compared directly to Western countries such as the Netherlands, most types of child maltreatment were much more common in Vietnam [
22]. The question is whether different types of child maltreatment have similar effects in Vietnam compared to Western cultures. In a study with an international sample, perceived normativeness of harsh discipline lowered the negative effects of harsh discipline [
25]. The differential effects of
maltreatment based on perceived normativeness have not been investigated yet. It may be that maltreatment has less severe effects as well in cultures where physical and emotional punishment are more normative.
In Vietnam, three studies on the association of child maltreatment with emotional and physical health functioning have been conducted [
20,
26,
27]. The results of these studies on emotional functioning were partially similar to those in Western cultures. When controlling for the other types of child maltreatment, emotional abuse was associated with depression, anxiety, and impaired self-esteem and neglect was related to depression [
26] and anxiety [
27]. Furthermore, sexual abuse was related to more depression, anxiety, and psychological distress while controlling for co-morbidity of physical abuse, emotional abuse, and neglect [
27]. However, no significant associations were found for physical abuse and emotional abuse with emotional problems, when controlling for neglect and sexual abuse [
27]. Regarding physical health outcomes of child maltreatment, a study in Vietnam showed that poorer perceived physical health was associated with emotional abuse for both boys and girls, and with physical abuse for girls only [
26]. However, these studies only used self-reported outcomes and were conducted with non-random samples or samples only representative of a small area in Vietnam.
The effects of child maltreatment may not be the same for boys and girls. A meta-analysis found larger effect sizes for the association between child maltreatment and health issues in samples with only females compared to samples including both males and females [
3]. However, meta-analyses on the association between child sexual abuse and adult mental health did not confirm significant moderation effects of gender [
28,
29].
In addition to gender, ethnicity may play a role in the effects of child maltreatment on the wellbeing of victims. The findings of studies with American samples showed that the effects of child maltreatment were more severe in African-American children than in children from other racial groups on the likelihood of being arrested for violence in young adulthood [
30], suffering from PTSD in adulthood [
31], and developing internalizing and externalizing behavior in both adolescence and adulthood [
32]. However, other studies found that the associations between child maltreatment and depression, heavy drinking, and violence were equivalent between Black and White adolescent and young adult men [
33]. Differences in the effects of child maltreatment based on ethnicity have not yet been examined in Vietnam. The Kinh is the ethnic majority accounting for 84% of the population. We will test whether the association of maltreatment with different outcomes is different for Kinh versus non-Kinh populations.
In sum, the current study aims to (1) examine the association between sexual abuse, physical abuse, emotional abuse, witnessing parental conflict, and neglect and child emotional, cognitive, and physical health functioning in a representative sample in Vietnam using both self-report and more objective measures (weight, height, study ranking, digit span test); (2) test the moderation effects of gender and ethnicity in these associations.
Discussion
We showed that child maltreatment was related to different aspects of child wellbeing including emotional, cognitive, and physical health functioning. More specifically, controlling for all other types of maltreatment, we found that life time sexual abuse and neglect were related to emotional dysfuctioning, poorer working memory, and physical health problems. Past year and life time physical abuse was related to emotional dysfunctioning and physical health problems. Past year and life time emotional abuse was associated with emotional dysfuctioning. Life time prevalence of witnessing parental conflict was linked to emotional dysfunctioning. Contrary to our hypotheses, we found that life time emotional abuse was positively associated with academic performance and no associations were found for weight status.
All types of child maltreatment were associated with emotional dysfunctioning. This finding is consistent with previous research in mostly Western countries [
2,
5,
6,
51]. Regarding cognitive functioning, sexual abuse and neglect were associated with poorer working memory performance. This is in line with several other studies on memory functioning [
1,
8,
11]. Physical health was related to physical abuse, sexual abuse, and neglect. This finding extends previous findings on the effects of child maltreatment on physical health in adulthood [
2,
3] by showing that also during childhood, physical abuse, sexual abuse, and neglect were associated with poorer perceived physical health while controlling for the effects of other types of child maltreatment.
Comparing our results with previous studies on the effects of child maltreatment in Vietnam reveals both similarities and differences. Earlier research in Vietnam found negative associations of some specific psychological conditions, for example depression and anxiety, with emotional abuse [
26], sexual abuse [
27], and neglect [
26,
27]. The current study shows that - in addition to emotional abuse, sexual abuse, and neglect - physical abuse and witnessing parental conflict were also related to emotional dysfunctioning while controlling for the co-morbidity of other types of child maltreatment. In addition, previous research with Vietnamese students of similar age as our sample found that physical health problems were linked to physical abuse (for girls) and emotional abuse [
26]. Our results extend these findings in that we also found associations with sexual abuse and neglect. Our measure of physical health was a general one whereas Nguyen (2006) measured specific physical health problems, which may explain differences in effects. Contrary to this earlier study however, we did not find an association with emotional abuse which could possibly be explained by the fact that in the current study emotional abuse was measured with a single item, which may not be representative of the full range of emotionally abusive behaviors. In addition, our study is the first to investigate and find associations between child maltreatment and cognition in Vietnam. Taking these and our other findings together, we showed that the consequences of child maltreatment in Vietnam seem to be more extensive than was thought based on previous research. Considering the high prevalence of all types of child maltreatment in Vietnam, with prevalence rates up to 39.0% for physical abuse and 59.6% for emotional abuse [
22], child maltreatment is a very serious threat to the general health of the Vietnamese population.
One of the central mechanisms of the association between child maltreatment and emotional, cognitive, and physical health outcomes may be alterations in the stress system and brain areas regulating the stress system. There is a wide range of evidence that the chronic stress of child maltreatment can induce long term impairments of the hypothalamic-pituitary-adrenal axis (HPA axis), a core part of the human stress system [
7,
52]. In turn, changes in the functioning of the HPA-axis are related to mental health problems such as anxiety and depression [
53]. Moreover, abnormal cortisol levels as a result of an impaired stress response system were also associated with lower cognitive functioning [
54]. More specifically, meta-analytic results showed that child maltreatment was associated with reduced hippocampal volume, a brain area that plays an important role in cognition, primarily memory, and controls the regulation of the HPA axis [
55]. In addition, impaired functioning of the HPA axis was found to alter metabolic processes, enhance the susceptibility to inflammation, and impair counter-regulatory control of immune responses [
56], all putting individuals at risk for health problems. Another possible mechanism through which child maltreatment could influence emotional, cognitive, and physical health functioning is through an increase in exhibiting risky behaviors. Child maltreatment experiences were found to increase the practice of risky behaviors, such as substance abuse, smoking, physical inactivity, multiple sex partners, which in turn can impair emotional, cognitive, and physical health functioning [
57].
Regarding physical and emotional abuse, not only life time experience, but also past year experience was related to emotional dysfunction, whereas for the other types of maltreatment only life time experiences were related to the outcomes. This finding suggests that past as well as recent experiences of physical and emotional abuse in adolescence can have devastating effects, whereas for the other types of maltreatment, the occurrence earlier in development may be more relevant. A reason that we did not find an association between past year sexual abuse and emotional dysfunction could be the low prevalence of sexual abuse in the past year which may not have allowed to detect significant association.
Among all types of child maltreatment, life time physical abuse and neglect were related to the largest number of dysfunctional child outcomes. While physical abuse has been studied frequently, research on neglect is often “neglected” [
58]. Our finding on the effects of neglect is consistent with the results of Spinhoven et al. (2010) [
59] that neglect has as much or even more detrimental effects than other types of child maltreatment, even after controlling for co-morbidity. Noticeably, total items in the household as a proxy for poverty was also related to four out of six outcome measures: physical health, underweight status, working memory performance, and academic performance. This finding may add to the findings on neglect and provides empirical evidence for the negative effect of poverty on cognitive functioning and physical health during childhood.
In contrast to our hypotheses, emotional abuse was related to better academic performance. Samples of previous studies reporting worse academic performance as a result of emotional abuse, all originated from Western and Middle Eastern countries and none originated from an East Asian country [
13,
14]. In East Asian countries, such as Vietnam, the so called “tiger parenting” practice is more common than in countries in other regions [
60,
61]. “Tiger parenting” was first introduced by Chua (2011) who described it as a harsh parenting style that focuses on making children reach high academic achievements and compliance with family obligations [
62]. This parenting style is in line with the “achievement/adjustment paradox” observed in Asian American children, who both reach higher academic achievement and have more psychological problems [
63], something that was also confirmed by the positive association between emotional abuse and emotional dysfunctioning found in this study. “Tiger parenting” has rarely been studied with Vietnamese parents; a qualitative study with a few Vietnamese-Australian mothers did find that Vietnamese parents had an authoritarian parenting style and expected high education achievements of their children [
64]. Although there are no studies on “tiger parenting” conducted in Vietnam and our study did not measure parenting style, the association between emotional abuse and better academic performance on the one hand and worse emotional functioning on the other hand might be an indication of “tiger parenting” practice. As mentioned before, only one item was used to assess emotional abuse and therefore this association needs to be further investigated in future research.
Regarding overweight and underweight status, the current study did not find associations with any type of child maltreatment. Although previous findings on the effects of child maltreatment on underweight status are inconsistent, there is evidence that child maltreatment is related to more overweight in children referred to Child Protective Services [
65] and there is meta-analytic evidence concerning this relation in adult samples [
15]. The current study was conducted with a general population sample of school-aged children, and the effects of child maltreatment on overweight could be less explicit in population-based samples compared to children referred by Child Protective Services.
No significant results were found for the moderation effects of gender and ethnicity. A reason that we did not find moderation by ethnicity could be the differences between minor and major ethnicities in Vietnam mainly reflect differences in economic status, so after controlling for the effects of economic status and other background information differences, the effects of child maltreatment across ethnicities were similar. These findings imply that the effects of child maltreatment during childhood were equivalent across different genders and ethnicities after controlling for socioeconomic inequalities.
Some limitations of this study should be mentioned. First, the cross-sectional design of this study does not allow the examination of causality of child maltreatment on emotional functioning, cognitive functioning and physical health status. However, research with animal models and a longitudinal twin-study on child development provide strong evidence of causal effects of child maltreatment on different outcomes [
66,
67]. Second, the different informed consent procedure in Hanoi compared with other provinces resulted in a lower participation rate for this region. The use of different consent procedures in one sample could affect the validity of the results of the study. Third, in terms of measurement, the emotional abuse scale only consisted of one item which limits the reliability and validity of this scale. Moreover, overweight and underweight status were identified through height and weight reported by the participants. It might be difficult for children to remember their precise height and weight, thus recall bias may affect the accuracy of weight problems in this study. Another limitation is the fact that a large percentage of working memory test results were invalid or missing (26.5%). Although we used a reliable multiple imputation process, the large number of missing values may still have introduced bias in the data analysis relating to working memory.