The present study partly supported the notion that some "unfavorable" parenting styles have the potential to predict a poorer disease status of a child's asthma and identified several specific types of a mothers' chronic stress and specific patterns in terms of stress coping behavior that may predict a poorer or better status of her child's asthma. In addition, it showed that these associations between a mother's psychosocial properties and her child's asthma are considerably different according to the child's age.
Previous studies suggested that a child's more severe asthmatic status was associated with its mother's more critical attitudes to the child [10
], and to higher parenting attitudes of passive rejection, active rejection, or doting [11
]. However, these studies were based on retrospective data, and interpretation of causality is rather difficult. The present study showed that for children aged < 7 years, a mothers' high active rejection (standardized partial correlation coefficient β = .22, P = .008) or interference (β = .17, P = .034) score was correlated with more severe disease at baseline
, but not at follow-up (Table ). However, for the children aged 7 years and over interference and inter-parental inconsistency, which were not correlated with the baseline severity (data not shown), were positively associated with the severity at follow-up
, suggesting the importance of prospective observation in examining such relationships.
The present study identified some factors related to the chronic stress and stress-coping behaviors of mothers that are associated with a child's asthmatic status in the subsequent year. Reportedly, the disease severity of an asthmatic child may decrease the asthma-associated quality of life of parents [27
], and a child's hospitalization in the past 12 months may increase a care-taker's depression or anxiety [28
]. In the present study, however, none of the maternal stress factors significantly related to a child's asthma at follow-up was materially associated with the severity at baseline (data not shown), except for a positive correlation with unfulfilled needs for acceptance for children aged < 7 years (β = .21, P = .011). Moreover, the associations between a mother's stress at baseline and her child's disease severity at follow-up remained after controlling for the severity at baseline. It is likely that the mothers' stress at baseline affected her child's disease status in the subsequent year. Wright et al reported, in a prospective study that followed asthmatic infants aged 2-14 months, that while parental stress predicted a child's wheezing, a child's wheezing did not affect parental stress [29
For the children aged < 7 years of the present study, a mother's stress/coping behaviors, especially chronic irritation and anger represented by the SI object dependence of anger and annoying barrier scales, and emotion-suppressive coping behavior represented by the SI unfulfilled needs for acceptance and altruism scales, were predictive of a more severe disease in the subsequent year. Because adjustment for adherence did not change these associations, it is unlikely that the association between a mother's stress and her child's disease was mediated by the mother's impaired care-taking functions. A mothers' stress (or wellbeing) may be verbally or non-verbally conveyed to her child, and affect the child's asthmatic status via a psycho-physiological pathway, such as by immunoreactivity to allergens or a vulnerability to airway infections. Research shows that while sadness induces cholinergic airway contraction, happiness causes airway relaxation [30
] and that cortisol response to psychosocial stress is blunted in asthmatic children [31
For the children aged 7 years and over, a child's asthma was not associated with the mother's stress relevant to chronic irritation or emotional suppression, suggesting that as children grow their allergic conditions become less affected by communication with their mother, even those who are under chronic stress. However, a higher egoism score was inversely related to the severity at follow-up, which was a rather unexpected finding. Among the parenting styles, a strong tendency toward maternal interference was related to a poor prognosis for asthmatic children. These associations were not altered by adjustment for adherence. For older children, a mother's egocentric, self-defensive behaviors and lack of excessive interference may rather assist her child in the development of autonomy, which in turn psycho-physiologically leads to the stabilization of allergic reactions.
The present study is subject to several limitations. First, the outcome variable disease severity was based on the mother's report of the intensity and frequency of her child's attacks in the past year. Our preliminary investigation indicated that the reports of mothers, most of whom had taken a program to learn about the management of their children's asthma, were in good agreement with the doctors' evaluation, supplemented by information from medical charts, yet they may include some errors originating from the quality of the mother's memory and interpretations. Random errors could have attenuated or even masked the association between the true disease status and the mother's parenting attitudes, stress, and coping styles. An example of a non-random, systemic error would be that a more distressed mother may overestimate her child's asthmatic status or rather become less aware of her child's attacks. This is unlikely, however, because the baseline correlation between the children's disease severity and factors representing the mothers' chronic stress was limited, if any. Second, this study only assessed the parenting styles and stress of mothers, not of fathers. Although all the participating mothers were those who regularly accompanied their children to the hospital and can thus be considered an important care-taker, some fathers may have been even stronger than the mother in affecting the child's disease status. Third, we discussed above the psycho-physiological pathway of the children as a possible linkage between a mother's stress and her child's disease, but this study lacks relevant information such as the children themselves' stress/wellbeing or asthma-associated immunological parameters. Wolf et al reported that parental stress and depression predicted increases in children's inflammatory profiles (eosinophil cationic protein and stimulated interleukin-4 production) over a six month period and that this pattern was not mediated by the child depression and anxiety [32
]. Fourth, the limited number of subjects, especially girls, did not allow an analysis stratified by both age category and sex. Future studies with a larger sample of girls should be done for each age cohort to address possible sex differences. Fifth, correlation coefficients between the baseline scores and the scores at follow-up ranged between 0.55 and 0.70 for most of the T-PCRp and SI scales in the present sample, suggesting that the parenting attitudes and stress/coping behaviors that they measure can be dynamic to some degree. Future studies with repeated measures with an interval (e.g., three months) for both psychosocial and outcome variables could address the question how temporal changes
in a mother's psychosocial factors are associated with their child's disease course.
The present study is among few studies that prospectively tested the hypothesis that certain parenting styles affect the clinical status of asthmatic children or that prospectively addressed the role of parental stress and coping behaviors on the disease status of asthmatic children [8
], or that showed how a mother's psychosocial properties may affect her child's asthma differently according to the child's age. We addressed a priori
hypotheses regarding specific parenting styles and parental stress and coping behaviors as risk factors for childhood asthmatic morbidity, used a fairly large sample with a high response rate, used validated instruments for psychosocial factors, and took into account a variety of known or potential confounding factors.