Journal of Behavioral Medicine

, Volume 36, Issue 6, pp 632–640

Beating and insulting children as a risk for adult cancer, cardiac disease and asthma

Authors

    • Plymouth University
  • Ahmed M. Alkhalaf
    • Plymouth University
  • Ben Whalley
    • Plymouth University
Article

DOI: 10.1007/s10865-012-9457-6

Cite this article as:
Hyland, M.E., Alkhalaf, A.M. & Whalley, B. J Behav Med (2013) 36: 632. doi:10.1007/s10865-012-9457-6

Abstract

The use of physical punishment for children is associated with poor psychological and behavioral outcomes, but the causal pathway is controversial, and the effects on later physical health unknown. We conducted a cross-sectional survey of asthma, cancer, and cardiac patients (150 in each category, 75 male) recruited from outpatient clinics and 250 healthy controls (125 male). All participants were 40–60 years old and citizens of Saudi Arabia, where the use of beating and insults is an acceptable parenting style. Demographic data and recalled frequency of beatings and insults as a child were assessed on an 8-point scale. Beating and insults were highly correlated (ρ = 0.846). Propensity score matching was used to control for demographic differences between the disease and healthy groups. After controlling for differences, more frequent beating (once or more per month) and insults were associated with a significantly increased risk for cancer (RR = 1.7), cardiac disease (RR = 1.3) and asthma (RR = 1.6), with evidence of increased risk for cancer and asthma with beating frequency of once every 6 months or more. Our results show that a threatening parenting style of beating and insults is associated with increased risk for somatic disease, possibly because this form of parenting induces stress. Our findings are consistent with previous research showing that child abuse and other early life stressors adversely affect adult somatic health, but provide evidence that the pathogenic effects occur also with chronic minor stress. A stress-inducing parenting style, even when normative, has long term adverse health consequences.

Keywords

Physical punishmentCorporal punishmentChildrenCancerCardiac diseaseAsthmaHealth

Introduction

Is physical punishment a good way of disciplining children? In 24 countries, physical punishment is illegal, either at school or in the child’s home (Zolotor & Puzia, 2010). In 96 countries it is illegal in school but ‘reasonable’ physical punishment is permitted by parents. In yet others, including the USA, some countries in the Middle East and Asia, physical punishment is legal in school and in the home (Zolotor & Puzia, 2010). In the USA, surveys of parents and caregivers show that the use of physical punishment decreases with age of the child but is used in about 80 % of preschool age children (Zolotor et al., 2011), and 90 % of American parents report using spanking at some time during their parenting history (Lansford et al., 2004). Although there are striking differences between countries (Cappa & Kahn, 2011), one international study suggests that about half of all children have experienced this form of discipline (Lansford et al., 2010). The use of physical punishment is highly contentious with arguments on both sides (Benjet & Kazdin, 2003; Holden, 2002). Some, including many pediatricians, reject the use of physical punishment (Oates, 2011), but rejection is not universal (Larzelere et al., 2010; McCormick, 1992).

A large body of research has examined the effect of physical punishment and other parenting styles on children and their later development. In general research shows a correlation between childhood physical punishment and both later behavioral problems and poor psychological functioning (Gershoff, 2002). Although the correlation between physical punishment and later problems is well established, the causal relationship is controversial, and it is the question of causality that drives the debate on policy.

There are three reasons for questioning whether the statistical association between physical punishment and poor outcome is due to a causal relationship. First, some research shows a statistical association in Western or European American samples, but not in many non-Western, Latin American and African American samples (Dodge et al., 2005; Lansford et al., 2004; Slade & Wissow, 2004). Additionally, authoritarian parenting style (which correlates with physical punishment, e.g., Jocson et al., 2012) is associated with poor mental health in Western but not Arabic and Asian countries, where authoritarian parenting is culturally accepted (Dwairy et al., 2006). These findings lead to the conclusion that it is not physical punishment per se, but some interaction between physical punishment and cultural factors, that is important. Some conclude that spanking is not harmful if consistent with cultural norms (Dodge et al., 2005). In contradiction to this conclusion, other studies show that higher levels of physical punishment are associated with worse outcome even when normative (Gershoff et al., 2010; Grogan-Kaylor, 2005).

A second reason for questioning the causal relationship between physical punishment and outcome is that the research is, for ethical reasons, correlational. Physical punishment will be correlated with other variables and these other variables may have the causal effect. The association between physical punishment and poor outcome remains even after extensive controlling for other potential confounding variables. For example, Taylor et al. (2010) controlled for mother’s age, education, income, race, religion, depression, use of drugs and alcohol, mother-father relationship, and intimate-partner violence but physical punishment remained a predictor of aggressive behavior if the physical punishment occurred more than two times in the previous month. Nevertheless, it is never possible to control for all possible confounds and critics can argue that some variable (e.g., some aspect of non-normativeness) has not been accounted for. Some argue that it is ‘inept parenting’ (Baumrind et al., 2002) rather than physical punishment that is the causal factor. The inept parenting argument, however, needs to explain exactly what inept parenting consists of and how it relates to physical punishment. Inept parenting has a number of components, which may include the type of punishment, but also a failure to provide a nurturing environment.

A final reason is that the causal relationship may go in the opposite direction. It is possible that naughty children elicit more spanking, and later behavioral problems are due not to the spanking but to their innate naughtiness. Parental disciplining behavior is logically related to the behavior of the child, and the finding that both non-physical as well as physical punishment are associated with adverse behavioral outcomes (Larzelere et al., 2010) is consistent with the hypothesis that bad behavior is the not the consequence of physical punishment but its cause. By contrast, other research shows that although fussiness in the first year of life predicts increased physical and non-physical punishment, the child’s aggression at an early age predicted neither later physical nor verbal punishment (Berlin et al., 2009).

To date, research on physical punishment in young children has focused on psychological and behavioral outcomes in older children. One study shows that physical punishment increases the risk of adult psychiatric disease (Afifi et al., 2006). To our knowledge no studies have examined the effect of physical punishment on adult physical health. Child abuse (i.e., unreasonable and non-normative physical punishment) and household dysfunction have both been shown to be related to increased incidence of adult somatic disease, including cancer, heart disease, lung disease, liver disease, skeletal fractures and irritable bowel syndrome (Beesley et al., 2010; Felitti et al., 1998; Fuller-Thomson & Brennenstuhl, 2009; Fuller-Thomson et al., 2010). Two possible mechanisms can explain these findings. One is that childhood stress increases the incidence of health harming behaviors. The other is that childhood stress causes causes epigenetic and other changes that predispose individuals to disease (Fish et al., 2003; Harper, 2005) through a raised non-specific inflammatory profile. Furthermore, it is likely that these two mechanisms interact in a vicious cycle: severe and chronic childhood stress leads to biological changes; biological changes predispose towards dysregulatory behaviors; and dysregulatory behaviors exacerbate the allostatic load on the biological system, leading to yet further biological changes (Miller et al., 2011). Some of the research on the adverse effects of childhood stress examines the biological consequences of severe chronic stress, such as abuse or trauma (Miller et al., 2011). Other research examines the consequences of low socio-economic status (SES) (Miller et al., 2009, 2011). The stress of low SES is multi-component and may involve either major stress or minor stress. Whether minor childhood stress (in the form of normative physical punishment) is sufficiently stressful to cause later disease has not been established.

Our study was motivated by the hypothesis that physical punishment is a signal of threat and, because it is a signal of threat, creates stress in the young child. Physical punishment is likely to be combined with other aspects of parenting that also signal threat create and stress. Physical punishment is therefore a strong indicator of a stress-inducing parenting style; however, physical punishment is not the only form of parenting that could lead to stress. We examined the relationship between parenting style and later physical disease in a society where harsh and stress-inducing forms of punishment are normative. By carrying out our research in this context, we avoid the criticism that it is not stress but the atypicality of the punishment that is the causal factor. By focusing on physical health rather than behavioral outcomes, we avoid the criticism that harsh punishment is being elicited by child characteristics.

Methods

Design

This is a cross-sectional questionnaire study where adults with and without diagnosed disease recall the frequency with which they were beaten as a child and provide demographic data.

Participants and cultural context

All participants were recruited in Saudi Arabia. Saudi Arabia is a country with a traditional but changing culture, but where many people consider the use of physical punishment and insults an acceptable and responsible way of raising children (Dwairy et al., 2006; Long, 2005). Harsh punishment or abuse is well documented, but until recently little action has been taken against perpetrators (Al-Mahroos, 2007; Eissa & Almuneef, 2010). Physical punishment in schools was banned in April 1996, and this ban has been ratified in subsequent legislation. Physical punishment remains legally permitted at home. In addition to physical punishment, a commonly used form of parenting is to insult the child, and where the insult is typically accompanied by yelling or shouting (Long, 2005). Thus, parenting style can involve threat through both physical and verbal punishment.

There were 4 groups of participants: out-patients with asthma, with cancer, or with cardiac disease, and a control group. Patients with cancer were recruited from the King Fahad Specialist Hospital in Dammam. Patients with cardiac disease were recruited from the Saud Al-Babtain Cardiac Center, in Dammam. Patients with asthma were recruited from two general hospitals: the Dammam central hospital and the King Fahad hospital in Hofuf. Health care is free to Saudi citizens. Our healthy control sample was recruited from administrators and nurses working at these hospitals. All participants were between 40 and 60 years old; they would have been children at a time the time physical punishment was permitted in schools.

Measures

Demographic questions recorded participants’ gender, age, education, and the education of their mother and father. Education was indicated via a 7 point scale (1 = none, 2 = elementary, 3 = intermediate, 4 = high school, 5 = Bachelor degree, 6 = Master degree, 7 = Higher graduate degree).

Frequency of beating as a child was measured by a single Arabic question which, translated into English reads “Were you beaten as a child?” Participants responded on an eight-point scale ranging from ‘never’ to ‘at least once a day’ (see Table 2). Frequency of insult was measured by a single question “Did your parents verbally insult you?” with same response scale.

Procedure

Patients were recruited, provided written informed consent and completed the questionnaires while attending outpatient clinics. A male researcher approached male patients and a female researcher female patients until the target number of 75 male and 75 female patients had completed the questionnaire for each type of disease. One hundred and twenty five male and 125 female control participants were recruited from nursing and administrative staff departments in the same way. Those in the control sample were asked if they had any disease, and if they did were excluded. Refusal to take part in the study was rare. Ethical approval was obtained from the relevant hospital or ministry authorities, and from the University of Plymouth Human Participants ethics committee.

Statistics

We computed non-parametric correlations between beating and insults. Differences in demographic characteristics of the groups (see Table 1) were compared by one-way ANOVA, before and after propensity score matching on demographic variables. Logistic regression was used to construct propensity scores (for an accessible introduction see Shadish & Steiner, 2010), predicting the probability of membership of each of the patient groups, as compared with the control group, based on participants’ age, gender, education and the education of their parents. Reports of beating and insult frequency were dichotomised by grouping the first four response categories on the scale and the last four response categories: punishment reported at every 6 months or less was coded infrequent (0) and punishment reported as monthly or more frequent was coded frequent (1). Risk ratios for the effect of frequent versus infrequent punishment were computed first without matching and then, as recommended by Baser, (2006), based on alternative propensity score matching procedures. These procedures were: 2 nearest neighbours with replacement; kernel density matching (bandwidth = 0.06), and radius matching (caliper = 0.05). The use of alternative procedures provides alternative estimates of risk ratio based on different assumptions used in the matching procedure. Significant variation in these results would be a cause for concern. All matching analyses were implemented using the Stata psmatch2 command (Leuven & Sianesi, 2003). To further explore the relationship between punishment frequency and outcomes we performed an unmatched logistic regression analysis in which patient/control group membership was regressed on indicators for different punishment frequencies.
Table 1

Demographic characteristics of our subsamples

 

Healthy

 

Cancer

 

Cardiac

 

Asthma

 

M

(SD)

M

(SD)

M

(SD)

M

(SD)

Age

47.37

(5.61)

50.47

(5.91)

52.97

(5.33)

49.55

(6.06)

Participants’ education

4.48

(0.84)

3.43

(1.09)

2.61

(1.12)

3.36

(1.38)

Father’s education

2.55

(1.42)

2.18

(1.19)

1.49

(0.78)

2.31

(1.30)

Mother’s education

2.15

(1.18)

1.91

(1.08)

1.36

(0.63)

2.03

(1.18)

Results

Demographic characteristics are shown in Table 1, and frequency of punishment in Table 2. The correlation between being beaten and being insulted was high (ρ = 0.846, p < 0.001): Of those participants who reported monthly or more frequent beating, 97 % also reported monthly insults.
Table 2

Frequencies of participants in each group who experienced different levels of beating and insult

 

Asthma

Cancer

Cardiac

Healthy

Total

Were you beaten by your parents?

 Never

33

11

37

113

194

 Once in my live

12

13

15

73

113

 Once a year

12

26

14

30

82

 Once every 6 months

28

35

11

11

85

 Once a month

30

25

12

8

75

 Once a week

16

29

14

3

62

 Once every 2 or 3 days

12

10

31

5

58

 At least once every day

7

1

16

7

31

 Total

150

150

150

250

700

Did your parents verbally insult you?

 Never

23

8

34

97

162

 Once in my live

13

9

11

45

78

 Once a year

12

5

9

42

68

 Once every 6 months

7

27

12

20

66

 Once a month

34

37

7

11

89

 Once a week

34

42

33

11

120

 Once every 2 or 3 day

17

18

31

10

76

 At least once every day

10

4

13

14

41

We used propensity score matching to control for the fact that frequently beaten participants differed on demographic variables. Prior to propensity score matching, infrequently-punished participants compared to frequently punished participants were significantly younger, more educated, and had more educated parents (all p values < 0.001). However, after matching, infrequently beaten or insulted participants compared to those frequently insulted or beaten did not differ on any of the demographic variables used to construct propensity scores (all p values > 0.14). To check that our matching procedure was appropriate we inspected propensity score histograms for each group, and this inspection indicated reasonable common support in all cases; matched analyses were performed on observations with common support and the minimum N was 391 (from a total of 400 available for each model). Estimated reductions in standardised bias following matching were substantial and ranged between 55 and 98 %. Although propensity matching can reduce the effect of selection bias between the two groups, it cannot eliminate it entirely; total elimination of bias would require perfect measurement of confounding variables.

Table 3 presents the increases in risk for participants who reported frequent (as opposed to infrequent) beating and being insulted as children, using different matching algorithms. In unmatched analyses, the effects of both beating and insults were large, with risk ratios ranging from 2.6–3.5. Matching for demographic variables produced smaller but still significant differences between participants who experienced frequent versus infrequent or no punishment in the form of beating or insults: median risk ratios from the 3 matched analyses ranged from 1.3 to 2.07 (all p values < 0.015). Results based on the 3 matching techniques were relatively consistent.
Table 3

Risk ratios for participants frequently versus infrequently beaten and verbally abused, including both unmatched results, and those following propensity score matching, using three different forms of propensity matching procedure

Outcome

Punishment

Unmatched

Nearest neighbourb

Kernel

Radius (0.05)

Cancer

Beating

    
 

 Risk ratio

2.71

1.69

1.74

1.76

 

 Differencea

0.46

0.30

0.31

0.37

 

 SE

0.05

0.07

0.06

0.06

 

 p

0.00

0.00

0.00

0.00

 

 N

400

392

392

392

 

Insult

    
 

 Risk ratio

3.56

2.09

2.07

2.06

 

 Difference

0.49

0.35

0.35

0.35

 

 SE

0.04

0.07

0.06

0.06

 

 p

0.00

0.00

0.00

0.00

 

 N

400

392

392

392

Cardiac

Beating

    
 

 Risk ratio

3.00

1.37

1.28

1.30

 

 Difference

0.51

0.21

0.16

0.17

 

 SE

0.05

0.08

0.07

0.06

 

 p

0.00

0.02

0.01

0.01

 

 N

400

391

391

391

 

Insult

    
 

 Risk ratio

2.65

1.57

1.29

1.29

 

 Difference

0.40

0.23

0.41

0.14

 

 SE

0.05

0.06

0.06

0.06

 

 p

0.00

0.00

0.01

0.01

 

 N

400

391

391

391

Asthma

Beating

    
 

 Risk ratio

2.71

1.64

1.67

1.64

 

 Difference

0.47

0.29

0.30

0.29

 

 SE

0.05

0.08

0.07

0.07

 

 p

0.00

0.00

0.44

0.00

 

 N

400

399

399

399

 

Insult

    
 

 Risk ratio

3.17

1.88

1.99

1.98

 

 Difference

0.46

0.32

0.33

0.33

 

 SE

0.05

0.06

0.06

0.05

 

 p

0.00

0.00

0.00

0.00

 

 N

400

398

398

398

aDifference: difference in probability of patient group membership for frequently versus infrequently punished participants

bThree methods of propensity score matching were compared: Nearest neighbour, kernel matching (Epanechnikov), and radius matching with a calliper of 0.05

Follow-up analyses, in which we simply conditioned on, rather than matched by, propensity scores, explored the frequency of punishment required to increase risk for each condition. Table 4 presents risk ratios for patients who experienced differing frequencies of beating and insults, as compared with participants who reported no beating or insults. Risks associated with both beating and insults appeared to be greatest when recalled as monthly, weekly, or every 2–3 days, though there was some evidence of increased risk of cancer and asthma for beating with a frequency of as low as once per 6 months. Unexpectedly, for all disease groups, the relationship with recalled beating or insults appeared to be curvilinear: risk ratios were attenuated at the very highest frequencies of reported punishment (i.e., at least once per day).
Table 4

Risk ratios for patient-group membership at different frequencies of beating and insult, compared with participants who reported no punishment

Frequency of punishment

Cancer

Cardiac

Asthma

Risk ratio

p

Risk ratio

p

Risk ratio

p

Insult

 Never

  Once in my life

2.20

0.04

0.81

0.23

1.27

0.35

  Once a year

1.38

0.49

0.98

0.91

1.29

0.32

  Once every 6 months

4.62

0.00

1.07

0.67

1.34

0.36

  Once a month

5.52

0.00

1.42

0.09

3.06

0.00

  Once a week

6.19

0.00

1.80

0.00

2.97

0.00

  Once every 2 or 3 days

4.05

0.00

1.45

0.04

2.15

0.01

  At least once every day

1.54

0.46

1.04

0.80

1.22

0.57

Beating

 Never

  Once in my life

1.68

0.11

0.80

0.12

0.79

0.33

  Once a year

3.60

0.00

1.23

0.20

1.32

0.26

  Once every 6 months

5.17

0.00

1.45

0.08

2.51

0.00

  Once a month

5.32

0.00

1.29

0.23

2.71

0.00

  Once a week

6.42

0.00

1.58

0.11

2.96

0.00

  Once every 2 or 3 days

3.66

0.01

1.72

0.01

2.06

0.03

  At least once every day

0.79

0.80

1.26

0.30

1.19

0.69

Results based on logistic models in which patient group membership is regressed on indicators for punishment frequency and propensity scores

Discussion

We examined whether a threatening form of parenting style was associated with the incidence of major disease in a population where a threatening parenting style is normative. We measured retrospective recall of the frequency of being beaten and the frequency of being insulted. The very high correlation we found between these two single item questions leads to two conclusions. First, that it is not possible to determine whether the relationship with outcome is due to beating or insults. Second, because of the high correlation between these two forms of punishment, it seems plausible that threat forms a meaningful dimension for describing parenting in Saudi Arabia.

After controlling for demographic factors, we found that those who reported being beaten or insulted as child were more likely to have cancer, cardiac disease, and asthma, when compared to a group of healthy individuals. These results suggest that a threatening parenting style influences not only behavior (as shown by previous research) but also the adult of major adult disease.

We found evidence of an increased risk of cancer and asthma with a beating frequency of only once per 6 months, which suggests that even infrequent beating can have pathogenic effects. There was some evidence that the relationship between frequency of threatening punishment and disease is curvilinear, and risk ratios for disease status were slightly lower for both beating and verbal insults at the highest (at least once every day) as opposed to top quartile (monthly/weekly). There are several possible reasons for this finding: First, the number of patients in this high-exposure group was small, and the curvilinear appearance may be due to chance fluctuation. The may also be survivor or responder artefacts in our data, whereby those beaten most frequently are censored from our dataset by death or disability, or experience greater distortions in recall of events from childhood. However, it is also possible that perceptions of threat decrease when beatings or insults are so frequent that they come to be expected. The latter explanation would be consistent with previous research showing that the child’s perception of threat and attitude towards physical punishment is more important than its frequency for mental health outcomes (Mulvaney & Mebert, 2010). Finally, it is possible that very frequent and less beating differ on some other aspect of parenting style which is protective in the case of very frequent beating.

Our results are consistent with research showing that various forms of early life stress have long-term consequences for later somatic health. Early life stresses that are associated with later poor health include physiological stress, such as poor fetal nutrition (Barker et al., 2009), and psychological stress, such as child abuse or trauma (Beesley et al., 2010; Fuller-Thomson & Brennenstuhl, 2009; Fuller-Thomson et al., 2010; Heim et al., 2009), family dysfunction (Felitti et al., 1998), and low socio-economic status (SES) (Miller et al., 2009) Additionally, poor fetal nutrition is associated with attention deficit hyperactivity disorder (ADHD) (Biederman et al., 2002) and with increased hostility (Räikkönen et al., 2008). Thus, our results are consistent with a wealth of data showing that adverse early life circumstances, irrespective of whether they are biological or psychological, all lead to poor psychological and physical health, either through a behavioral route, a biological route, or, as is more likely, some interaction between the two (Hyland, 2011; Miller et al., 2011).

Limitations of the study include the possibility of non-accurate recall of childhood experiences, the non-randomness of the selection of the different groups, and the possible omission of confounding variables from the models used to construct our propensity scores. The risk of bias in analysis was minimized by our use of propensity score matching. However, where matching variables are imperfect measures of the non-random selection of participants then bias may persist (Shadish & Steiner, 2010). Future studies might include additional matching variables, including other measures of parenting style, but this limitation is mitigated by our hypothesis: We suggest that it is not beating or insults that is important but rather the threat these methods of punishment evoke. Other aspects of parenting style may also be associated with increased poor health, but are likely to part of a general package where the parent acts in threatening way, thereby inducing stress in the child. Stress has well known immunological effects that predispose to disease (Kemeny & Schedlowski, 2007).

Further evidence in support of a causal inference from our data, comes from the different pattern between physical punishment and education in Arabic countries compared to Western countries. In the West, education is positively related to health and negatively with beating frequency. By contrast in our study education was unrelated to beating in the healthy sample, but there was a negative association in the disease samples. Wealth is positively related to beating in the Syrian Arab Republic (Cappa & Kahn, 2011). Thus, the relationship between wealth, education, and beatings is probably complex and varies with cultural factors.

Conclusions, implications and further research

Parenting styles in Saudi Arabia vary along a dimension where at one end physical punishment and insults are frequently used and at the other end they are never used. We found that this dimension of parenting was related to adult incidence of cancer, cardiac disease and asthma. Our results lend support to those who suggest that physical punishment should not be used. Our interpretation, however, is that it is not physical punishment per se but the threat produced by the use of physical punishment that leads to negative outcomes. It is difficult to see how physical punishment can fail to be threatening, though one possible interpretation of our data is that if the child is beaten so frequently (at least once a day), then the beating no longer becomes a threat, or at least the child adapts to it. We think this interpretation unlikely, as there are other explanations for the curvilinear relationship suggested in Table 4. We do not advocate daily beating!

Although our research was carried out in Saudi Arabia, it will almost certainly generalise to other cultures. Our research was carried out in a society where beating and insulting the child is considered an acceptable form of parenting. The adverse effects of this form of parenting are likely to be increased in societies where this form of punishment is considered abnormal.

Our data demonstrate extended consequences of childhood punishment late into adult life, thereby increasing the urgency of understanding this common parenting practice. However, replication is needed with a better control sample and in other countries. It is a primary concern to increase confidence that childhood punishment and adult illness are not confounded, for example, by using natural experiments or related counter-factual methods.

Several types of further research may be helpful. Better understanding of children’s perceptions of punishment may show when punishment is or is not stressful. If physical punishment is a component of early childhood stress, it is unclear how this interacts with other aspects of parenting such inconsistency, neglect, or the lack of a nurturing relationship. Additionally, the balance between reward and punishment in outcome may be more important than the type of punishment—there is evidence that reward is a more effective way of shaping behavior than punishment (Grusec, 2011). Further research is needed to examine the inter-relationship between these different factors, and how these different factors interact to produce a context that varies in stress. This research will help explain a phenomenon which is often overlooked—some children are raised in what are considered appalling conditions but nevertheless thrive and grow up to be healthy and long lived adults.

A final research question concerns to what extent the biological changes caused by childhood stress are reversible. The question of reversibility is seldom addressed. One theoretical approach suggests that stress is a one-way ticket involving permanent epigenetic change—or at least, the possibility of reversibility is not considered (Miller et al. 2011). Yet another suggests that some biological changes are potentially reversible if the adult is exposed to a highly benign environment (Hyland, 2011). The question of reversibility requires empirical investigation: The answer is important because of its implication for therapy.

Copyright information

© Springer Science+Business Media New York 2012