CHAPTER 3

MEDICAL, BEHAVIOR, AND SOCIAL PROBLEMS

Lianne Fisher, Susan Thompson, Mark Ferrari, Lynn A. Savoie, and Sarah Lukie

When parents were asked whether they had initially had any concerns about adopting an RO child, 94% of them mentioned potential health problems, 31% mentioned concerns about developmental delays, and only 18% mentioned having concerns as to potential behavioral, emotional, or social problems the child might have. Developmental delays have already been discussed in the previous chapter. Here, we will discuss the children's other problems in the order of importance the parents would have given them before adoption, that is, first medical, then behavioral-emotional and social. It should be noted, however, that this is not the order of importance that parents would now give to these areas.

Physical growth and medical problems

Children's weight and height

Besides specific medical conditions, children arriving from orphanages were very small and malnourished. According to their parents' reports, at the time the RO children were adopted, 85% of them fell below the 10th percentile for weight (i.e., fewer than 10% of children in the North American population were smaller than they were), with 59% falling below the 5th percentile (McMullan & Fisher, 1992). Using Romanian norms, the same results were found. These results are supported by the findings of Benoit et al. (in press), who found that 3 months after adoption 50% of a sample of 16 Romanian children adopted to Canada after 6 months of age were below the 5th percentile in weight, and 44% were below the 5th percentile for height. Rutter (1996) has reported that half of the Romanian children adopted to the U. K. had weights below the third percentile.

Parts of this chapter have been accepted for publication (Fisher, Ames, Chisholm, & Savoie, in press).

Medical problems

At Time 1, 85% of RO children were reported to have or to have had at least one medical problem, a higher proportion than was reported for CB** (39%) and EA** children (42%). Ninety-three percent of RO children's medical problems were present from the time they were adopted, as were 65% of EA children's medical problems.

The most common problems of the RO children w ere intestinal parasites (including giardia) (31%), hepatitis B (28%), and anemia (1 5%), although there was a wide range of problems. Giardia and intestinal parasites were found only in RO children, whereas hepatitis B and anemia were found in both the RO and EA groups. Other types of problems found only in the RO and EA groups included skin problems (e.g., scabies), problems of diarrhea or vomiting (which may have been symptoms of their intestinal problems), and bone disease (e.g., rickets). In contrast, eye problems, ear problems, and respiratory problems were reported across all, three groups.

Occasionally, parents would request particular medical tests for their child and these requests would be refused. Usually, however, parents had no idea what tests to request, and had to depend on their physician's judgment as to whether tests should be done and what those tests should be.

Although at Time 2 the percent of RO children reported as having at least one medical problem (72%) had decreased* from the 85% reported at Time 1, still, a larger proportion of RO children (48%) than of CB children (1 5%) had 2 or more medical problems reported for them **. The EA group was intermediate, and did not differ from either the RO or CB groups.

To examine whether the average severity of medical problems differed across groups, the 51 different medical problems mentioned by parents in the interview were rated by a class of 1 8 nursing students from the University of British Columbia. Ratings were on a 7-point scale from I = "no problem at all for parents" to 7 = "severe problem for parents". The average problem severity ratings of the RO, CB, and EA groups were all between 2 and 3, and the groups did not differ significantly.

Summary: Physical and medical problems

Most RO and EA children had greatly improved their height and weight status and their general health since coming to Canada. Still, RO children remained smaller than CB children in both height and weight; they were also shorter than EA children. The longer RO children had spent in orphanage, the shorter they were for their age. At Time 2 the RO children still had more medical problems than did CB children, but the problems were not of greater severity than those of CB or EA children. The intestinal parasites and anemia characteristic of RO children had been overcome, and their health was greatly improved. Because there is no cure for hepatitis B, it remained as a problem. At neither Time 1 nor Time 2 did EA children differ from CB children in terms of medical problems.

Behavior problems at Time 1

Number of behavior problems at Time 1

The number of behavior problems was measured with the Child Behavior Checklist (CBCL) for 2- to 3-year-olds (Achenbach, 1 992; Achenbach, Edelbrock, & Howell, 1 987). This is a 100-item test on which parents indicate whether an item is 0 ("not true of my child"), 1 ("somewhat or sometimes true of my child"), or 2 ("very true or very often true of my child"). Besides a Total Problem score, problems are divided into Internalizing and Externalizing dimensions. The Internalizing dimension taps withdrawn, anxious types of behavior, for example, "child is withdrawn" or "child looks unhappy". The Externalizing dimension is characterized by actingout behaviors. Examples of items are "child has a hot temper" and "child destroys others' things".

Types of behavior problems, Time I

Although we used the CBCL to measure the number of problems children had, we did not depend on it for a description of what those problems were, because it contained only problems common to North American children and might therefore not be able to tap particular problems arising from orphanage experience. To describe those particular problems, we asked RO children's parents to describe any difficulties they had experienced with their children since their arrival in Canada. Parents of children in the CB and EA groups were asked to describe any problems they had experienced with their children for the time period that corresponded to the time spent in Canada by that child's RO match. Following this open-ended question, parents in all groups were asked specifically if any difficulties had been experienced. in each of the following problem areas: eating problems, sleeping problems, stereotyped behavior problems, sibling problems, and peer problems. For each problem mentioned, parents were asked when this problem started or occurred, what they had done about it, how the problem had changed, and whether they had sought help or advice on how to deal with it. All interviews were audiotaped.

Only the two most troublesome problems in each of the problem areas were coded. All answers to interview questions were independently coded by two coders, and discrepancies between coders were resolved by discussion with a coding supervisor. It is important to note that a problem was coded only if the parent perceived it as a problem; if the parent mentioned a behavior but did not consider it to be a problem, it was not coded.

Eating problems, Time 1

RO children were given all food by bottle in a sort of pureed mush for the first 1-1/2 to 2 years. Refusal to chew is also a difficulty that has been reported in North American children (Wasserman, 1 987), one that occurs when solid foods are first introduced to the child. Its later appearance in RO children is likely, then, to represent merely the appearance of a typical childhood problem delayed because of lack of environmental opportunity.

The problem of eating excessively has also been reported in prior research. on children with orphanage experience (Flint, 1978; Lowrey, 1940; Provence & Lipton, 1 962). Several RO families reported that they continued to offer their child food following a

meal, and that additional food was never refused. It was the parents, not the children, who finally said "enough". Given the orphanage children's malnourishment and small size, this eating "too much" probably represented their bodies' natural "catch-up" mechanism when food became available. In addition, the RO children had to learn when to stop eating; never before had they eaten enough that they could learn what it felt like to be full.

Sleeping Problems, Time I

RO children had significantly more sleeping problems than did EA children **, but not more than CB children. There was one distinctive problem of RO children. Twenty-eight percent of them were reported to lie quietly in their beds without signaling wakeup; this problem was not reported for any CB or EA child. Again, this seems to be a behavior directly attributable to orphanage life, where lying quietly in bed was the most common activity for young children, and where no one would have come to children if they had cried or called upon awaking. This problem disappeared when children learned that their parents would come if they called, or that they could get out of bed by themselves.

Stereotyped behavior Problems, Time I

Stereotyped behaviors have also been reported in normative samples of North American infants less than one year old (Thelen, 1981). In the weeks before infants first start to crawl, they typically rock back and forth on their hands and knees in a stereotyped fashion. Thelen (1 979) has suggested that this rocking is a built-in motor pattern preparation for crawling. She has also documented other stereotypies that happen when children are in sitting or standing positions, and has showed that these occur shortly after children have matured enough to assume such postures (Thelen, 1979). It is our belief that the same sorts of mechanisms may have been operating in the RO children for the stereotyped behaviors of rocking from side to side while standing, and rocking the trunk back and forth while sitting. Rocking may well be a preparatory behavior for balance, walking, and forward movement. Children in Romanian orphanages showed these motor patterns, but then had nowhere to move beyond the length of their cribs. In the absence of any opportunity to develop further, these, ordinarily preparatory behaviors were continued, and persisted as habits.

In addition to being signs of environmentally-blocked development, these behaviors could have had at least two functions that helped them to persist within the orphanage setting and even beyond. First, they could be self-stimulating attempts to adapt to the bland and unresponsive orphanage environment. Second, they could have become self-soothing measures in times of distress. Self-rocking may have given children the soothing that adults would provide in a non-orphanage setting. Research on North American infants has also suggested that stereotypies may be self-soothing coping behaviors. Increases in stereotyped behaviors were seen when North American children were placed in restrictive conditions, e.g., a car seat (Thelen, 1981). Children classifed as generally high in stereotypy had also been talked to and held less (Thelen, 1981).

Although stereotyped behavior was the most frequently reported behavior problem of RO children, it was also reported as having improved in 55% of cases and as being completely resolved in 43%. Only 2% of children with stereotyped behavior had not improved at all in the 1 1 months after adoption. This improvement fits well with the functions we have proposed for stereotyped behaviors: as the adoptive home provided increased stimulation, and as the child developed ways of letting his or her needs be known to a responsive parent, these types of behaviors naturally decreased.

The degree of improvement in stereotyped behavior problems was negatively correlated with the amount of time RO children had spent in the orphanage (r = -.34**) and positively correlated with the length of time they had been in Canada (r = .31 **). That is, the longer the child had been in orphanage, the less improvement she or he showed by the Time 1 interview, and the longer the child had been in Canada, the greater was the improvement.

Sibling and peer problems, Time 1

Overall, 32% of RO children were reported to have a problem with peers. The specific problem for 1 2% of RO children was avoiding peer contact, and for 1 0% it was being overwhelmed by peer attention. These specific problems were not reported for either CB or EA children. On the CBCL item of "withdrawn, doesn't get involved with others", RO children scored higher than CB children **.

Thus, in both their relations with their siblings and their relations with their peers at Time 1, RO children seemed somewhat overwhelmed by the other children, and tended to withdraw or avoid them. "Withdrawal from peers" has also been reported in prior literature on children with orphanage experience (Flint, 1978; Goldfarb, 1943). While this behavior was consistent with RO children's higher Internalizing scores on the CBCL, avoidance of peers may seem surprising for children who in orphanage had been reared in groups, never out of sight of other children. It has been observed (Ames, 1990), however, that children in cribs seldom attempted to interact with the child in the crib next to theirs, because that other child would usually not give any response.

Older orphanage children (aged 1-1/2 to 3 years) were only slightly less passive. Because many of them were not able to walk or talk, their ability to socialize with other children was limited. When children moved into orphanages for 3- to 6-year-olds they were given language lessons, and they seemed more lively (Ames, 1990). Compared to North American children of the same age, however, their social skills were extremely underdeveloped, and they did not know how to play. After adoption, they were confronted with siblings and peers who were not like the roommates they had known. When faced with noisy, busy, and boisterous children who wanted to involve them in activities and play they did not understand, approximately 1/4 of the RO children simply withdrew. Withdrawal necessarily, meant that new social skills could not be learned, so it is not surprising that parents reported that between adoption and the Time 1 interview the lowest amount of improvement (60%) had been made in the area of sibling and peer problems.

How parents dealt with behavior problems

Parents of RO children sought professional advice concerning 33% of their children's behavior problems. Sixty-two percent of the advice was sought from physicians, but nursery school or day care teachers, physiotherapists, psychologists, and public health nurses were occasionally consulted. Parents were most likely to seek advice concerning eating (45%.of problems) and stereotyped behavior

(37%); very little advice was sought on sibling (1 1 %) And peer (22%) problems.

RO parents generally used gentle means (e.g. ignoring the problem, reassuring the child) rather than harsh means (e.g., scolding, punishing, or restraining the child) to deal with all of the behavior problems. In more than 3/4 of eating, sleeping, and stereotyped behavior problems, and in 100% of sibling and peer problems, RO parents employed gentle means. Because so few parents employed harsh means, it was not possible to determine whether harsh or gentle means worked better in reducing problems. In a general sense, however, the predominantly gentle means that were used appear to have been effective, given the high rates of improvement reported between adoption and the Time 1 interview (60% to 98% of cases in the various problem areas).

Discussion: Behavior problems at Time I

At Time 1 RO children had a larger total number of behavior problems and more Internalizing problems than did the other two groups, and several of their problems were distinctive of their group, that is, they were not present in CB or EA children. The fact that EA children were not different from CB children in any problem areas or in CBCL scores suggests that it was orphanage experience rather than prenatal or perinatal experience that was responsible for the problems. Indeed, it is possible to attribute the RO children's distinctive problems to specific aspects of their orphanage life. For example, the eating problems (unwillingness to eat solids, and not seeming to know when to stop eating) seem attributable to lack of opportunity to eat solids or to learn the feeling of having had enough to eat.' Not signalling being awake seems, a direct carryover from an environment in which no one cared whether a child was awake or asleep. Stereotyped behaviors, we propose, were probably motor movements that were not allowed to develop because of environmental restriction of movement, and that were maintained even after restriction was removed because of their stimulating and/or self-soothing properties. Avoidance of other children after adoption seems to have been wariness toward children who behaved unlike any children the child had seen in orphanage.

Because it is possible to trace these distinctive problems back to their roots in orphanage life, even though some of them (e.g., stereotyped behaviors) would be considered signs of "abnormality" in children reared in home environments, we do not see them necessarily as signs of underlying brain damage or emotionallyproduced pathology. It is very useful to use age norms as markers for the identification of "abnormal" behaviours in children reared in normal home environments. For children with unusual early experiences, however, the same problems may not indicate abnormality, but rather, adaptation to the rearing environment. Only when a particular problem does not resolve within a reasonable amount of time after children have been exposed to opportunities should behavior be regarded as "abnormal". It was very important, then, to see how much behavior problems had improved at Time 2.

Behavior groblems at Time 2

Number of behavior problems, Time 2

On the 4- to 1 8-year version of the CBCL there are also some scales that do not contribute to either the Internalizing or the Externalizing scores. The RO children scored higher than CB** children on the Thought Problems scale, and higher than both CB** and EA** children on the 'Attention Problems scale.

Were RO children really less "internalizing" at Time 2?

It appears that in spite of no longer scoring higher than the other groups on the Internalizing dimension at Time 2, RO children had not lost all the behaviours that gave them a high Internalizing score at Time 1.

The (Internalizing) behaviors on which RO children scored higher than RO and/or EA children on the 2- to 3-year-old version of the CBCL at Time 1 are listed in the box on the left of Figure 2. Most of the items measuring these behaviors also appear in the 4- to i 8year-old version of the CBCL, but on different scales, as shown in the boxes on the right of Figure 2. RO children's scores on all the items shown in the right-hand boxes did not change significantly from Time 1 to Time 2.

Figure 2

SHIFT OF CBCL ITEMS FROM YOUNGER VERSION TO OLDER VERSION

It may be seen from Table 7 that RO children still scored higher than CB and EA children on these items at Time 2.

The relation of number of behavior problems to time in orphanage and in Canada

Time in orphanage - As shown in the two left columns of Table 8, at both Time 1 and Time 2 correlations between the total time RO children spent in orphanage and their CBCL Total problem**, Internalizing and Externalizing** scores were all positive. The longer children had lived in orphanage, the more behavior problems of all kinds they had.

Time in Canada - At Time 1, the length of time RO children had been in Canada was negatively related to their Internalizing* scores (i.e., the longer they had been in Canada, the lower the Internalizing score), but unrelated to their Total and Externalizing scores (Table .8).

At Time 2, most of the RO children were tested when they were 4-1/2 years old. This meant that the time in Canada was not independent from time in orphanage: the longer the time spent in orphanage, the shorter the time the child had spent in Canada by age 4-1/2. In order to disentangle these two variables we sent CBCL questionnaires to RO families again at a later date. At this time the children's median age was 76.5 months (range 64 to 148 months). Thirty-four families completed and returned the CBCL. There was no significant difference between the RO children's 4-1/2-year-old scores and this second set of scores.

Using these later "Time 2" scores, length of time in Canada was negatively correlated w ith RO children's Total**,

Externalizing**, and Internalizing* scores, as shown in Table 8.

The longer the total time the children had been in Canada, the lower were their CBCL scores.

Types of behavior problems at Time 2

The number of stereotyped behavior problems in RO children decreased** from Time 1 to Time 2, but in spite of this improvement, at Time 2, three years after adoption, RO children still had more stereotyped behavior than CB** and EA** children. Forty-one percent of all RO children still displayed at least some stereotyped behaviour, although it had decreased in frequency in all children. Decreases over time in eating problems, difficulties with siblings, sleeping problems, and stereotyped behaviors have also been reported for Romanian adoptees by Marcovitch, Cesaroni, Roberts, and Swanson (1 995).

Social problems with Peers

Because at Time 2 the children we're all 4-1/2 years old or older, their interactions with peers had become more important in their lives. First, to find out whether children in all groups had different amounts of opportunity to interact with their peers, in the interview we asked parents a. series of questions concerning how frequently the child played with friends the same age, how important the parent thought it was for the child to engage in social activities with peers, how often the parents had friends to the home for social occasions, and how often the parents socialized away from home with friends. On none of these questions were there differences between RO and CB, RO and EA, or CB and EA groups, so it appears that all groups of parents had approximately equal attitudes towards their own and their children's socializing, and provided the children with equal opportunities to play with their peers, both in structured settings (preschool, day care) and informally.

To measure the social behavior of the children in greater detail, we used three paper-and-pencil questionnaires:

Social Skills Ratinci System (SSRS), Preschool version

(Gresham & Elliott, 1 990). This is a series of rating scales designed to measure children's social skills and problem behaviors. Two versions of the SSRS were used: the 49-item version filled out by parents emphasized social behaviors needed in the home or in smaller peer groups, and the 40-item version filled out by teachers emphasized social behaviors more essential in the classroom or larger group situation. Each item was scored as 0 = child "never", 1 sometimes" does, or 2 = "very often" does this behavior, e. g., ticontrols temper when arguing with other children".

The Preschool Behavior Ouestionnaire (PBO: Behar & Stringfield, 1 974). This questionnaire has 30 items (e.g., "fights with other children", "tells lies"), for each of which the parent indicated that it "does not apply" (score 0), "sometimes applies" (1), or "frequently applies" (2) to the child.

The Child Behavior Checklist (CBCL: Achenbach, 1991). This questionnaire, which measures all sorts of behavior problems (including social problems) has been described above. Initially, to measure social behavior we looked only at the 8-item Social Problems subscale, on which parents indicate whether each item is 0 ("not true of my child"), 1 ("somewhat or sometimes true of my child"), or 2 ("very true or very often true of my child").

Group differences on the three social behavior questionnaires

All three measures were intercorrelated **, and on each of them RO children scored as having poorer social behavior than CB ** children. For this reason, and because several of the items were the same on two or three of the questionnaires, we decided to construct our own measure of social behavior problems. This was done in two steps. First, nine developmental psychologists selected all the social items from the three questionnaires (including the entire CBCL, not just the Social Problems scale). "Social" was defined as behavior that involves an interaction between people.

Second, we performed a factor analysis (a statistical procedure that allowed us to see whether there were different groups of items each representing a different type of social problem). Only one clear factor was obtained, so the sum of a child's score on the social items that fell on this factor was taken as that child's Social Behavior Problem (SBP) score. Some sample items from the SBP are shown in Table 9.

Because each of the 58 items could receive 0, 1, or 2 points, total scores on the Parent SBP could range from 0 to 116.

A similar factor analysis was performed on the social items on the three questionnaires completed by teachers. As with the parents, only one clear factor was obtained. Its 57 items were very similar to those on the parent SBP score. Teachers' SBP scores were positively correlated with parents' SBP scores in each group: RO r = .60**; CB r = .37*; EA r = 70**. That is, there was a tendency for a teacher and a parent judging the same child to agree on whether the child had few or many social behavior problems (SBP).

Group differences on SBP scores

(r = .37 **), to having a lower IQ (r = -.44 **), to having younger adoptive parents (mothers r = -.36 **, fathers r = -.36*) and a lower family income (r = -.34 **), to the family having adopted more than one child from Romania (r .44 **), and to the parent feeling less attached to the child (r -.42**).

How parents dealt with behavior problems, Time 2

At Time 2 CB parents used less harsh means than they had at Time 1 **, whereas the RO group had not changed. Because the CB group had become less harsh and RO parents had not, at Time 2 the RO parents used harsh means to deal with a larger proportion of problems (20%) than CB parents (5%) did **. EA parents were in the middle, not significantly different from either the RO or the CB group.

Which type of problem has been most worrisome for parents?

As mentioned at the start of this chapter, before RO parents went to Romania, 94% of them were concerned about health problems their adopted child might have, 31 % were concerned about developmental delay, and only 18% said they were concerned about behavioral, emotional, or social problems the child might have. By the Time 2 interview, their actual concerns were very different. When parents were asked. "Of all the problems we have talked about either today or in our last interview [Time 1]. What is the one that stands out as being the most troublesome?" Only 10% mentioned developmental delay (academic problems, speech problems), and only 18% mentioned medical problems, with over 1/2 of the medical concerns being hepatitis B. The other 72% of "most troublesome problems" were behavioral, emotional, or social.

In summary, by Time 2, RO children had overcome most of the medical problems they had on arrival in their homes, and medical problems (except for hepatitis B) had become less of a concern for their parents. Developmental delays, which had greatly declined in severity, had also become less of a concern to parents (although it is possible that academic problems may cause greater concern in the future as more of the children enter school). What most parents had not anticipated before they adopted was that their RO child might have more behavioral, emotional, and social problems than other children. These problems have assumed much more importance with time. By Time 2 RO parents were dealing with the specific problems related to their children's orphanage experience (eating, sleeping, stereotypies, peer and sibling problems) in a harsher manner than CB parents were with similar (but much less frequent) problems in their children.

Given the concerns that parents expressed at the start, we believe that many of them had not realized that their RO children, more than other children, would require greater parental efforts at teaching them how to focus attention, how to interact with their peers, how to deal with fears and negative emotions, and generally, how to behave as well as their parents want them to behave. Adopting an orphanage child requires remediation in a I I areas-behavioral, emotional, and social, as well as medical and intellectual.

Service utilization

Avery important practical question is whether, with their larger numbers of medical, developmental, and behavioral-social problems, the RO -children were using more services for remediation of these problems. To measure service use, parents were asked during the interview how many times during the past 3 months the child had visited a hospital, an emergency room, the family doctor, any other doctor, a mental health professional, or a dentist. There were no differences among the groups in how often they had used these services.

In addition, parents were asked to name any kind of specialized teaching or education their child had ever used in Canada. The specific services used are shown in Table 10. More RO children** (67%) and EA children** (23%) used at least one specialized teaching service than did CB children (9%). When the Infant Development Programme was excluded, more RO children (41%) than CB children** (9%) and EA children** (10%) used at least one specialized teaching service, and CB and EA children did not differ on their use of specialized teaching services. This result is in accord with results obtained by Groze & lleana (in press) on a volunteer sample of parents of children adopted from Romania to the United States, 2/3 of whom had spent time in institutions. In response to a questionnaire, 24% of school-aged adoptees' parents reported that their children were in special education classes at least part of the time.

The amount of service utilization was correlated with the number of problems in all groups (RO L,= .61 **; CB Es = .38**; EA L, .40**), and was uncorrelated with caregiver stress, or with demographic characteristics or social isolation of the family, so it appears that service was being utilized equitably according to need.

When asked whether there were services they wanted to access but were unable to do so, 95% of CB and EA parents could not identify any such services. Fifty-two percent of RO parents also indicated that they had been able to find all the services they wanted.

Of those who had not been able to find services:

6 parents expressed the wish for an adoption support group, especially one for parents of older children.

4 parents wanted advice concerning behavior problems.

4 parents wanted speech therapy for their child, especially access to speech therapy when the child was young.

3 wanted respite care. ,

2 wanted a one-to-one aide for their child in preschool and

1 each wanted "doctors who believe a parent's concerns"

a Fetal Alcohol Syndrome clinic, and transportation to services.

In light of their children's problems, these desired services seem reasonable and warranted.

Back to A.R.I.A.'s Homepage