Available online at www.sciencedirect.com
ScienceDirect Comprehensive Psychiatry 55 (2014) 283 – 289 www.elsevier.com/locate/comppsych
Association between maternal behavior in infancy and adult mental health: A 30-year prospective study Angela P. Fan a,⁎, Stephen L. Buka b , Russell O. Kosik c , Ying-Sheue Chen a , Shuu-Jiun Wang a , Tung-Ping Su a , William W. Eaton d a
School of Medicine, National Yang-Ming University, Taipei, Taiwan Department of Epidemiology, Brown University, Providence, RI, USA c Santa Clara Valley Medical Center, San Jose, CA, USA d Department of Mental Health, School of Public Health, The Johns Hopkins University, Baltimore, MD, USA b
Abstract Background: Existing theories suggest that the mother–infant relationship has a potentially significant effect on long-term adult mental health, but there are few empirical data to support this view. Even fewer prior studies have examined the specific dynamics of the mother– infant relationship and their association with adult mental health. Methods: A total of 1752 inner-city infants born between 1960 and 1965 were followed prospectively as a part of the Collaborative Perinatal Project (CPP) and the Johns Hopkins Pathways to Adulthood Study. Multiple observations of development and an extensive adult interview were performed. Maternal behavior was observed and systematically rated at the infant's 4-month pediatric neurological evaluation and at 8 months by a developmental psychologist. Factor analysis was used to organize the maternal behavior variables into different types of dysfunctional mother–infant relationships. Adult mental health was assessed at the follow-up interview, when the infant had reached the age of 27– 33 years, by the General Health Questionnaire (GHQ) and self-perception of current mental health. Results: There was a significant association between unsupportive maternal behavior at 8 months and subsequent poor adult mental health (Fisher's exact test, p = 0.026). There was no association between overly involved maternal behavior and poor mental health as an adult. After adjustment for potential confounding variables, the elevated rates of poor adult mental health in children of mothers who exhibited unsupportive maternal behavior at 8 months persisted (OR = 1.41 [95% CI = 1.00–1.97], p b 0.05). Conclusion: Infants who experience unsupportive maternal behavior at 8 months have an increased risk for developing psychological sequelae later in life. © 2014 Elsevier Inc. All rights reserved.
Abbreviations: CPP, Collaborative Perinatal Project; GHQ, General Health Questionnaire; JHCPS, Johns Hopkins Collaborative Perinatal Study; NCPP, National Collaborative Perinatal Project; OR, odds ratio; PAS, The Pathways to Adulthood Study; PTSD, posttraumatic stress disorder. Funding source: This study is supported in part by NIMH Grant MH14592 and MH53188, the Robert Wood Johnson Foundation, the William T. Grant Foundation, and the Johns Hopkins Population Center. Financial disclosure statement: There is nothing to disclose. Conflict of interest: The authors have no financial relationships relevant to this article to disclose. “What's Known on this Subject”: Existing theories suggest that the mother–infant relationship has a potentially significant effect on long-term adult mental health, but there are few empirical data to support this view. Even fewer prior studies have examined the specific dynamics of the mother–infant relationship and their impact on adult mental health. “What This Study Adds”: This is a 30-year prospective cohort study that includes 1752 children and their mothers. Our findings suggest that children whose mothers exhibit unsupportive attachment behaviors when they are 8 months old are at a significantly increased risk for developing poor mental health as an adult. ⁎ Corresponding author. P.O. Box 22072, Taipei, Taiwan 100, ROC. E-mail address: [email protected] (A.P. Fan). 0010-440X/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.comppsych.2013.08.024
284
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
1. Introduction Attachment, that is to say the secure bonding between infant and mother (or other regular caretaker) that takes place during the days, weeks, and months after birth, is believed to be an essential developmental step that lays the foundation for future social and emotional development [1,2]. According to many theoretical positions, when the mother responds lovingly and sensitively to the infant's needs, the infant develops a sense of basic trust that is essential for later learning and normative development [3]. Erikson [1] suggests that this basic trust is a vital aspect of personality development. Several researchers have provided support for the idea that warm and caring parent–infant interactions during the first few months of life, particularly in response to the infant's signals, is necessary for secure attachment, whereas a lack thereof may place the infant at risk for a range of negative psychological sequelae [3,4]. Weich et al. [5] in 2009 conducted a meta-analysis of 23 papers published between 1970 and 2008 examining the effects of poor parenting on adult mental disorders. They found that abusive relationships significantly increase the risk of depression, anxiety, and PTSD in the grown child, while maternal emotional unavailability significantly increases the risk of suicide in adolescents. Other researchers have shown that poor family interactions during childhood can negatively impact the mental health of youths [6]. These studies offer support to the conjectures that previous child psychologists have put forth, focusing on the mother as the cornerstone of a secure mother infant relationship. Certainly, there are a variety of factors that can contribute to a mother's inability to care for her infant as deeply as she might like. Mothers who are depressed or who have many young children and may be preoccupied with responsibilities at home or in the workplace may lack the ability or energy to respond to their babies [7,8]. There is empirical evidence suggesting that mothers who experience high levels of stress, especially if their babies are premature or difficult to care for, are less able than mothers of normal babies to relate to their babies in a sensitive manner [9–11]. There is also research demonstrating that by 11 to 17 months of age, infants of depressed mothers exhibit reduced activity in the right frontal area of the brain [12]. These findings raise the possibility that maternal behavior not only influences an infant's developing psychosocial functioning but also the development of his/her central nervous system [13]. There are few previous studies that have examined the long-term association between atypical patterns of maternal behavior and the subsequent mental health of the grown child. Further, most of these previous studies have been retrospective, short term, based on clinical samples, and have relied on subjective reports [5]. In this paper, we present results from 1752 offspring born in the early 1960s to innercity mothers enrolled in the Collaborative Perinatal Project (CPP), who were assessed prospectively in adulthood at age
27 to 33 years old as a part of the Pathways to Adulthood: A Three Generation Urban Study. 2. Methods 2.1. Sample The Pathways to Adulthood Study (PAS) had its origin in the Johns Hopkins Collaborative Perinatal Study (JHCPS), a part of the national Collaborative Perinatal Project (CPP) of the National Institute of Neurologic and Communicative Disorders and Stroke. Prospective data on first-generation mothers and their second-generation children are available. The CPP was a multi-institutional, trans-disciplinary, collaborative project that attempted to identify factors operating during the prenatal, perinatal, and early childhood periods that adversely influence subsequent neurological and cognitive development. Twelve university–medical centers cooperated in a single study design, which called for the systematic collection of data by way of the prospective observation and examination of approximately 60,000 pregnancies from the prenatal period through the first seven (or, at some institutions eight) years of life. Reports from the CPP have been summarized by Niswander and Gordon [14], Hardy et al. [15], and Broman et al. [16]. The Pathways Study sample (n = 2,694) was randomly selected from the 3006 JHCPS children born between January 1st, 1960, and July 31, 1965, who successfully completed either the 7- and/or 8-year follow-up examinations that were part of the CPP (99% received both). From July 1992 to January 1994, both mothers and their children participated in extensive follow-up interviews, bridging the period of time from when the children were age 7 to 8 until they were 27 to 33, and providing current outcome data. This report concerns the 1752 offspring who completed the full adult mental health interviews. The interview procedure, potential biases, missing data, and attrition from the sample have been discussed in detail elsewhere [17]. Some differences at birth were found between the offspring with complete adult interviews, those with partial data, and those not located [18]. Mothers of children who could not be located in adulthood were more often younger (p b 0.001), poorly educated (p b 0.001), unmarried (p b 0.001), and poor (p b 0.001) at the time of their children's birth than were mothers of those children who were interviewed. Children who were located but did not complete full interviews tended to have mothers with characteristics similar to mothers of those who were interviewed [18], but as the data are limited to surviving subjects, attrition is likely to have resulted in an underestimation of the effects of unsupportive maternal behavior, as it is those at greater risk of poorer outcomes who are preferentially lost to follow-up. Eighteen percent of the mothers were white, and 82% were black. Twenty-eight percent had completed a high school education at the time their child was born. Among the
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
285
Table 1 Characteristics of the study sample.
Variable Male White Maternal education (completed high school) G1 not married at G2 birth
Total, n (%)
Poor mental health, n (%)
Normal mental health, n (%)
p-Value
1752 803 (45.8) 319 (18.2) 483 (28.3) 488 (28.5)
279 118 (42.3) 69 (24.7) 56 (20.5) 84 (30.8)
1473 685 (46.5) 250 (17.0) 427 (29.8) 404 (28.0)
0.20 0.02⁎ 0.02⁎ 0.36
N Low attachment Overly involved Unhealthy behavior at 4 months G1 age at G2 birth Household income
1619 1619 1520 1752 1625
Mean (SD) 0.06 (0.16) 0.04 (0.13) 0.04 (0.11) 24.95 (7.18) 4115.43 (2060)
Mean (SD) 0.09 (0.20) 0.04 (0.13) 0.05 (0.14) 24.70 (7.25) 3963.32 (1993.60)
Mean (SD) 0.06 (0.16) 0.04 (0.13) 0.04 (0.11) 25.10 (7.16) 4144.22 (2061.35)
p-Value 0.04⁎ 0.66 0.24 0.39 0.19
⁎ A p-value b 0.05 was considered statistically significant.
children, 46% were male and 54% were female. Table 1 summarizes the characteristics of the study population (both children and mothers) at the time of the child's birth. 2.2. Measures Outcomes were measured by self-reports obtained through questionnaire data provided by the adult offspring during a follow-up interview conducted at age 27–33. At that time, subjects completed a General Health Questionnaire (GHQ-28) [19,20] and were asked about their selfperception of their present mental/emotional health. GHQ Mental Health Score in our study was defined as the sum of the seven questions related to depression on the GHQ. Because each question had a possible score of 1–4 points, subjects could obtain a possible score of 7–28. Self-assessment of present mental/emotional health was assessed by asking subjects, “At the present time, would you say your mental and emotional health are excellent, very good, good, fair, or poor?” Participants who scored in the top decile on the GHQ Mental Health Score (11 or higher) and/or who reported their current mental/emotional health as “fair” or “poor” were classified as having “poor mental health.” As a result, 279 subjects (15.9%) were labeled as having poor mental health. Maternal behavior data were collected by direct observations made during the child's CPP follow-up examinations at ages 4 months (three maternal behaviors rated) and 8 months (seven maternal behaviors rated). The 4-month pediatric examination was conducted between the 14th and 20th weeks of life by a pediatrician with special training in neurology or by a pediatric neurologist. In addition to the standard checkup of the baby, the doctor also evaluated the mother's behavior and her relationship with the child during the examination. Maternal behavior measurements included the following: (1) mother's responsiveness to the child's physical needs; (2) mother's focus of attention during the examination; and (3) mother's attitude toward the child's test performance. The 8-
month psychological examination was conducted by a child psychologist when the child was 7½ to 8½ months old. Maternal behavioral measurements included the following: (1) mother's expression of affection; (2) mother's verbal evaluation of the child; (3) mother's physical handling of the child; (4) mother's management of the child during the testing; (5) mother's reaction to the child's needs; (6) mother's reaction to the child's test performance; and (7) mother's focus of attention during the examination. Each behavior was initially rated on a scale from 1 to 5 with the midpoint reflective of adaptive functioning. Response options, however, did not routinely fall along a continuum, but instead the different poles often reflected qualitatively different behaviors [21]. For example, for the maternal behavioral measurement “Mother's reaction to the child's test performance,” one endpoint reflected the mother's indifference to the child's test performance, while the other reflected the mother's excessive defensiveness or criticism of failures in the child's test performance. Accordingly, consistent with previous analyses of these behavioral ratings (Donatelli et al.), two variables were derived for each original item, each utilizing a 3-point scale (0 = not true; 1 = somewhat true; 2 = very true). Thus, for example, for the original item “Mother's reaction to child's test performance,” two new variables were created, “Mother indifferent to child's test performance” and “Mother defensive or critical of failures in child's test performance” (see Appendix 1). An individual who was rated as highly defensive [5] on the initial item was rescored as “very true” [2] on the new “Mother defensive or critical of failures in child's test performance” variable and 0 (none) on the “Mother indifferent to child's test performance” variable. 2.3. Analysis These 20 maternal behavior ratings were entered into a factor analysis with varimax rotation, yielding three primary factors as described below. From these we calculated three scale scores of maternal behavior and used the t test to
286
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
month exam than mothers of subjects with normal adult mental health (p = 0.040). There were no significant differences between poor adult mental health and the other two factors, “Overly Involved” at the 8-month exam and “Unhealthy Behavior at Four Months.” The seven items that comprised the “Low Attachment” factor were further examined individually (Table 2). Of the seven, four items were significantly associated with the development of poor adult mental health at α = 0.05. They were “Mother was negative and harsh with the child, rarely or never using an affectionate term” (p = 0.04); “Mother unaware of or slow to recognize child's needs” (p = 0.05); “Mother made no spontaneous effort to facilitate the testing” (p = 0.02); and “Mother inconsiderate and/or awkward in handling the child” (p = 0.04). We further dichotomized the “Low Attachment” factor for each of the seven items (by assigning 1 for any positive answer and 0 for a negative answer) and examined the association between each of the items and poor adult mental health using the chi-square test and Fisher's exact test (data not shown). Results of both the chi-square test (p = 0.04) and Fisher's exact test (p = 0.026) offered further evidence that children whose mothers exhibited the characteristics contained in the “Low Attachment” factor had a significantly increased risk of developing poor mental health as an adult. Overall, 19.9% of children whose mothers exhibited “Low Attachment” developed poor mental health as an adult, but only 15.1% of children whose mothers did not exhibit “Low Attachment” developed poor mental health as an adult. The gender-specific distribution was also examined. Twenty-two percent of male children who experienced “Low Attachment” from their mothers developed poor mental health, but only 13.3% of male children who did not experience “Low Attachment” from their mothers developed poor mental health. Meanwhile, 18.1% of female children who experienced “Low Attachment” from their mothers developed poor mental health, but only 16.7% of female
examine whether certain dimensions of maternal behavior were associated with poor adult mental health. We also used the chi-square test and Fisher's exact test to examine the relationship between these three forms of maternal behavior and poor adult mental health in a categorical manner. Finally, we used regression analysis to control for the effects of several potential confounding factors including maternal age, education attainment, household income, marital status, gender, and race.
3. Results The 20 maternal behavior ratings were entered into a factor analysis with varimax rotation, yielding three primary factors (see Appendix 1). These three factors each represented a distinct dimension of atypical maternal behavior. The “Low Attachment” factor represented maternal behavior that was unsupportive of the child's needs at 8 months and included seven items characterized by indifference, rough handling, and criticism towards the child. The “Overly Involved” factor represented maternal behavior that was overprotective at 8 months and included seven items characterized by excessive pride, caution, and affection. Finally, the “Unhealthy Behavior at Four Months” factor included six ratings of maternal behavior characterized by either unsupportive or overly involved behavior at 4 months. As shown in Table 1, there were only modest sociodemographic differences between the 279 participants who reported poor adult health and the remaining 1473 subjects. The groups were comparable in terms of gender, household income, maternal marital status, and maternal age at birth. However, a higher proportion of white subjects reported poor mental health (p = 0.02) (Table 1). Mothers of subjects who reported poor adult mental health were significantly more likely to exhibit “Low Attachment” behaviors at the 8-
Table 2 The association between poor mental health and each of the low attachment maternal behavior variables. Low attachment maternal behavior variables
No = 0\; yes = 1 or 2
N
%Poor mental health
Relative risk
Fisher's exact test 1-side p-value
Harsh and negative expression
No Yes No Yes No Yes No Yes No Yes No Yes No Yes
1485 128 1502 113 1482 102 1471 146 1542 58 1556 59 1555 54
15.4% 21.9% 15.5% 21.2% 15.7% 22.5% 15.4% 22.6% 15.7% 17.2% 15.6% 25.4% 15.7% 24.1%
1.42
0.04⁎
1.37
0.74
1.43
0.05
1.47
0.02⁎
1.10
0.43
1.63
0.04⁎
1.54
0.08
Indifferent to child's test performance Unaware or slow to child's need Made no effort to facilitate the testing Negative and critical of child Inconsiderate in handling the child Centered attention on her own concerns
⁎ A p-value b 0.05 was considered statistically significant.
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
287
Table 3 Logistic regression models for the association between low attachment maternal behavior and poor mental health in the grown child. Characteristics G2 profile Race (black) Gender (male) G1 profile Maternal education (did not complete high school) G1 age at G2 birth (risk) G1 not married at G2 birth Household income Mother's behavior Low attachment
Model 1
Model 2
Model 3
0.57 (0.42–0.78)⁎
0.60 (0.44–0.83)⁎
0.57 (0.41–0.81)⁎ 0.90 (0.68–1.19)
1.51 (1.08–2.10)⁎
1.41 (1.00–2.00)⁎ 1.02 (0.76–1.37) 1.21 (0.86–1.69) 1.00 (1.00–1.00)
1.39 (1.00–1.92)⁎
1.41 (1.00–1.97)⁎
1.44 (1.04–1.98)⁎
⁎ A p-value b 0.05 was considered statistically significant.
children who did not experience “Low Attachment” from their mothers developed poor mental health. Finally, logistic regression results showed that even after adjusting for race, gender, mother's education, age, marital status, and household income, experiencing “Low Attachment” as a child was still significantly associated with the development of poor mental health as an adult (OR = 1.41 [95% CI = 1.00–1.97], p b 0.05) (Table 3).
4. Discussion Using the 30-year prospective Pathways to Adulthood Study, we examined data from 1752 children and their mothers to test the hypothesis that certain types of atypical maternal behavior are associated with poor mental health in adulthood. Our findings suggest that children whose mothers exhibit unsupportive attachment behaviors when the child is 8 months old are at a significantly increased risk for developing poor mental health as an adult. Further, children whose mothers are overly involved when the child is 8 months old do not have an increased risk of developing poor mental health. Factor analysis was used to create the three factors “Low Attachment,” “Overly Involved,” and “Unhealthy Behavior at Four Months.” Retention of these three factors only yielded a cumulative variance of 35.8%. Although retention of additional factors would have yielded a higher cumulative variance, it was decided to retain only these three factors because they were best representative of the hypotheses we wished to test, namely to examine how maladaptive maternal behavior affects the mental health of the adult child. However, maladaptive maternal behavior can come in two forms. It can either be indifferent, unloving, and unsupportive, or it can be overprotective and characterized by excessive affection and caution. Both opposing forms of maladaptive maternal behavior can have negative effects on the child. Although the indifferent and unsupportive form may intuitively seem to be more likely to negatively impact a child's mental health, excessive affection and caution can create an unhealthy amount of dependence. We felt it important to distinguish
these two forms of maladaptive maternal behavior. Because the 8-month exam was conducted by a child psychologist, who was skilled at understanding the differences between the opposing forms of maladaptive maternal behavior, the “Low Attachment” factor, a measure of unsupportive maternal behavior, and the “Overly Involved” factor, a measure of excessive affection, were both composed only of data collected at the 8-month exam. Because the 4-month exam was conducted by a pediatrician and not a child psychologist, the “Unhealthy Behavior at Four Months” factor was most accurate in merely representing that the physician felt something about the mother's behavior was abnormal at the 4-month exam. Further characterization of that abnormal behavior we felt was beyond the scope of the pediatrician's ability to perform in a short clinic visit. The outcome, poor mental health, was intended to be a broad measure of the mental status of the subjects. We therefore included two separate measures of mental health in this variable. The first was the GHQ Depression subscale. Because 10% is approximately the lifetime prevalence of depression in the general population, the bottom 10% of patients, those who scored 11 or greater on the GHQ Depression subscale, were classified as having poor mental health. The second was, the subject's self-assessment of their own mental health. Using this second measure allowed us to identify those subjects who felt as if they had poor mental health but did not score as such on the GHQ Depression subscale. Thus the addition of this second variable accounted for any lack of validity in the GHQ Depression subscale's ability to detect poor mental health. We therefore felt that the combination variable that included both the GHQ Depression subscale score as well as the self-assessment was best reflective of overall mental health. Because the maternal behavior measures, being those included in the NCPP dataset, are more than 40 years old, they therefore were not the newest and best ways to accurately measure maternal behavior. However, because their shortcomings likely led to an underestimation of the number of mothers who exhibited poor behavior, they were unlikely to have any effect on the results. Additionally, although lifetime mental conditions as diagnosed by a
288
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
medical professional would serve as a more accurate reflection of adult mental health, because such data were not collected by the NCPP, this was impossible, and therefore the poor mental health variable was used as the outcome measure. Finally, as in any long-term study, attrition played a significant role. In this study, it is probably equally as likely or even more likely that subjects with poor mental health left the study as it is that subjects who experienced poor parenting left the study. Thus attrition most likely led to an underestimation of the effect of maternal behavior on adult mental health. Therefore, the effect size, as determined by our study, is likely conservative.
5. Implications Early parental sensitivity to the infant's needs is of particular importance in setting the stage for the future physiological and psychological development of the child [3]. The quality of the attachment between mother and child is important for the overall social and emotional development of the child [11,22], including expectations of adult availability, problem solving, social relationships, and ability to cope with new or stressful situations [23]. The mother's own experience in childhood, her marital relationship, the quality of her social support network, and her own mental health, all affect her ability to appropriately care for her infant [7,9,24,25]. Pediatricians, family physicians, and other providers of health care to infants and children ought to be aware of these possible mechanisms that link a lack of adequate parenting and adult mental illness. Clinically, physicians and other healthcare staff must be attentive to mothers whose behavior may not be optimally supportive of their children's development due to lack of good parenting information and/or role models, being overwhelmed by depression and/or personal or family problems, lack of social support, or lack of basic resources. Referral for appropriate help may be indicated. On a preventive level healthcare workers should provide anticipatory guidance, counseling, and education to parents, which sensitive primary care providers should already do for children of all ages. Poor and unmarried mothers are more likely to respond inappropriately to their children, and so these are the mothers who should be targeted by prevention initiatives. Community programs are an excellent potential source of aid. Certain countries, such as Australia and New Zealand [26], offer far more in the way of health supervision and parenting education for mothers and children than other countries, such as the United States. In conclusion, maternal behavior toward the infant has a significant long-term impact on the child's risk of developing poor mental health in adulthood. Further this impact is limited to maternal behavior that does not promote secure bonding with the infant. That is to say, children of overprotective mothers do not see the same psychological sequelae that children of unsupportive mothers do. We suggest that primary care physicians involved in the care of young
children and their mothers be sensitized to this association so that they may be properly prepared to intervene.
Appendix 1. Factor analysis Step 1 Original variables. 4-month maternal behaviors X Mother's responsiveness to the child's physical needs Y Mother's focus of attention during the examination Z Mother's attitude toward the child's test performance 8-month maternal behaviors A Mother's expression of affection B Mother's verbal evaluation of the child C Mother's physical handling of the child D Mother's management of the child during the actual testing E Mother's reaction to the child's needs F Mother's reaction to the child's test performance G Mother's focus of attention during the examination Step 2 Recode variable (Condition 1: 1 = 2, 2 = 1, else = 0; variable label: “L.” Condition 2: 4 = 1, 5 = 2, else = 0; variable label: “H”). X X(L): Mother unaware of or slow to recognize child's physical needs X(H): Mother overly absorbed with child's physical needs Y Y(L): Mother centered her attention on the child Y(H): Mother centered her attention on her own concerns and/or problems extraneous to the situation Z Z(L): Mother indifferent to child's test performance Z(H): Mother defensive or critical of failures in child's test performance A A(L): Mother was negative and harsh with the child, rarely or never using an affectionate term A(H): Mother was overly affectionate both verbally and physically B B(L): Mother generally negative and critical of child B(H): Mother only positive about child and ignorant of less desirable behaviors C C(L): Mother inconsiderate and/or awkward in handling the child C(H): Mother extremely gentle and/or overly cautious in handling the child D D(L): Mother made no spontaneous effort to facilitate the testing D(H): Mother frequently disrupted the testing to help the child with given tasks E E(L): Mother unaware of or slow to recognize child's needs E(H): Mother immediately responsive to child despite no existence of needs F F(L): Mother indifferent to child's test performance F(H): Mother defensive or critical of failures in child's test performance G G(L): Mother centered her attention on the child G(H): Mother centered her attention on her own concerns and/or problems extraneous to the situation Step 3 Factor analysis. Variable
Factors 1
2
3
A(L)
0.728
0.023
F(L)
0.718
E(L)
0.683
− 0.018 − 0.089 0.050
0.113 0.103
Cumulative Cronbach's Name explained α variance 13.844%
0.741
Low attachment
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289 Step 3 (continued) Variable
Factors 1
2
3
D(L)
0.614
0.142
B(L)
0.539
− 0.010 0.067
C(L)
0.510
0.060
G(H) E(H)
0.166 0.657
F(H)
0.467 − 0.019 0.106
B(H) C(H)
0.005 0.010
0.592 0.571
A(H)
0.557
D(H)
− 0.060 0.086
0.532
G(L)
0.104
0.493
Y(H)
0.081 0.128
0.644
X(L)
− 0.013 − 0.123 0.169
− 0.095 − 0.010 0.803
0.563
Z(L)
0.180
X(H)
− 0.043 0.025
− 0.115 − 0.124 0.111
Z(H)
0.616
− 0.137 − 0.090 0.131 − 0.065 − 0.025 0.113 − 0.035 0.122
Cumulative Cronbach's Name explained α variance
[8] [9]
[10] 26.031%
0.634
Overly involved [11]
[12]
[13]
[14] 35.755%
0.307
Unhealthy behavior at 4 months
[15]
[16] 0.498 0.425
− 0.090 0.065 Model fit KMO = 0.686 Barlett test of sphericity: χ 2 = 5040.811⁎
Y(L)
[7]
Extraction method: factor analysis. Rotation method: Varimax with Kaiser normalization. ⁎ p b 0.05 was considered statistically significant.
[17]
[18]
[19] [20]
References [1] Erikson E. Childhood and society. 2nd ed. New York, NY: Norton; 1963 [Revised and Expanded. Edition]. [2] Main M, Kaplan N, Cassidy J. Security in infancy, childhood and adulthood. A move to the level of representation. Growing points of attachment theory and research. Monogr Soc Res Child Dev 1985;50:66-104. [3] Miller WB, Feldman SS, Pasta DJ. The effect of the nurturant bonding system on child security of attachment and dependency. Soc Biol 2002;49(3–4):125-59. [4] Hipwell A, Keenan K, Kasza K, Loeber R, Stouthamer-Loeber M, Bean T. Reciprocal influences between girls' conduct problems and depression, and parental punishment and warmth: a six year prospective analysis. J Abnorm Child Psychol 2008;36:663-77. [5] Weich S, Patterson J, Shaw R, Stewart-Brown S. Family relationships in childhood and common psychiatric disorders in later life: systematic review of prospective studies. Br J Psychiatry 2009;194:392-8, http:// dx.doi.org/10.1192/bjp.bp.107.042515. [6] Mellins CA, Brackis-Cott E, Dolezal C, Leu CS, Valentin C, MeyerBahlburg HF. Mental health of early adolescents from high-risk
[21]
[22] [23]
[24] [25]
[26]
289
neighborhoods: the role of maternal HIV and other contextual, selfregulation, and family factors. J Pediatr Psychol 2008;33(10):1065-75. Riley AW, Coiro MJ, Broitman M, Colantuoni E, Hurley KM, Bandeen-Roche K, et al. Mental health of children of low-income depressed mothers: influences of parenting, family environment, and raters. Psychiatr Serv 2009;60(3):329-36. Burke L. The impact of maternal depression on familial relationships. Int Rev Psychiatry 2003;15(3):243-55. Crnic KA, Greenberg MT, Robinson NM, Ragozin AS. Maternal stress and social support: effects on the mother–infant relationship from birth to eighteen months. Am J Orthopsychiatry 1984;54:224-35. Nagata M, Nagai Y, Sobajima H, Ando T, Honjo S. Depression in the mother and maternal attachment—results from a follow-up study at 1 year postpartum. Psychopathology 2003;36(3):142-51. Bosquet M, Egeland B. Associations among maternal depressive symptomatology, state of mind and parent and child behaviors: implications for attachment-based interventions. Attach Hum Dev 2001;3(2):173-99. Dawson G, Klinger LG, Panagiotides H, Hill D, Spieker S. Frontal lobe activity and affective behavior of infants of mothers with depressive symptoms. Child Dev 1992;63:725-37. Gómez FM. Early experiences and environment: their impact on neurodevelopment and the creation of mental processes. Vertex 2004;15:105-14. Niswander KR, Gordon M. The women and their pregnancies: the Collaborative Perinatal Study of The National Institute of Neurological Disease and Stroke. Edition. Philadelphia, PA: WB Saunders Company; 1972. Hardy JB, Drage JS, Jackson EC. The first year of life. The Collaborative Perinatal Project of the National Institute of Neurological and Communicative Disorders and Stroke. Edition. Baltimore, MD: The Johns Hopkins University Press; 1979. Broman SH, Nichols PL, Kennedy WA. Preschool IQ, prenatal and early developmental correlates. Edition. Lawrence Erlbaum Associates Publishers; 1975. Hardy JB, Shapiro S, Mellits D, Skinner EA, Astone NM, Ensminger M, et al. Self-sufficiency at ages 27–33 years: factors present between birth and 18 years that predict educational attainment among children born to inner-city families. Pediatrics 1997;99:80-7. Hardy JB, Astone NM, Brooks-Gunn J, Shapiro S, Miller TL. Like mothers, like child: intergenerational patterns of age at first birth and associations with childhood and adolescent characteristics and adult outcomes in the second generation. Dev Psychol 1998;34:1220-32. Goldberg D. Manual of the General Health Questionnaire. Edition. NFER Publishing Company; 1978. Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. Psychol Med 1979;9:139-4522. Hardy JB, Shapiro S. Pathway to adulthood: a three generation urban study, 1960–1994 (Baltimore, Maryland). The second generation children user's manual. The Johns Hopkins Medical Institution; 1997. Bowlby J. A secure base: clinical application of attachment theory. Edition. London, UK: Routledge; 1988. Drotar D, Sturm L. Personality development, problem solving, and behavior problems among preschool children with early histories of nonorganic failure-to-thrive: a controlled study. J Dev Behav Pediatr 1992;13:266-73. Bell SM, Ainsworth MD. Infant crying and maternal responsiveness. Child Dev 1972;43:1171-90. Main M, Hesse E. Adult lack of resolution of attachment-related trauma related to infant disorganized/disoriented behavior in the Ainsworth strange situation: linking parental states of mind to infant behavior in a stressful situation. Attachment in the preschool years: Theory, research and intervention. Chicago, IL: University of Chicago Press; 1990. p. 339-426. Rowe HJ, Fisher JR. The contribution of Australian residential early parenting centres to comprehensive mental health care for mothers of infants: evidence from a prospective study. Int J Ment Health Syst 2010;4:6.
ScienceDirect Comprehensive Psychiatry 55 (2014) 283 – 289 www.elsevier.com/locate/comppsych
Association between maternal behavior in infancy and adult mental health: A 30-year prospective study Angela P. Fan a,⁎, Stephen L. Buka b , Russell O. Kosik c , Ying-Sheue Chen a , Shuu-Jiun Wang a , Tung-Ping Su a , William W. Eaton d a
School of Medicine, National Yang-Ming University, Taipei, Taiwan Department of Epidemiology, Brown University, Providence, RI, USA c Santa Clara Valley Medical Center, San Jose, CA, USA d Department of Mental Health, School of Public Health, The Johns Hopkins University, Baltimore, MD, USA b
Abstract Background: Existing theories suggest that the mother–infant relationship has a potentially significant effect on long-term adult mental health, but there are few empirical data to support this view. Even fewer prior studies have examined the specific dynamics of the mother– infant relationship and their association with adult mental health. Methods: A total of 1752 inner-city infants born between 1960 and 1965 were followed prospectively as a part of the Collaborative Perinatal Project (CPP) and the Johns Hopkins Pathways to Adulthood Study. Multiple observations of development and an extensive adult interview were performed. Maternal behavior was observed and systematically rated at the infant's 4-month pediatric neurological evaluation and at 8 months by a developmental psychologist. Factor analysis was used to organize the maternal behavior variables into different types of dysfunctional mother–infant relationships. Adult mental health was assessed at the follow-up interview, when the infant had reached the age of 27– 33 years, by the General Health Questionnaire (GHQ) and self-perception of current mental health. Results: There was a significant association between unsupportive maternal behavior at 8 months and subsequent poor adult mental health (Fisher's exact test, p = 0.026). There was no association between overly involved maternal behavior and poor mental health as an adult. After adjustment for potential confounding variables, the elevated rates of poor adult mental health in children of mothers who exhibited unsupportive maternal behavior at 8 months persisted (OR = 1.41 [95% CI = 1.00–1.97], p b 0.05). Conclusion: Infants who experience unsupportive maternal behavior at 8 months have an increased risk for developing psychological sequelae later in life. © 2014 Elsevier Inc. All rights reserved.
Abbreviations: CPP, Collaborative Perinatal Project; GHQ, General Health Questionnaire; JHCPS, Johns Hopkins Collaborative Perinatal Study; NCPP, National Collaborative Perinatal Project; OR, odds ratio; PAS, The Pathways to Adulthood Study; PTSD, posttraumatic stress disorder. Funding source: This study is supported in part by NIMH Grant MH14592 and MH53188, the Robert Wood Johnson Foundation, the William T. Grant Foundation, and the Johns Hopkins Population Center. Financial disclosure statement: There is nothing to disclose. Conflict of interest: The authors have no financial relationships relevant to this article to disclose. “What's Known on this Subject”: Existing theories suggest that the mother–infant relationship has a potentially significant effect on long-term adult mental health, but there are few empirical data to support this view. Even fewer prior studies have examined the specific dynamics of the mother–infant relationship and their impact on adult mental health. “What This Study Adds”: This is a 30-year prospective cohort study that includes 1752 children and their mothers. Our findings suggest that children whose mothers exhibit unsupportive attachment behaviors when they are 8 months old are at a significantly increased risk for developing poor mental health as an adult. ⁎ Corresponding author. P.O. Box 22072, Taipei, Taiwan 100, ROC. E-mail address: [email protected] (A.P. Fan). 0010-440X/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.comppsych.2013.08.024
284
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
1. Introduction Attachment, that is to say the secure bonding between infant and mother (or other regular caretaker) that takes place during the days, weeks, and months after birth, is believed to be an essential developmental step that lays the foundation for future social and emotional development [1,2]. According to many theoretical positions, when the mother responds lovingly and sensitively to the infant's needs, the infant develops a sense of basic trust that is essential for later learning and normative development [3]. Erikson [1] suggests that this basic trust is a vital aspect of personality development. Several researchers have provided support for the idea that warm and caring parent–infant interactions during the first few months of life, particularly in response to the infant's signals, is necessary for secure attachment, whereas a lack thereof may place the infant at risk for a range of negative psychological sequelae [3,4]. Weich et al. [5] in 2009 conducted a meta-analysis of 23 papers published between 1970 and 2008 examining the effects of poor parenting on adult mental disorders. They found that abusive relationships significantly increase the risk of depression, anxiety, and PTSD in the grown child, while maternal emotional unavailability significantly increases the risk of suicide in adolescents. Other researchers have shown that poor family interactions during childhood can negatively impact the mental health of youths [6]. These studies offer support to the conjectures that previous child psychologists have put forth, focusing on the mother as the cornerstone of a secure mother infant relationship. Certainly, there are a variety of factors that can contribute to a mother's inability to care for her infant as deeply as she might like. Mothers who are depressed or who have many young children and may be preoccupied with responsibilities at home or in the workplace may lack the ability or energy to respond to their babies [7,8]. There is empirical evidence suggesting that mothers who experience high levels of stress, especially if their babies are premature or difficult to care for, are less able than mothers of normal babies to relate to their babies in a sensitive manner [9–11]. There is also research demonstrating that by 11 to 17 months of age, infants of depressed mothers exhibit reduced activity in the right frontal area of the brain [12]. These findings raise the possibility that maternal behavior not only influences an infant's developing psychosocial functioning but also the development of his/her central nervous system [13]. There are few previous studies that have examined the long-term association between atypical patterns of maternal behavior and the subsequent mental health of the grown child. Further, most of these previous studies have been retrospective, short term, based on clinical samples, and have relied on subjective reports [5]. In this paper, we present results from 1752 offspring born in the early 1960s to innercity mothers enrolled in the Collaborative Perinatal Project (CPP), who were assessed prospectively in adulthood at age
27 to 33 years old as a part of the Pathways to Adulthood: A Three Generation Urban Study. 2. Methods 2.1. Sample The Pathways to Adulthood Study (PAS) had its origin in the Johns Hopkins Collaborative Perinatal Study (JHCPS), a part of the national Collaborative Perinatal Project (CPP) of the National Institute of Neurologic and Communicative Disorders and Stroke. Prospective data on first-generation mothers and their second-generation children are available. The CPP was a multi-institutional, trans-disciplinary, collaborative project that attempted to identify factors operating during the prenatal, perinatal, and early childhood periods that adversely influence subsequent neurological and cognitive development. Twelve university–medical centers cooperated in a single study design, which called for the systematic collection of data by way of the prospective observation and examination of approximately 60,000 pregnancies from the prenatal period through the first seven (or, at some institutions eight) years of life. Reports from the CPP have been summarized by Niswander and Gordon [14], Hardy et al. [15], and Broman et al. [16]. The Pathways Study sample (n = 2,694) was randomly selected from the 3006 JHCPS children born between January 1st, 1960, and July 31, 1965, who successfully completed either the 7- and/or 8-year follow-up examinations that were part of the CPP (99% received both). From July 1992 to January 1994, both mothers and their children participated in extensive follow-up interviews, bridging the period of time from when the children were age 7 to 8 until they were 27 to 33, and providing current outcome data. This report concerns the 1752 offspring who completed the full adult mental health interviews. The interview procedure, potential biases, missing data, and attrition from the sample have been discussed in detail elsewhere [17]. Some differences at birth were found between the offspring with complete adult interviews, those with partial data, and those not located [18]. Mothers of children who could not be located in adulthood were more often younger (p b 0.001), poorly educated (p b 0.001), unmarried (p b 0.001), and poor (p b 0.001) at the time of their children's birth than were mothers of those children who were interviewed. Children who were located but did not complete full interviews tended to have mothers with characteristics similar to mothers of those who were interviewed [18], but as the data are limited to surviving subjects, attrition is likely to have resulted in an underestimation of the effects of unsupportive maternal behavior, as it is those at greater risk of poorer outcomes who are preferentially lost to follow-up. Eighteen percent of the mothers were white, and 82% were black. Twenty-eight percent had completed a high school education at the time their child was born. Among the
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
285
Table 1 Characteristics of the study sample.
Variable Male White Maternal education (completed high school) G1 not married at G2 birth
Total, n (%)
Poor mental health, n (%)
Normal mental health, n (%)
p-Value
1752 803 (45.8) 319 (18.2) 483 (28.3) 488 (28.5)
279 118 (42.3) 69 (24.7) 56 (20.5) 84 (30.8)
1473 685 (46.5) 250 (17.0) 427 (29.8) 404 (28.0)
0.20 0.02⁎ 0.02⁎ 0.36
N Low attachment Overly involved Unhealthy behavior at 4 months G1 age at G2 birth Household income
1619 1619 1520 1752 1625
Mean (SD) 0.06 (0.16) 0.04 (0.13) 0.04 (0.11) 24.95 (7.18) 4115.43 (2060)
Mean (SD) 0.09 (0.20) 0.04 (0.13) 0.05 (0.14) 24.70 (7.25) 3963.32 (1993.60)
Mean (SD) 0.06 (0.16) 0.04 (0.13) 0.04 (0.11) 25.10 (7.16) 4144.22 (2061.35)
p-Value 0.04⁎ 0.66 0.24 0.39 0.19
⁎ A p-value b 0.05 was considered statistically significant.
children, 46% were male and 54% were female. Table 1 summarizes the characteristics of the study population (both children and mothers) at the time of the child's birth. 2.2. Measures Outcomes were measured by self-reports obtained through questionnaire data provided by the adult offspring during a follow-up interview conducted at age 27–33. At that time, subjects completed a General Health Questionnaire (GHQ-28) [19,20] and were asked about their selfperception of their present mental/emotional health. GHQ Mental Health Score in our study was defined as the sum of the seven questions related to depression on the GHQ. Because each question had a possible score of 1–4 points, subjects could obtain a possible score of 7–28. Self-assessment of present mental/emotional health was assessed by asking subjects, “At the present time, would you say your mental and emotional health are excellent, very good, good, fair, or poor?” Participants who scored in the top decile on the GHQ Mental Health Score (11 or higher) and/or who reported their current mental/emotional health as “fair” or “poor” were classified as having “poor mental health.” As a result, 279 subjects (15.9%) were labeled as having poor mental health. Maternal behavior data were collected by direct observations made during the child's CPP follow-up examinations at ages 4 months (three maternal behaviors rated) and 8 months (seven maternal behaviors rated). The 4-month pediatric examination was conducted between the 14th and 20th weeks of life by a pediatrician with special training in neurology or by a pediatric neurologist. In addition to the standard checkup of the baby, the doctor also evaluated the mother's behavior and her relationship with the child during the examination. Maternal behavior measurements included the following: (1) mother's responsiveness to the child's physical needs; (2) mother's focus of attention during the examination; and (3) mother's attitude toward the child's test performance. The 8-
month psychological examination was conducted by a child psychologist when the child was 7½ to 8½ months old. Maternal behavioral measurements included the following: (1) mother's expression of affection; (2) mother's verbal evaluation of the child; (3) mother's physical handling of the child; (4) mother's management of the child during the testing; (5) mother's reaction to the child's needs; (6) mother's reaction to the child's test performance; and (7) mother's focus of attention during the examination. Each behavior was initially rated on a scale from 1 to 5 with the midpoint reflective of adaptive functioning. Response options, however, did not routinely fall along a continuum, but instead the different poles often reflected qualitatively different behaviors [21]. For example, for the maternal behavioral measurement “Mother's reaction to the child's test performance,” one endpoint reflected the mother's indifference to the child's test performance, while the other reflected the mother's excessive defensiveness or criticism of failures in the child's test performance. Accordingly, consistent with previous analyses of these behavioral ratings (Donatelli et al.), two variables were derived for each original item, each utilizing a 3-point scale (0 = not true; 1 = somewhat true; 2 = very true). Thus, for example, for the original item “Mother's reaction to child's test performance,” two new variables were created, “Mother indifferent to child's test performance” and “Mother defensive or critical of failures in child's test performance” (see Appendix 1). An individual who was rated as highly defensive [5] on the initial item was rescored as “very true” [2] on the new “Mother defensive or critical of failures in child's test performance” variable and 0 (none) on the “Mother indifferent to child's test performance” variable. 2.3. Analysis These 20 maternal behavior ratings were entered into a factor analysis with varimax rotation, yielding three primary factors as described below. From these we calculated three scale scores of maternal behavior and used the t test to
286
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
month exam than mothers of subjects with normal adult mental health (p = 0.040). There were no significant differences between poor adult mental health and the other two factors, “Overly Involved” at the 8-month exam and “Unhealthy Behavior at Four Months.” The seven items that comprised the “Low Attachment” factor were further examined individually (Table 2). Of the seven, four items were significantly associated with the development of poor adult mental health at α = 0.05. They were “Mother was negative and harsh with the child, rarely or never using an affectionate term” (p = 0.04); “Mother unaware of or slow to recognize child's needs” (p = 0.05); “Mother made no spontaneous effort to facilitate the testing” (p = 0.02); and “Mother inconsiderate and/or awkward in handling the child” (p = 0.04). We further dichotomized the “Low Attachment” factor for each of the seven items (by assigning 1 for any positive answer and 0 for a negative answer) and examined the association between each of the items and poor adult mental health using the chi-square test and Fisher's exact test (data not shown). Results of both the chi-square test (p = 0.04) and Fisher's exact test (p = 0.026) offered further evidence that children whose mothers exhibited the characteristics contained in the “Low Attachment” factor had a significantly increased risk of developing poor mental health as an adult. Overall, 19.9% of children whose mothers exhibited “Low Attachment” developed poor mental health as an adult, but only 15.1% of children whose mothers did not exhibit “Low Attachment” developed poor mental health as an adult. The gender-specific distribution was also examined. Twenty-two percent of male children who experienced “Low Attachment” from their mothers developed poor mental health, but only 13.3% of male children who did not experience “Low Attachment” from their mothers developed poor mental health. Meanwhile, 18.1% of female children who experienced “Low Attachment” from their mothers developed poor mental health, but only 16.7% of female
examine whether certain dimensions of maternal behavior were associated with poor adult mental health. We also used the chi-square test and Fisher's exact test to examine the relationship between these three forms of maternal behavior and poor adult mental health in a categorical manner. Finally, we used regression analysis to control for the effects of several potential confounding factors including maternal age, education attainment, household income, marital status, gender, and race.
3. Results The 20 maternal behavior ratings were entered into a factor analysis with varimax rotation, yielding three primary factors (see Appendix 1). These three factors each represented a distinct dimension of atypical maternal behavior. The “Low Attachment” factor represented maternal behavior that was unsupportive of the child's needs at 8 months and included seven items characterized by indifference, rough handling, and criticism towards the child. The “Overly Involved” factor represented maternal behavior that was overprotective at 8 months and included seven items characterized by excessive pride, caution, and affection. Finally, the “Unhealthy Behavior at Four Months” factor included six ratings of maternal behavior characterized by either unsupportive or overly involved behavior at 4 months. As shown in Table 1, there were only modest sociodemographic differences between the 279 participants who reported poor adult health and the remaining 1473 subjects. The groups were comparable in terms of gender, household income, maternal marital status, and maternal age at birth. However, a higher proportion of white subjects reported poor mental health (p = 0.02) (Table 1). Mothers of subjects who reported poor adult mental health were significantly more likely to exhibit “Low Attachment” behaviors at the 8-
Table 2 The association between poor mental health and each of the low attachment maternal behavior variables. Low attachment maternal behavior variables
No = 0\; yes = 1 or 2
N
%Poor mental health
Relative risk
Fisher's exact test 1-side p-value
Harsh and negative expression
No Yes No Yes No Yes No Yes No Yes No Yes No Yes
1485 128 1502 113 1482 102 1471 146 1542 58 1556 59 1555 54
15.4% 21.9% 15.5% 21.2% 15.7% 22.5% 15.4% 22.6% 15.7% 17.2% 15.6% 25.4% 15.7% 24.1%
1.42
0.04⁎
1.37
0.74
1.43
0.05
1.47
0.02⁎
1.10
0.43
1.63
0.04⁎
1.54
0.08
Indifferent to child's test performance Unaware or slow to child's need Made no effort to facilitate the testing Negative and critical of child Inconsiderate in handling the child Centered attention on her own concerns
⁎ A p-value b 0.05 was considered statistically significant.
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
287
Table 3 Logistic regression models for the association between low attachment maternal behavior and poor mental health in the grown child. Characteristics G2 profile Race (black) Gender (male) G1 profile Maternal education (did not complete high school) G1 age at G2 birth (risk) G1 not married at G2 birth Household income Mother's behavior Low attachment
Model 1
Model 2
Model 3
0.57 (0.42–0.78)⁎
0.60 (0.44–0.83)⁎
0.57 (0.41–0.81)⁎ 0.90 (0.68–1.19)
1.51 (1.08–2.10)⁎
1.41 (1.00–2.00)⁎ 1.02 (0.76–1.37) 1.21 (0.86–1.69) 1.00 (1.00–1.00)
1.39 (1.00–1.92)⁎
1.41 (1.00–1.97)⁎
1.44 (1.04–1.98)⁎
⁎ A p-value b 0.05 was considered statistically significant.
children who did not experience “Low Attachment” from their mothers developed poor mental health. Finally, logistic regression results showed that even after adjusting for race, gender, mother's education, age, marital status, and household income, experiencing “Low Attachment” as a child was still significantly associated with the development of poor mental health as an adult (OR = 1.41 [95% CI = 1.00–1.97], p b 0.05) (Table 3).
4. Discussion Using the 30-year prospective Pathways to Adulthood Study, we examined data from 1752 children and their mothers to test the hypothesis that certain types of atypical maternal behavior are associated with poor mental health in adulthood. Our findings suggest that children whose mothers exhibit unsupportive attachment behaviors when the child is 8 months old are at a significantly increased risk for developing poor mental health as an adult. Further, children whose mothers are overly involved when the child is 8 months old do not have an increased risk of developing poor mental health. Factor analysis was used to create the three factors “Low Attachment,” “Overly Involved,” and “Unhealthy Behavior at Four Months.” Retention of these three factors only yielded a cumulative variance of 35.8%. Although retention of additional factors would have yielded a higher cumulative variance, it was decided to retain only these three factors because they were best representative of the hypotheses we wished to test, namely to examine how maladaptive maternal behavior affects the mental health of the adult child. However, maladaptive maternal behavior can come in two forms. It can either be indifferent, unloving, and unsupportive, or it can be overprotective and characterized by excessive affection and caution. Both opposing forms of maladaptive maternal behavior can have negative effects on the child. Although the indifferent and unsupportive form may intuitively seem to be more likely to negatively impact a child's mental health, excessive affection and caution can create an unhealthy amount of dependence. We felt it important to distinguish
these two forms of maladaptive maternal behavior. Because the 8-month exam was conducted by a child psychologist, who was skilled at understanding the differences between the opposing forms of maladaptive maternal behavior, the “Low Attachment” factor, a measure of unsupportive maternal behavior, and the “Overly Involved” factor, a measure of excessive affection, were both composed only of data collected at the 8-month exam. Because the 4-month exam was conducted by a pediatrician and not a child psychologist, the “Unhealthy Behavior at Four Months” factor was most accurate in merely representing that the physician felt something about the mother's behavior was abnormal at the 4-month exam. Further characterization of that abnormal behavior we felt was beyond the scope of the pediatrician's ability to perform in a short clinic visit. The outcome, poor mental health, was intended to be a broad measure of the mental status of the subjects. We therefore included two separate measures of mental health in this variable. The first was the GHQ Depression subscale. Because 10% is approximately the lifetime prevalence of depression in the general population, the bottom 10% of patients, those who scored 11 or greater on the GHQ Depression subscale, were classified as having poor mental health. The second was, the subject's self-assessment of their own mental health. Using this second measure allowed us to identify those subjects who felt as if they had poor mental health but did not score as such on the GHQ Depression subscale. Thus the addition of this second variable accounted for any lack of validity in the GHQ Depression subscale's ability to detect poor mental health. We therefore felt that the combination variable that included both the GHQ Depression subscale score as well as the self-assessment was best reflective of overall mental health. Because the maternal behavior measures, being those included in the NCPP dataset, are more than 40 years old, they therefore were not the newest and best ways to accurately measure maternal behavior. However, because their shortcomings likely led to an underestimation of the number of mothers who exhibited poor behavior, they were unlikely to have any effect on the results. Additionally, although lifetime mental conditions as diagnosed by a
288
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289
medical professional would serve as a more accurate reflection of adult mental health, because such data were not collected by the NCPP, this was impossible, and therefore the poor mental health variable was used as the outcome measure. Finally, as in any long-term study, attrition played a significant role. In this study, it is probably equally as likely or even more likely that subjects with poor mental health left the study as it is that subjects who experienced poor parenting left the study. Thus attrition most likely led to an underestimation of the effect of maternal behavior on adult mental health. Therefore, the effect size, as determined by our study, is likely conservative.
5. Implications Early parental sensitivity to the infant's needs is of particular importance in setting the stage for the future physiological and psychological development of the child [3]. The quality of the attachment between mother and child is important for the overall social and emotional development of the child [11,22], including expectations of adult availability, problem solving, social relationships, and ability to cope with new or stressful situations [23]. The mother's own experience in childhood, her marital relationship, the quality of her social support network, and her own mental health, all affect her ability to appropriately care for her infant [7,9,24,25]. Pediatricians, family physicians, and other providers of health care to infants and children ought to be aware of these possible mechanisms that link a lack of adequate parenting and adult mental illness. Clinically, physicians and other healthcare staff must be attentive to mothers whose behavior may not be optimally supportive of their children's development due to lack of good parenting information and/or role models, being overwhelmed by depression and/or personal or family problems, lack of social support, or lack of basic resources. Referral for appropriate help may be indicated. On a preventive level healthcare workers should provide anticipatory guidance, counseling, and education to parents, which sensitive primary care providers should already do for children of all ages. Poor and unmarried mothers are more likely to respond inappropriately to their children, and so these are the mothers who should be targeted by prevention initiatives. Community programs are an excellent potential source of aid. Certain countries, such as Australia and New Zealand [26], offer far more in the way of health supervision and parenting education for mothers and children than other countries, such as the United States. In conclusion, maternal behavior toward the infant has a significant long-term impact on the child's risk of developing poor mental health in adulthood. Further this impact is limited to maternal behavior that does not promote secure bonding with the infant. That is to say, children of overprotective mothers do not see the same psychological sequelae that children of unsupportive mothers do. We suggest that primary care physicians involved in the care of young
children and their mothers be sensitized to this association so that they may be properly prepared to intervene.
Appendix 1. Factor analysis Step 1 Original variables. 4-month maternal behaviors X Mother's responsiveness to the child's physical needs Y Mother's focus of attention during the examination Z Mother's attitude toward the child's test performance 8-month maternal behaviors A Mother's expression of affection B Mother's verbal evaluation of the child C Mother's physical handling of the child D Mother's management of the child during the actual testing E Mother's reaction to the child's needs F Mother's reaction to the child's test performance G Mother's focus of attention during the examination Step 2 Recode variable (Condition 1: 1 = 2, 2 = 1, else = 0; variable label: “L.” Condition 2: 4 = 1, 5 = 2, else = 0; variable label: “H”). X X(L): Mother unaware of or slow to recognize child's physical needs X(H): Mother overly absorbed with child's physical needs Y Y(L): Mother centered her attention on the child Y(H): Mother centered her attention on her own concerns and/or problems extraneous to the situation Z Z(L): Mother indifferent to child's test performance Z(H): Mother defensive or critical of failures in child's test performance A A(L): Mother was negative and harsh with the child, rarely or never using an affectionate term A(H): Mother was overly affectionate both verbally and physically B B(L): Mother generally negative and critical of child B(H): Mother only positive about child and ignorant of less desirable behaviors C C(L): Mother inconsiderate and/or awkward in handling the child C(H): Mother extremely gentle and/or overly cautious in handling the child D D(L): Mother made no spontaneous effort to facilitate the testing D(H): Mother frequently disrupted the testing to help the child with given tasks E E(L): Mother unaware of or slow to recognize child's needs E(H): Mother immediately responsive to child despite no existence of needs F F(L): Mother indifferent to child's test performance F(H): Mother defensive or critical of failures in child's test performance G G(L): Mother centered her attention on the child G(H): Mother centered her attention on her own concerns and/or problems extraneous to the situation Step 3 Factor analysis. Variable
Factors 1
2
3
A(L)
0.728
0.023
F(L)
0.718
E(L)
0.683
− 0.018 − 0.089 0.050
0.113 0.103
Cumulative Cronbach's Name explained α variance 13.844%
0.741
Low attachment
A.P. Fan et al. / Comprehensive Psychiatry 55 (2014) 283–289 Step 3 (continued) Variable
Factors 1
2
3
D(L)
0.614
0.142
B(L)
0.539
− 0.010 0.067
C(L)
0.510
0.060
G(H) E(H)
0.166 0.657
F(H)
0.467 − 0.019 0.106
B(H) C(H)
0.005 0.010
0.592 0.571
A(H)
0.557
D(H)
− 0.060 0.086
0.532
G(L)
0.104
0.493
Y(H)
0.081 0.128
0.644
X(L)
− 0.013 − 0.123 0.169
− 0.095 − 0.010 0.803
0.563
Z(L)
0.180
X(H)
− 0.043 0.025
− 0.115 − 0.124 0.111
Z(H)
0.616
− 0.137 − 0.090 0.131 − 0.065 − 0.025 0.113 − 0.035 0.122
Cumulative Cronbach's Name explained α variance
[8] [9]
[10] 26.031%
0.634
Overly involved [11]
[12]
[13]
[14] 35.755%
0.307
Unhealthy behavior at 4 months
[15]
[16] 0.498 0.425
− 0.090 0.065 Model fit KMO = 0.686 Barlett test of sphericity: χ 2 = 5040.811⁎
Y(L)
[7]
Extraction method: factor analysis. Rotation method: Varimax with Kaiser normalization. ⁎ p b 0.05 was considered statistically significant.
[17]
[18]
[19] [20]
References [1] Erikson E. Childhood and society. 2nd ed. New York, NY: Norton; 1963 [Revised and Expanded. Edition]. [2] Main M, Kaplan N, Cassidy J. Security in infancy, childhood and adulthood. A move to the level of representation. Growing points of attachment theory and research. Monogr Soc Res Child Dev 1985;50:66-104. [3] Miller WB, Feldman SS, Pasta DJ. The effect of the nurturant bonding system on child security of attachment and dependency. Soc Biol 2002;49(3–4):125-59. [4] Hipwell A, Keenan K, Kasza K, Loeber R, Stouthamer-Loeber M, Bean T. Reciprocal influences between girls' conduct problems and depression, and parental punishment and warmth: a six year prospective analysis. J Abnorm Child Psychol 2008;36:663-77. [5] Weich S, Patterson J, Shaw R, Stewart-Brown S. Family relationships in childhood and common psychiatric disorders in later life: systematic review of prospective studies. Br J Psychiatry 2009;194:392-8, http:// dx.doi.org/10.1192/bjp.bp.107.042515. [6] Mellins CA, Brackis-Cott E, Dolezal C, Leu CS, Valentin C, MeyerBahlburg HF. Mental health of early adolescents from high-risk
[21]
[22] [23]
[24] [25]
[26]
289
neighborhoods: the role of maternal HIV and other contextual, selfregulation, and family factors. J Pediatr Psychol 2008;33(10):1065-75. Riley AW, Coiro MJ, Broitman M, Colantuoni E, Hurley KM, Bandeen-Roche K, et al. Mental health of children of low-income depressed mothers: influences of parenting, family environment, and raters. Psychiatr Serv 2009;60(3):329-36. Burke L. The impact of maternal depression on familial relationships. Int Rev Psychiatry 2003;15(3):243-55. Crnic KA, Greenberg MT, Robinson NM, Ragozin AS. Maternal stress and social support: effects on the mother–infant relationship from birth to eighteen months. Am J Orthopsychiatry 1984;54:224-35. Nagata M, Nagai Y, Sobajima H, Ando T, Honjo S. Depression in the mother and maternal attachment—results from a follow-up study at 1 year postpartum. Psychopathology 2003;36(3):142-51. Bosquet M, Egeland B. Associations among maternal depressive symptomatology, state of mind and parent and child behaviors: implications for attachment-based interventions. Attach Hum Dev 2001;3(2):173-99. Dawson G, Klinger LG, Panagiotides H, Hill D, Spieker S. Frontal lobe activity and affective behavior of infants of mothers with depressive symptoms. Child Dev 1992;63:725-37. Gómez FM. Early experiences and environment: their impact on neurodevelopment and the creation of mental processes. Vertex 2004;15:105-14. Niswander KR, Gordon M. The women and their pregnancies: the Collaborative Perinatal Study of The National Institute of Neurological Disease and Stroke. Edition. Philadelphia, PA: WB Saunders Company; 1972. Hardy JB, Drage JS, Jackson EC. The first year of life. The Collaborative Perinatal Project of the National Institute of Neurological and Communicative Disorders and Stroke. Edition. Baltimore, MD: The Johns Hopkins University Press; 1979. Broman SH, Nichols PL, Kennedy WA. Preschool IQ, prenatal and early developmental correlates. Edition. Lawrence Erlbaum Associates Publishers; 1975. Hardy JB, Shapiro S, Mellits D, Skinner EA, Astone NM, Ensminger M, et al. Self-sufficiency at ages 27–33 years: factors present between birth and 18 years that predict educational attainment among children born to inner-city families. Pediatrics 1997;99:80-7. Hardy JB, Astone NM, Brooks-Gunn J, Shapiro S, Miller TL. Like mothers, like child: intergenerational patterns of age at first birth and associations with childhood and adolescent characteristics and adult outcomes in the second generation. Dev Psychol 1998;34:1220-32. Goldberg D. Manual of the General Health Questionnaire. Edition. NFER Publishing Company; 1978. Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. Psychol Med 1979;9:139-4522. Hardy JB, Shapiro S. Pathway to adulthood: a three generation urban study, 1960–1994 (Baltimore, Maryland). The second generation children user's manual. The Johns Hopkins Medical Institution; 1997. Bowlby J. A secure base: clinical application of attachment theory. Edition. London, UK: Routledge; 1988. Drotar D, Sturm L. Personality development, problem solving, and behavior problems among preschool children with early histories of nonorganic failure-to-thrive: a controlled study. J Dev Behav Pediatr 1992;13:266-73. Bell SM, Ainsworth MD. Infant crying and maternal responsiveness. Child Dev 1972;43:1171-90. Main M, Hesse E. Adult lack of resolution of attachment-related trauma related to infant disorganized/disoriented behavior in the Ainsworth strange situation: linking parental states of mind to infant behavior in a stressful situation. Attachment in the preschool years: Theory, research and intervention. Chicago, IL: University of Chicago Press; 1990. p. 339-426. Rowe HJ, Fisher JR. The contribution of Australian residential early parenting centres to comprehensive mental health care for mothers of infants: evidence from a prospective study. Int J Ment Health Syst 2010;4:6.
