ADHD medications and academic achievement in elementary school

By Nestor Lopez-Duran PhD

A few weeks ago I discussed a research study that examined the effects of the medication Concerta (methylphenidate) on performance variability during cognitive tasks in children with ADHD. But does this translate to improvements in school work? Does the research support the effectiveness of ADHD meds in more tangible outcomes, such as grades or academic achievement?

Surprisingly, there is a lack of longitudinal long term research exploring the effectiveness of ADHD medication across multiple grades. Instead, most ADHD research examining academic outcomes are relatively short (within one year) or have very small sample sizes. However, in the latest issue of the Journal of the American Academy of Pediatrics, Dr. Richard Scheffler and a team from the University of California at Berkeley, reported the findings of a comprehensive long term examination of the effectiveness of ADHD medications on academic achievement.

The authors examined a representative population cohort of children entering Kindergarten in the late 1990s. The study cohort included 11,890 children who entered Kindergarten in 1998. These children were examined yearly until the end of their 5th grade. The authors gathered information on whether the child had an ADHD diagnosis, whether the child was taking medication (more of this below), and the mathematics and reading achievement levels of all the kids during the 5 year study.

The estimation of the medication use was a bit tricky. During 5th grade, the families were asked whether the child was taking medication for ADHD at that time. If the child was not taking medication, the authors assumed that the child had not taken ADHD medication during the duration of the study. If the child was taking a medication in 5th grade, the parents were then asked to report the length of medication use, which was used to estimate past years use.

The authors then compared children who had been medicated to those with a diagnosis of ADHD but who had not received any medications.

The Results:

  1. 9% of the sample had a life-time diagnosis of ADHD by 5th grade. This does not mean that 9% of 5th graders had ADHD. It means that by 5th grade, 9% of those who entered kindergarten in 1998 had received an ADHD diagnosis sometime during their lives.
  2. 68 % of kids with ADHD had taken medication for their condition.
  3. While controlling for a number of individual and family characteristics, medicated ADHD kids had significantly higher mathematics achievement scores across the different grades than the non-medicated ADHD kids. Although this difference is statistically significant, the authors reported that the gains represent the average gain expected during 0.19 school year over a 6 year period.
  4. There was no difference between those medicated at a single year vs. those medicated at multiple years in their mathematics achievement scores.
  5. Children who were medicated in multiple years had significantly higher reading achievement scores than the non-medicated ADHD peers. This reflects gains of 0.29 school years over 6 years.

Despite the limitations of this study regarding how medication use was estimated (retrospectively via parental report, no information on dosage, gaps in administration, etc), there is one very compelling overall finding: If we are to assume that severity of ADHD is associated with the likelihood of medication use (the more severe the more likely you are to be medicated), these findings show that medications are effective in improving academic achievement even among these severe kids. But we cant test that hypothesis because the study did not include data about the initial severity of ADHD prior to medication use. Someone could also argue that the effects observed were not due to the medication, but instead to other untapped family characteristic that differentiated those who tried medications vs. those who did not. That is, it is possible that factors that make a family more likely to try medication for their ADHD kids contribute to the kids better long-term academic performance.
The reference: Scheffler, R., Brown, T., Fulton, B., Hinshaw, S., Levine, P., & Stone, S. (2009). Positive Association Between Attention-Deficit/ Hyperactivity Disorder Medication Use and Academic Achievement During Elementary School PEDIATRICS, 123 (5), 1273-1279 DOI: 10.1542/peds.2008-1597

Effects of bullying on children with special needs

By Nestor Lopez-Duran PhD

During the past few months Ive discussed a couple of studies on bullying, including an examination of the relationship between bullying and psychotic symptoms, and a study exploring the factors that make kids more likely to be bullies or victims. Today I report on a related study that was presented at the APS convention last weekend. The authors of this study examined the long term effects of bullying on children receiving special education services.

A number of studies have shown that kids with special needs are at higher risk for educational and behavioral problems. Other studies have also shown that being bullied is associated with a number of negative long term outcomes. Yet, less is known about the long-term effects of bullying on children receiving special education services. For example, are these kids at an even higher risk for developing emotional disorders in response to bullying? Or on the contrary, are these kids more resilient to the effects of bullying than kids not in special education?

To answer this question a research team from the University of South Florida surveyed 1,439 middle school children attending schools in Florida. 155 of these kids (approximately 11%) had a special education classification (specific learning disorder, speech disorder, or emotionally handicapped). These kids were assessed in 2003 and then 4 years later in 2007. The authors examined:

  • Bullying behavior (history of being a bully or a victim)
  • Attendance
  • GPA
  • Discipline records (number of referrals and suspensions)

The results:

  1. Among the kids with a special education classification, 14% were victims of bullying, and 8% were bullies. Among kids with a regular education status, 12% were victims and 5% were bullies. There was no difference in the rates of victimization or bullying between kids in special education and those in regular education.
  2. At time 1, being a bully was associated with more discipline referrals, suspensions, and lower GPA. However, the association between bullying on discipline referrals appeared to be stronger in the non-special education kids.
  3. Also at time 1, being in special education was associated with higher discipline referrals and higher suspensions.
  4. When compared to their time 1 levels, victims of bullies in special education had better GPAs than victims in regular education.
  5. Also when compared to time 1 levels, bullies in special education showed a reduction in referrals while bullies in regular education showed an increase in referrals.

There are two very salient findings. First, there was no difference in the rate of bullying between students in special education and those in regular education. This is a bit surprising as past studies have shown that kids with special needs (for example learning disabilities) are at higher risk for being victimized (see for example Faye, 2003 for a review). Why the discrepancy? One possibility is that the rates of bullying may be changing. Specifically, most data on special education and bullying comes from studies conducted in the 90s. It is possible that the rates of bullying (or at least the self-report of bullying) in general education classrooms is increasing and now stands equal (about 10-15%) than among students in special education. Another possibility is that previous studies were conducted with very small sample sizes and that this study more accurately reflects the rates of bullying in general population.

A second surprising finding is that both bullies and victims in special education seem to perform better overtime than bullies and victims in regular education. Why? This effect may reflect the effectiveness of the special education program. That is, it is likely that kids in special education receive more targeted interventions that help them modulate the harmful effects of bullying.

The References:
Feldman, Gesten, Rojas, Totura, Smith-Schrandt, Alexander, Scanga & Brown (2009). A longitudinal evaluation of bullying and victimization among adolescents of varying exceptionalities. Poster presented at the Annual Convention of the Association for Psychological Science. San Francisco, May 2009.

Mishna, Faye (2003). Learning Disabilities and Bullying: Double Jeopardy Journal of Learning Disabilities, 36 (4)

Reading disabilities and speech delays: Literacy outcomes in children with Speech Sound Disorder

By Nicole Hess MS, CCC-SLP

Friday’s Column “Focus on Language” by Nicole Hess.

One of my jobs as a speech language pathologist is to evaluate children who have speech delays, and one of the most common questions that I answer for parents is whether their child will have problems learning how to read. It has been long held that children with speech sound disorder (SSD) have poorer outcomes in reading ability (reading disability, RD). In fact, research suggests that SSD and reading disabilities have a comorbidity rate of 25-30%. It is thought that an underlying phonological processing deficit is responsible for both RD and SSD. However, research also points towards a more complex association between SSD and RD. For example, when a child’s difficulty only lies in articulation and does not involved language ability, literacy outcomes are quite good. However, if the deficits include vocabulary development (semantics) or sentence construction (syntax) the child is more likely to develop a reading disability. Yet, previous research has also shown that that if a child’s speech (e.g., articulation) delays persist to when the child is beginning to read, such delays will have an impact on reading skills. Therefore, it is unclear whether reading disabilities are affected by language delays in the context of SSD, by persistent speech difficulties, or by a combination of these two factors.
Peterson and colleagues published an article in the Journal of Speech, Language and Hearing Research last month that investigated whether it was language ability or persistence of speech sound disorder that contributed to the development of reading disorders. They took one hundred twenty-three 7- to 9-year-old children, including 86 with a history of SSD and 37 with normative language development (controls). The children completed various measures of language and reading ability. The original assessment of these children occurred when they were between 5-6 and then again at 7-9.

As expected, the children with SSD did more poorly than the controls in reading ability. Within this group, language impairment, but not SSD persistence, predicted poor literacy outcomes. This would suggest that SSD persistent does not play a role in reading development. However, SSD persistence was related to deficits in phonological awareness, which is in turn associated with some types of reading difficulties. In addition, the authors found that those with SSD persistence AND a language impairment (LI) had the worst outcomes, with 66% of these children also showing a reading disability.

Their other hypothesis was that broad language ability at the age of 5-6 would predict later literacy better than persistence of speech errors. This was also confirmed. It appears that phonological processing alone does not predict reading ability. Instead, multiple deficits is a better predictor of reading disorders. Children who had only a SSD did as well as the controls on reading ability tests.

The study had a few limitations. It did not assess reading ability in children older than 7-9. Since reading demands become more complex in later years, it is unknown if SDD persistence alone would affect future reading abilities.

For the working speech language pathologist it is critical that we look at both language and articulation when a child comes to us with a speech delay. For parents, it is also important to ask about the childs broader language abilities even when, on the surface, it appears that the child only has an articulation/speech impairment. Thus, it is important to know if there are underlying language issues that need to be addressed in order to prevent reading difficulties. Most importantly, nonverbal intelligence, vocabulary development and syntax should be targeted.

The reference: Peterson, R., Pennington, B., Shriberg, L., & Boada, R. (2009). What influences literacy outcome in children with Speech Sound Disorder? Journal of Speech, Language, and Hearing Research DOI: 10.1044/1092-4388(2009/08-0024)

Proactive aggression: Ill hit if you dont hit back

By Nestor Lopez-Duran PhD

Aggression researchers have increasingly supported the notion of two specific types of aggressive behaviors: proactive and reactive aggression. Reactive aggression is usually fear-based and impulsive in nature. We all remember the child that would hit at the sightliest sense of threat or anxiety. In contrast, proactive aggression is predatory and calculated such as what you see in some types of bullying behaviors. Kids with high levels of proactive aggression are not necessarily reacting to the perception of threat, but instead may engage in aggression coldly to obtain rewards or impose their will.

Studies with both humans and non-human primates have shown that these two types of aggression have distinct physiological profiles. For example, last year we published a study showing that reactive aggressive children have significant higher endocrine responses during a stress task while proactive aggressive children do not (see Lopez-Duran et al. 2008 DOI: 10.1007/s10802-008-9263-3). However, less is known about how these two types of aggressive behavior differ in their underlying neurocognitive processes. For example, we know little about how executive functioning (e.g., planning and response inhibition) as well as cognitive bias (hostile attributions) affect the presence of reactive and proactive aggression in children.

A study scheduled to be published in the Journal of Abnormal Child Psychology explored neurocognitive and social cognition processes in reactive and proactive aggression. The authors examined 83 boys ages 9 to 12. These children completed a neuropsychological battery of tests that measured various aspects of executive functioning. Teachers also completed a measure of child aggression that assesses proactive and reactive aggressive behaviors. Finally, the authors measured attributional biases using a story-based measure that explores the extent to which children attribute hostile intentions to others in ambiguous situations.

The results:

  1. Problems with response inhibition were associated with reactive aggression but not with proactive aggression.
  2. Surprisingly, hostile attributions was not directly related to either type of aggression. However:
  3. Hostile attributions interacted with planning deficits in predicting both types of aggression. Here the story gets equally interesting and complicated.

As you can see below, when kids had high levels of hostile attributions, deficits in planning ability predicted reactive aggression. In contrast, with low levels of hostile attributions, the deficits in planning ability predicted lower levels of reactive aggression.

So if you think others have hostile intentions AND you have difficulties planning, you may be more likely to show reactive aggression. But if you dont think others have hostile attributions, your difficulties in planning actually reduces your changes of aggression.

The opposite pattern was found for proactive aggression. As seen below, for those with hostile attributions, poor planning ability was associated with a decrease in aggression. In contrast, for those with low levels of hostile attributions, poor planning ability was associated with increased aggression.

What does this mean? It seems that in proactive aggressive kids, hostile attributions may increase the demands for planning. So for example, if you are a predatory bully, thinking that another child has hostile attributions and may fight back requires you to plan more. Thus, if you have deficits in planning ability you may be less likely to engage in the aggressive act.

In sum, this article provides an interesting picture on how executive functioning and social cognition interact in the phenomenology of aggressive responses. The results confirm the notion that reactive aggression is associated with deficits in response inhibition, which have implications for the development of treatment interventions for these children. The interaction of planning ability and hostile attribution is intriguing and likely to also help us develop better treatment interventions for proactive aggression.

The reference: Ellis, M., Weiss, B., & Lochman, J. (2009). Executive Functions in Children: Associations with Aggressive Behavior and Appraisal Processing Journal of Abnormal Child Psychology DOI: 10.1007/s10802-009-9321-5

How your baby moves can predict her IQ

By Nestor Lopez-Duran PhD

It is fascinating that infant movements can serve as a window into their developing brain. Numerous studies have shown that the quality of infant movements, especially among premature babies, strongly predicts whether the infant will have motor and neurological problems. The basic idea is that in normal development, infants move in very predictable ways, such as deviations from this norm may reflect anomalies in brain development.

Most of the original research reports on infant movements have been focused on motor problems. It makes sense that motor movements would reflect the development of motor regions of the brain. However, some researchers have suggested that infant movements also reflect the integrity of regions of the brain near the motor cortex that are in charge of cognitive and emotional control. Thus, it is sensible to predict that anomalies in infant motor movements may also predict cognitive and social functioning later in life.

In an upcoming issue of the Journal of Child Psychology and Psychiatry, Dr. Phillipa R. Butcher and a team from the University of Groningen report findings from a large longitudinal study of preterm infants. The authors used video tapes made at 11 to 16 weeks post-term in 65 infants born at or before 33 weeks of gestation. These infants were the followed for many years and completed a battery of neurocognitive tests when they were 7 to 11 years of age.

The authors were primarily interested in three types of movements:

1. Fidgeting Movements: These are small, circular movements of varying speed that appear around 6 weeks post-term.

2. Concurrent Movements. At this age, normative concurrent movements included kicking, manipulating clothing, and playing with fingers.

3. Concurrent Postural Patterns. These include for example the ability to hold the head in the midline and manipulating fingers so that the fingers are independent of one another (as opposed to always having the fists closed or open).

The results:

  1. While controlling for maternal IQ and attention problems, an index of motor quality score was a significant predictor of total IQ, Verbal IQ, and Performance IQ when the child was between 7 and 11 years of age.
  2. This association was driven exclusively by the presence and absence of normal and atypical postural patterns. That is, fidgeting movements and concurrent movements did not predict IQ, but it was postural patterns that was the strong predictor of IQ.
  3. There was no association between infant movements and behavior or emotional problems during middle childhood (internalizing and externalizing behavior problems).

The graphic below present a very clear picture of the findings. Note for example that the proportion on children with IQ in the 100 to 115 rage increased linearly as a function of the presence of normal postural patterns during infancy. Among those with less than 2 patterns, none of the children scored in the 100 to 114 rage, while among those with more than 2 postural patterns more than 50% of the children scored in that 100-114 range.

IQ scores as a function of infant postural patterns

The authors then commented on one additional important finding. The association between postural patterns and IQ was not explained by the presence of neurological problems. That is, even among the kids without clear neurological problems (such as Cerebral Palsy), infant movements still predicted IQ scores. These findings have important implications of preventive interventions. For example, the careful examination of infant motor patterns may help us determine which children may be at higher risk for cognitive deficits and could benefit from  intensive early intervention programs.
The Reference: Butcher, P., van Braeckel, K., Bouma, A., Einspieler, C., Stremmelaar, E., & Bos, A. (2009). The quality of preterm infants’ spontaneous movements: an early indicator of intelligence and behaviour at school age Journal of Child Psychology and Psychiatry DOI: 10.1111/j.1469-7610.2009.02066.x

Bully victims may be at risk for developing psychotic symptoms

By Nestor Lopez-Duran PhD

Research has consistently shown that the consequences of bullying are severe and range from impaired academic performance to increased risk for suicide. A smaller, but not less influential, line of research has examined the association between severe psychotic disorders (for example schizophrenia) and history of abuse. This research has shown that adults who experience psychotic disorder are more likely than non-affected adults to have a history of childhood trauma, including peer victimization. Could this mean that bullying may increase the risk for developing schizophrenia?

One way to start to examine this question would be to explore whether childhood victimization predicts the presence of early signs of psychotic disorders. This is the strategy employed by a team of British researchers who published their findings in this months issue of the prestigious Archives of General Psychiatry. This population-based study examined over 6,000  children at the ages of 8, 10, and 12 who were participating in a longitudinal study of human development in England. The authors measured the history of victimization at age 8 and 10 as predictors of psychotic symptoms at age 12. Psychotic symptoms included the presence of hallucinations (e.g., seeing or hearing things that are not there) or delusions (e.g., believing that people can read you thoughts).

The results:

  1. How common is bullying? 2,823 children, or 46% of the sample reported experiencing some type of bullying. 14% of the sample reported chronic victimization.
  2. Being victimized during middle childhood doubled the risk of experiencing definite psychotic symptoms in early adolescence (OR 1.94).
  3. The frequency of bullying was a key predictor of psychotic symptoms. Specifically, experiencing chronic bullying increased the risk of having psychotic symptoms by 252%.
  4. The type of bullying also played a role. While all types of bullying predicted an increase in the risk for psychotic symptoms, experiencing overt victimization (being beaten) combined with experiencing relational victimization (social exclusion, spreading rumors, etc) increased the risk of psychotic symptoms by 360% when compared to those who did not experience victimization.
  5. These findings remained stable after controlling for a number of potential explanatory variables, such as prior psychopathology, family adversity, and IQ.

Do these results indicate that victimization cause psychotic symptoms? No. The results are consistent with the hypothesis that victimization may lead to psychotic symptoms, but the nature of the study prevents us from making statements about causation. Although we use the terms increase the risk for developing x, this terminology is actually statistical terminology that refers to the probability for finding a specific outcome at a specific time. For example, in regards to the finding #2, being victimized in middle childhood increased the probability that the child would have psychotic symptoms at age 12. This does not address the question of why or how such probability is increased.

The authors correctly discussed this issue. Specifically, there is the possibility that children who were on path to developing psychotic disorders also engaged in behaviors during early childhood that made them more likely to be victims of bullying. In such a case, being victimized does not cause the psychotic symptoms. Instead, being victimized may have been the result of factors (such as extreme shyness) associated with later development of psychotic symptoms.

However, it is interesting that the authors found a dose response. That is, the more bullying the child experienced the higher the possibility of experiencing psychotic symptoms.  Although one could argue that those at greater risk for developing psychotic symptoms elicited more frequent and severe bullying episodes, dose response effects are usually observed mostly in situations whether the predictor (in this case bullying) has a causative role in the outcome (psychotic symptoms). So this dose effect supports the notion that peer victimization may contribute to the development of psychotic symptoms in childhood and adolescence.
The reference: Schreier, A., Wolke, D., Thomas, K., Horwood, J., Hollis, C., Gunnell, D., Lewis, G., Thompson, A., Zammit, S., Duffy, L., Salvi, G., & Harrison, G. (2009). Prospective Study of Peer Victimization in Childhood and Psychotic Symptoms in a Nonclinical Population at Age 12 Years Archives of General Psychiatry, 66 (5), 527-536 DOI: 10.1001/archgenpsychiatry.2009.23

Be Sociable, Share!

  • Tweet

Concerta for ADHD: A placebo controlled study of methylphenidate and attention problems

By Nestor Lopez-Duran PhD

A university-based randomized, placebo controlled research study of Concerta (methylphenidate), examines the effects of methylphenidate in regulating attention lapses.

Research studies on the neurocognitive profiles (memory, attention, executive functioning, etc) of kids with attention-deficit-hyperactivity disorder have one common denominator: there is no unified or common neuropsychological profile that characterizes ADHD. That is, there is no specific pattern of memory, language, attention, etc., deficits that are universally common in kids of ADHD. Such diverse neuropsychological profile reflects the heterogeneity and diversity of ADHD itself. However, there is one finding that is consistently observed in most studies. When compared to non-affected kids, those with ADHD have significantly variable or uneven performance across most tasks. This is noticeable on cognitive tasks that require sustained periods of attention. On these tasks, children with ADHD show marked within-task variability, oscillating between normative and impaired performance. Thus, instead of consistent impaired attention, most kids with ADHD show variable attention, or rapid lapses in attention.

Yet, despite the popularity of stimulant medications in the treatment of ADHD, there has been no single randomized, placebo-controlled study on the effects of methylphenidate on lapses of attention. The last issue of the Journal of Abnormal Child Psychology published a very sound study on the effects of methylphenidate on attention problems.

The study was conducted at the University at Buffalo (SUNY) and included 49 children with a diagnosis of ADHD. The sample included 39 boys and 10 girls, age 9 to 12, of average IQ, and average academic achievement. Twenty nine were diagnosed with ADHD-combined type, 8 were diagnosed with ADHD-inattentive type, and 2 were diagnosed with ADHD-hyperactive/impulsive type. All kids were not taking medication at the start of the trials. Specifically, those kids who were taking stimulant medication were asked to stop at least 24 hours prior to the first day, and those who were taking Strattera were asked to stop the medication at least 7 days before the first day (because Strattera takes longer to clear from the body). These kids were randomly assigned to a medication or a placebo condition. The medication consisted of long active Concerta provided at dosages ranging from .3 to .6 mg/kg.

During the testing, the children completed a simple computerized test of attention called the X and O Discrimination Task, which is very sensitive to lapses in attention. Specifically, the kids were presented with either the letter X or the letter O, and they had to respond by pressing two different corresponding keys as fast as possible. The task included 10 practice trials and 100 task trials.

To measure variability the authors focused on the right tail of reaction time distribution. When a child is presented with the X or O the computer records how long it takes the child to respond (in milliseconds). Since ADHD is characterized by variability in attention, the interest in the responses to the X or O task was not really on how how fast the child responded, but instead on how variable the child responded. For example, a child with little variability in responses would look like this:

Note that on the horizontal axis you can see the milliseconds and on the vertical axis you can see the number of times the child responded at each speed. In this example, the child responded 50 times at 500ms, 40 times at 600ms, and 10 times at 700ms. This could be considered a highly stable pattern of performance in that most of the responses were between 500 and 700ms. Now compare that performance to a performance like this:

In this case the responses are still highly stable (all responses within 300ms of each other) but the responses were slower. So this is what you would see if the child was simply slow in responding. Now, compare the two previous graphs to the pattern of responses below:
In this final case, the child was not necessarily slow, but the responses were more variable in that in some trials the child was very fast (in 5 of the trials the child responded within 400ms) while in some trials the child was very slow. Since this type of variability (rather than simple slowness) is commonly observed in ADHD, the authors examined whether Concerta affected the variability in responses during the task.

The results:

1. The authors found a significant effect of the medication on speed, and this effect was dose-dependent. That is, those taking the placebo were significantly slower responding to the task than kids taking a low dose of methylphenidate. Furthermore, kids that were taking a low dose of methylphenidate were significantly slower than those taking a high dose of the medication. Therefore, the results show a strong effect of the medication in increasing speed of responding.

2. The authors also found that the medication resulted in a significant reduction in variability (deviation from the mode), however this was not dependent upon dosage. That is, kids who took the medication showed significantly more consistent/stable responses than kids who took the placebo, but there was no difference in response variability between those taking a low vs. a high dosis of methylphenidate.

3. The increase in speed was not at the cost of accuracy. On the contrary, Kids taking the medication were not only faster and less variable than those taking the placebo, but they were also more accurate.

This study thus provides strong empirical support for the effectiveness of Concerta in facilitating attentional processes. Specifically, methylphenidate seems to 1) improve speed of responding (likely by facilitating sensory-motor processing) and 2) reduce variability in performance (likely by reducing lapses in attention).

Now some final thoughts about this study. This study was conducted by an academic team (university-based) and was financed by the National Institutes of Health (NIH) NOT by the makers of the drug.
Reference: Journal of Abnormal Child Psychology DOI: 10.1007/s10802-009-9316-2

Weight loss for teens: Family support effective. Family therapy, not so much.

By Nestor Lopez-Duran PhD

A family-based psychoeducational weight loss treatment for teens appears to be more effective than traditional family therapy.

Psychologists often work on the premise that any therapy is better than no therapy. Usually this tenet applies, in that most research shows that for most conditions therapy is better than no therapy. But unfortunately there are exceptions. Some interventions, in very specific conditions, may actually do more harm than good. It seems that short-term family systems therapy as part of a weight loss program for teenage girls may be one of such cases.

A team from the University of South Carolina conducted a controlled clinical trial of a family-based weight loss intervention program for teen girls. The authors were primarily interested in examining the effects of two family variables in weight loss: family nurturance (warmth) and family cohesion (closeness). These family factors are associated with a number of positive behavioral outcomes in teens, so it was hypothesized that interventions that help foster nurturance and cohesion would facilitate the effectiveness of weight loss programs.

In the study the authors randomly assigned 42 families with an overweight teenage girl to one out of three treatment conditions: 1) a family-based psychoeducational program, 2)a family-based psychoeducational program PLUS a multi-family therapy group, and 3) and a non-treatment (waiting list) condition. The psychoeducational program encouraged weight-loss behaviors and parental support but it did not include a specific caloric restriction goal, as the authors indicated that this is usually not effective with adolescents. The multi-family therapy program consisted of weekly 45 minute sessions of traditional group therapy during which families were able to address many issues, including the challenges they were encountering during the implementation of the psychoeducational program. Finally, the waiting list group simply waited until the end of the first phase of the study and then were given the opportunity to receive the treatment. These interventions lasted for 16 weeks.

The Results:
After 4 months, the average BMI of the participants of all three groups did not change significantly, and this index was practically identical between the groups. That is, the two treatment conditions were not more effective than the non-treatment group in reducing the body mass index.

However, a major difference between the groups was observed in energy intake. The girls in the psychoeducational group only displayed a significant reduction in energy intake. Instead, the girls in the family therapy plus psychoeducation displayed an increase in energy intake and this was comparable to the increase observed in the non-treatment group. Therefore, it seems that the psychoeducational intervention helped these girls eat less, but when family therapy was added, the benefits of the psychoeducational intervention vanished.

Concerning the effects of family nurturance and cohesion. The authors found that families that improved in family nurturance during the intervention also showed a significant reduction in energy intake. This suggests that family warmth and support helps adolescent girls lower their food intake.

Finally, family therapy led to higher levels of family conflict. The authors suggested that conflict issues may have arisen during therapy that were not fully addressed given the duration of the treatment. However, 4 months of therapy is within the norm for group therapy programs, thus it seems that this particular form of therapy, for this particular issue and population (weight loss in overweight but otherwise psychologically healthy teenage girls) is not effective in facilitating weight loss and may actually lead to increased family conflict. Please note however, that family systems therapy has been found effective for many conditions, including eating disorders (e.g., anorexia). This study ONLY assessed family therapy as part of a weight loss program in otherwise healthy girls.
Reference:Kitzman-Ulrich, H., Hampson, R., Wilson, D., Presnell, K., Brown, A., OBoyle, M. (2009). An Adolescent Weight-Loss Program Integrating Family Variables Reduces Energy Intake Journal of the American Dietetic Association, 109 (3), 491-496 DOI: 10.1016/j.jada.2008.11.029