Morning bright light therapy: a helpful tool for reducing comorbid symptoms of affective and behavioral dysregulation in juvenile depressed inpatients? A pilot trial
Abstract
Abstract.Objective: In recent years, bright light therapy (BLT) has been used to treat depression and to stabilize circadian rhythms. In this study we evaluated whether it is also helpful for comorbid symptoms of affective and behavioral dysregulation in depressive inpatients. Method: This article reports a secondary analysis comparing two subgroups of depressive participants with comorbid affective and behavioral dysregulation, captured with the dysregulation-profile of the Strengths and Difficulties Questionnaire (SDQ-DP; n = 16 vs. n = 11). Participants were randomly allocated to active BLT (10,000 lux) or control BLT (approx. 100 lux), and received 45 minutes of BLT for 2 weeks. SDQ-DP scores, sleep parameters, and circadian preference were assessed at baseline, after the intervention, and 3 weeks later. Results: No direct effects on SDQ-DP scores were observed. Sleep improved in both conditions. Only in the active BLT condition was a circadian phase advance found. Correlation and regression analyses indicated an indirect, circadian effect for improved SDQ-DP scores. Conclusions: The data of this pilot trial should be considered preliminary and merely descriptive. Further research is warranted.
Zusammenfassung.Fragestellung: Lichttherapie wird zur Behandlung von Depressionen sowie zur Stabilisierung des zirkadianen Rhythmus’ eingesetzt. Die vorliegende Pilotstudie untersucht die Wirksamkeit von Lichttherapie bei stationären depressiven Jugendlichen mit Symptomen komorbider affektiver und behavioraler Dysregulation. Methodik: In dieser Studie werden Ergebnisse einer Sekundäranalyse berichtet, in der zwei Gruppen von depressiven Patienten mit komorbiden Symptomen affektiver und behavioraler Dysregulation (erfasst durch das Dysregulations-Profil des Fragebogens für Stärken und Schwächen – SDQ-DP; n = 16 vs. n = 11) randomisiert entweder der aktiven Lichttherapie (10 000 lux) oder der Kontrollbedingung zugeordnet (ca. 100 lux) wurden. Alle erhielten täglich 45 Minuten morgendliche Lichttherapie über einen Zeitraum von zwei Wochen. Werte des SDQ-DP, Schlafparameter und die zirkadiane Präferenz wurden vor Beginn der Studie, nach zwei Wochen Intervention und weitere drei Wochen später erhoben. Ergebnisse: Symptome der affektiven und behavioralen Dysregulation zeigten keine direkten Verbesserungen als Folge der Lichttherapie. Schlafparameter zeigten Verbesserungen unabhängig der Lichtbedingung. Eine Vorverlagerung des zirkadianen Rhythmus’ wurde nur bei den Teilnehmern der aktiven Lichtbedingung gefunden. Die Ergebnisse der Korrelationen und Regressionen legen die Vermutung nahe, dass eine Verbesserung der Symptome affektiver und behavioraler Dysregulation durch einen indirekten, zirkadianen Weg möglich ist. Schlussfolgerungen: Weitere Studien sind notwendig, um die vorläufigen Ergebnisse dieser Pilotstudie zu replizieren.
Introduction
A mixed psychiatric condition conceptualized as severe affective and behavioral dysregulation in children and adolescents has gained interest in recent years and can be captured by specific dysregulation subscales derived from existing items on either the Child Behavior Checklist (CBCL-Dysregulation profile [CBCL-DP]; Althoff, Verhulst, Rettew, Hudziak & van der Ende, 2010; Holtmann et al., 2007; Hudziak, Althoff, Derks, Faraone & Boomsma, 2005) or the Strengths and Difficulties Questionnaire (SDQ-Dysregulation profile [SDQ-DP]; Holtmann, Becker, Banaschewski, Rothenberger & Roessner, 2011; Woerner, Becker & Rothenberger, 2004). Predominant symptoms of children and adolescents with severe affective and behavioral dysregulation include irritability, aggression, “affective storms,” hyperarousal, mood instability, and suicidal behavior (Holtmann et al., 2007). Moreover, severe dysregulation of affect and behavior has consistently been found to be associated with difficulties falling sleep and sleeping through the night as well as a reduced need for sleep (Ayer et al., 2009; Holtmann, Bölte, Goth & Poustka, 2008; Legenbauer, Heiler, Holtmann, Fricke-Oerkermann & Lehmkuhl, 2012; Mehl et al., 2006). Since a complex interaction exists between sleep and both internalizing and externalizing symptoms (Fallone, Owens & Deane, 2002; Meijer, Reitz, Dekovic, van den Wittenboer & Stoel, 2010), one gateway for the treatment of severely dysregulated youth – and for the prevention of devastating long-term consequences – may lie in the resolution of their sleep problems. Interventions targeting sleep disturbances in youth showed improvements of both behavioral and emotional functioning (Harvey, Mullin & Hinshaw, 2006).
One type of intervention influencing sleep disturbances and shifting circadian preference is morning bright light therapy (BLT; Gooley, 2008). The use of BLT as an adjuvant, innovative treatment option for affective and behavioral problems was recently proposed (Heiler, Legenbauer, Bogen, Jensch & Holtmann, 2011). Initially developed as a treatment for seasonal affective disorder, BLT was subsequently shown to improve depressive symptoms, sleep quality, and reset circadian parameters (Even, Schroeder, Friedman & Rouillon, 2008). In a primary analysis of a 2-week randomized controlled study on the feasibility and clinical effects of BLT in 57 depressed juvenile inpatients, compared to a dim light control condition BLT led to specific improvements of subjective sleep parameters as well as to a shift of circadian preference toward morningness (Bogen, Legenbauer, Gest & Holtmann, 2015). Enhanced sleep quality and the circadian phase advance largely predicted the reduction of depressive symptoms. Based on these results, this explorative secondary analysis examines whether BLT also has beneficial effects on the regulation of sleep and circadian phase in youths with comorbid severe affective and behavioral dysregulation as well as on the symptom level of affective and behavioral dysregulation (Heiler et al., 2011). Because one may assume an intimate relationship between sleep, circadian preference, affect regulation, and emotional and behavioral functioning (Neitzert Semler & Harvey, 2005; Van Dongen, Maislin, Mullington & Dinges, 2003; Gau et al., 2007; Harvey et al., 2006), we hypothesize that changes in sleep quality and sleep restoration, shifts of circadian preference and improvements in affective regulation are correlated with each other. Improvements of sleep and a circadian phase advance as a result of BLT are assumed to be of predictive value for a reduction of affective and behavioral problems.
Methods
Sampling
A group of 57 medication-naïve participants between 12 and 18 years were recruited from an inpatient psychiatric care setting. All of them had a diagnosis of moderate to severe depression according to ICD-10. Thereof, 27 participants additionally showed symptoms of affective and behavioral dysregulation as assessed with the SDQ-DP. For a detailed participant flow chart of the original sample see Bogen et al. (2015).
Design and Procedure
One week after admission to a tertiary care hospital for child and adolescent psychiatry, eligible participants were informed about study objectives and the study design by one of the main investigators. They randomly received 2 weeks of either active (10,000 lux) or inactive (100 lux) BLT in addition to treatment as usual on the respective units (TAU – including individual and group psychotherapy in addition to medical, psychological, and nursing care; approaches such as in-hospital schooling, occupational therapy, arts therapies, sports and movement therapy, and family-focused interventions). In order to reduce possible expectation bias, participants were informed that the aim of the study was to compare to different types of BLT. Interventions were conducted following the “Clinician’s Manual for Light and Wake Therapy” (Wirz-Justice, Benedetti & Terman, 2009), and were conducted five times a week, 45 minutes each, for a total duration of 2 weeks. Behavioral assessments were made before entering the study (T1), after 2 weeks of intervention (T2; T1-T2 → 2 weeks in between), and 3 weeks later (T3; from T2-T3 → 3 weeks in between).
This study was approved by the Medical Ethical Committee of the Ruhr University Bochum, Germany (Reg.-No. 3996-11). Written consent was given by both the participant and the person with primary custody. For a more detailed description of the study design and trial procedures see Bogen et al. (2013).
Measures
Strengths and Difficulties Questionnaire (SDQ)
The SDQ was used to identify participants with symptoms of affective and behavioral dysregulation. Normally, the SDQ is conducted as a brief behavioral screening questionnaire comprising 25 attributes assessing emotional symptoms, conduct problems, hyperactivity/inattention, peer-relationship problems, and prosocial behavior (Goodman, 1997, 2001; Klasen et al., 2000). More recently, a score derived from certain existing items of the SDQ has been used to assess symptoms of affective and behavioral dysregulation (SDQ-DP; Holtmann et al., 2011a) and to monitor short-term changes of SDQ-DP scores during the BLT trial (Goodman, 1997). The following items were identified as capturing this so-called SDQ-DP: Item 13 “often unhappy, down-hearted or tearful”; item 8 “many worries, often seems worried”; item 12 “often fights with other children or bullies them”; item 22 “steals from home, school or elsewhere”; item 2 “restless, overactive, cannot stay still for long”. The self-report version of the SDQ was used since self-ratings from adolescents were shown to better contribute to the diagnostic process in the clinical setting than ratings by parents (Arman, Amel & Maracy, 2013; Van der Meer, Dixon & Rose, 2008). Psychometric properties of the SDQ-DP were robustly tested and validated (Holtmann et al., 2011a; Woerner et al., 2004). Instructions were given that the SDQ was used to assess affective and behavioral symptoms over the last week.
Sleep Questionnaire B – revised (SF-B/ R)
The SF-B/R was used to subjectively assess sleep and sleep difficulties (Görtelmeyer, 2011). It contains 31 questions assessing the last 2 weeks. Sleep indices can be calculated including “sleep quality” and “restorative sleep.” For clinical populations, internal consistency can be considered moderate to excellent (alpha = .68 to .92). Test-retest reliability indicates that relatively stable sleep-related behavior and experiences are assessed (r = .53 to r = .91).
Morningness – Eveningness Questionnaire (MEQ)
The MEQ was administered to estimate circadian preference (Griefahn, Künemund, Bröde & Mehnert, 2001; Horne & Ostberg, 1976). This questionnaire assesses circadian preference based upon 19 questions that can be added up to a total score, with higher scores indicating “morningness” (scores ranging from 59 to 86: moderate to definite morningness) and lower scores implying “eveningness” (scores ranging from 16 to 41: moderate to definite eveningness), with a neutral circadian preference with scores in between (scores ranging from 42 to 58: neutral type). Full-scale internal consistency can be considered sufficient (.82; Shahid, Khairandish, Gladanac & Shapiro, 2012); a high test-retest reliability was repeatedly reported with up to r =.96 (Griefahn, 2002; Kerkhof, 1984).
Statistical Analyses
Secondary analyses of the data were also conducted (for primary data analyses, see Bogen et al., 2015). The analyses were based on an intention-to-treat sample (ITT sample). Missing values were replaced by using “last-observation-carried-forward” values. This method has shown to be a very conservative data substitution, likely to underestimate within-group changes (Prakash, Risser & Mallinckrodt, 2008). In order to investigate the course of means of SDQ-DP scores, sleep quality, restorative sleep, and circadian preference during BLT, 3 × 2 repeated measures GLM analyses were conducted with time x light condition, with an alpha level set at .05 to detect significant interactions. Furthermore, bivariate Pearson’s correlations were calculated to investigate possible linear relationships between changes in SDQ-DP scores, sleep quality, restorative sleep, and circadian preference using difference scores (T1-T2 and T1-T3, respectively). Standard multiple regression analyses were performed to further analyze the predictive value of changes in sleep parameters and circadian preference on changes in SDQ-DP scores.
Results
Sample Characteristics
Based on self-report ratings, 27 out of 57 participants (47 %) were classified as SDQ-DP positive (exceeding the SDQ-DP cutoff of 5). Significant group differences at baseline were found for age and MEQ total score, and were taken into consideration in further analyses1. No significant group differences could be found in relation to gender or number of co-morbidities (p < .05). Details are shown in Table 1.
Differences Between BLT Conditions for Sleep Parameters, Circadian Preference and SDQ-DP Scores During the Course of BLT
Results of sleep parameters including sleep quality and restorative sleep indicated that participants in both BLT conditions showed improvements, so that GLM measures did not reveal significant time x light condition interactions. If we take into account baseline group differences for circadian preference, the scores of participants in the active BLT condition shifted more toward morningness than scores of the control condition, resulting in a statistically significant time x light condition interaction effect. For test statistics please see Table 2.
Against our expectations, BLT did not have an additional direct influence on the symptom level of affective and behavioral dysregulation as measured with the SDQ-DP. Although active BLT proved to positively influence the course of SDQ-DP scores, especially in the short term, GLM repeated measures revealed no significant time x light condition interaction effect. Please see the course of SDQ-DP scores in Figure 1.
Correlations Between Changes in SDQ-DP Scores, Sleep Parameters, and Circadian Preference
We found significant moderate to large negative correlations between difference scores of SDQ-DP with difference scores of sleep quality, restorative sleep and circadian preference. Details are shown in Table 3.
Predictive Value of Sleep Parameters and Circadian Preference for SDQ-DP Scores
In order to estimate the proportion of variance in the change of SDQ-DP scores that can be accounted for by changes of sleep quality, restorative sleep, and circadian preference, we performed standard multiple regressions. Directly after 2 weeks of BLT (T1-T2), the regression model did not explain a significant proportion of variance. Including follow-up measures (T1-T3), a total of 50 % of variance of SDQ-DP difference scores were explained by the coefficients. Changes of sleep quality thereby showed to be a statistical significant predictor for changes in SDQ-DP scores. Statistical indices of the linear regression models are displayed in Table 4.
Discussion
The results of this explorative analysis indicate no additional direct effect of BLT on the symptom level of affective and behavioral dysregulation as assessed with the SDQ-DP. However, the results of the correlation and regression analyses indicated that improved sleep parameters and a circadian phase advance were associated with a reduction of SDQ-DP scores. Only improved sleep quality was of predictive value for a reduction of SDQ-DP scores in the long-term. Based on these results, one might assume that improved sleep and an advanced circadian phase might have positively influenced SDQ-DP scores as a secondary effect by an indirect, circadian mechanism, as was already assumed by Heiler et al. (2011).
Whereas a circadian phase advance was observed in the active BLT condition, neither direct light-induced nor indirect secondary influences of the circadian phase advance on sleep and SDQ-DP were found. Both the daily structures on the units and the therapeutic setting may have positively affected sleep quality and restorative sleep beyond the additional effects of 2 weeks of added BLT. Another aspect to consider is a possible placebo effect, which might have led to improved sleep in both conditions. Placebo effects are relatively high in BLT trials since it is difficult to “blind” a participant when using broad-spectrum intense white light as the active condition (Golden et al., 2005). Existing approaches to find a suitable placebo are the use of either dim light (< 300 lux), light of a specific wavelength (dim red light), or using low-density negative air ion exposure as placebo conditions (Goel et al., 2005). However, these approaches often led participants to guess which condition they have been allocated to, thereby influencing expectations. To minimize expectation bias elicited by either bright or dim light, in this trial all participants, irrespective of group allocation, received the same kind of neutral information regarding the use and efficacy of BLT. Since this study was conducted in an acute psychiatric setting and not in a sterile laboratory, setting effects as well as expectation bias cannot fully be excluded and might have elicited additional effects. Further research including different types of placebo conditions is warranted to clarify this issue.
It was assumed that, next to improved sleep, a circadian phase advance would also be reflected in the regression models as an essential factor for a reduction of SDQ-DP scores as was also found by Ryback et al. (2006). This, surprisingly, was not the case in this sample. It is common knowledge that circadian preference, part of normative development in adolescence, shifts toward eveningness (Crowley, Acebo & Carskadon, 2007; Kim, Dueker, Hasher & Goldstein, 2002). It is, however, unclear how much time is needed for a circadian phase advance to become clinically relevant or to indirectly influence other symptoms such as affective and behavioral dysregulation as secondary effect. Since circadian preferences show rather long-term effects on affect and behavior (Gau et al., 2007; Goldstein et al., 2007; Lange & Randler, 2011; Van der Heijden et al., 2013), the time frame in this sample might have been too short to assess the follow-up effects of a shifted circadian rhythm.
Limitations
Several limitations have to be discussed. First of all, this article reports results of a secondary analysis including subgroups with only few participants of a larger sample, which is a main weakness. Also, BLT was added to TAU on the respective units. TAU was identical for all participants, although we did not assess, for example, the number of psychology sessions or potential other confounders such as menstrual phase, another limitation of the study. Affective and behavioral problems were assessed by using the SDQ-DP score (Woerner et al., 2004), which only constitutes a screening measurement and cannot be compared to a thorough clinical evaluation. There are no norms for the SDQ for a 2-week window as it was used in this study. However, it has proved to be a robustly tested and validated possibility of assessment (Holtmann et al., 2011a). Whereas in population-based studies prevalence rates of affective and behavioral dysregulation generally vary between 1 %–3.8 %, in clinical contexts the rate is considerably higher. It is important to keep in mind that prevalence of affective dysregulation may vary due to age (Jucksch et al., 2011). In the present sample, however, the rate of participants with a positive SDQ-DP was considerably higher with 47 % of all participants. One has to consider that the SDQ-DP partly assesses depression-related symptoms with the items “often unhappy, down-hearted or tearful” as well as “many worries, often seems worried”. Since all of the participants had a diagnosis of moderate to severe depression, one might assume that the high rate of SDQ-DP positive in this sample might rather have reflected the severity and variety of affective and behavioural difficulties instead of the specific phenotype discussed in the literature (Holtmann et al., 2011; Jucksch et al., 2011).
Another limitation constitutes the short time frame in which BLT treatment was possible. Although a short duration (2 weeks) seems to be just as effective as a longer duration (5 weeks; Levitt & Levitan, 2003) for depressive symptoms, for symptoms of affective and behavioral dysregulation no empirical recommendations are yet available. This study was conducted in an intensive psychiatric care setting. Therefore, taking into account the first week in which the participants got the possibility to familiarize with the setting, daily interventions and the possibility to monitor follow-up effects were restricted. The effects of sleep and circadian preference on affect can, particularly, be seen in the long term (Gau et al., 2007; Goldstein et al., 2007; Lange & Randler, 2011; Van der Heijden et al., 2013). In future studies, it seems essential to monitor the long-term effects of BLT to follow-up on the secondary effects of a circadian shift toward morningness. It can be assumed that more time is needed for a phase advance to positively affect affective and behavioral dysregulation.
In summary, although no direct effects on the symptom level of affective and behavioral dysregulation were found, results indicated that sleep quality and a shift of circadian preference toward morningness might positively influence SDQ-DP scores by an indirect, circadian mechanism. These results seem promising, although they need further testing, since the data presented must be considered as preliminary with merely descriptive, hypothesis generating results that should be interpreted cautiously.
Funding
This work was supported by the LWL Research Institute Bochum, Germany [grant no. V006-2010].
Conflicts of interest: MH served in an advisory or consultancy role for: Lilly, Novartis, and Bristol-Myers Squibb, and received conference attendance support or was paid for public speaking by AstraZeneca, Janssen-Cilag, Lilly, Neuroconn, Novartis, Medice and Shire. The present work is unrelated to the above grants and relationships. The other authors have no conflicts of interest.
Literature
(2010). Adult outcomes of childhood dysregulation: A 14-year follow-up study. Journal of the American Academy of Child and Adolescent Psychiatry, 29, 1105–1116.
(2013). Comparison of parent adolescent scores on Strengths and Difficulties Questionnaire. International Journal of Research in Medical Sciences, 18, 501–505.
(2009). Child Behavior Checklist Juvenile Bipolar Disorder (CBCL-JBD) and CBCL Posttraumatic Stress Problems (CBCL-PTSP) scales are measures of a single dysregulatory syndrome. Journal of Child Psychology and Psychiatry, 50, 1291–1300.
(2013). Morning light therapy for juvenile depression and severe mood dysregulation: Study protocol for a randomized controlled trial. Trials, 14, .
(2015). Lighting the mood of depressed youth: Feasibility and efficacy of a 2 week-placebo controlled bright light treatment for juvenile inpatients. Journal of Affective Disorders, 190, 450–456, DOI: 10.1016/j.jad.2015.09.026.
(2007). Sleep, circadian rhythms, and delayed sleep phase in adolescence. Sleep Medicine, 8, 602–612.
(2003). The cumulative cost of additional wakefulness: Dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep, 26, 117–126.
(2008). Efficacy of light therapy in nonseasonal depression: A systematic review. Journal of Affective Disorders, 108, 11–23.
(2002). Sleepiness in children and adolescents: Clinical implications. Sleep Medicine Reviews, 6, 287–306.
(2007). Association between sleep problems and symptoms of attention-deficit/hyperactivity disorder in young adults. Sleep, 30, 195–201.
(2005). Controlled trial of bright light and negative air ions for chronic depression. Psychological Medicine, 35, 945–955.
(2005). The efficacy of light therapy in the treatment of mood disorders: A review and meta-analysis of the evidence. American Journal of Psychiatry, 162, 656–662.
(2007). Time of day, intellectual performance, and behavioral problems in Morning versus Evening type adolescents: Is there a synchrony effect? Personality and Individual Differences, 42, .
(1997). The Strengths and Difficulties Questionnaire: A research note. Journal of Child Psychology and Psychiatry, 38, 581–586.
(2001). Psychometric properties of the Strengths and Difficulties Questionnaire. Journal of the American Academy of Child and Adolescent Psychiatry, 40, 1337–1345.
(2008). Treatment of circadian rhythm sleep disorders with light. Annals Academy of Medicine Singapore, 37, 669–676.
(2011). [SF-A/R and SF-B/R – Sleep questionnaires A and B – Revised]. Göttingen: Hogrefe.
(2002). Einsatz eines Fragebogens (D-MEQ) zur Bestimmung des Chronotyps bei der Zuweisung eines Schichtarbeitsplatzes. Zeitschrift für Arbeitswissenschaft, 3, 142–149.
(2001). Zur Validität der deutschen Übersetzung des Morningness-Eveningness-Questionnaires von Horne und Östberg. Somnologie, 5, 71–80.
(2006). Sleep and circadian rhythms in children and adolescents with bipolar disorder. Development and Psychopathology, 18, 1147–1168.
(2013). Association of eveningness with problem behavior in children: A mediating role of impaired sleep. Chronobiology International, 30, 919–929.
(2011). Severe mood dysregulation: In the “light” of circadian functioning. Medical Hypotheses, 77, 692–695.
(2011). Psychometric validity of the Strengths and Difficulties Questionnaire – dysregulation profile. Psychopathology, 44, 53–59.
(2007). Prevalence of the Child Behaviour Checklist – Pediatric bipolar disorder phenotype in a German general population sample. Bipolar Disorder, 9, 895–900.
(2008). CBCL-pediatric bipolar disorder phenotype: Severe ADHD or bipolar disorder? Journal of Neural Transmission, 115, 155–161.
(2011). Child Behavior Checklist – Dysregulation Profile predicts substance use, suicidality, and functional impairment: A longitudinal analysis. Journal of Child Psychology and Psychiatry, 52, 139–147.
(1976). A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. International Journal of Chronobiology, 4, 97–110.
(2005). Prevalence and genetic architecture of Child Behaviour Checklist – juvenile bipolar disorder. Biological Psychiatry, 58, 562–568.
(2011). Severe affective and behavioural dysregulation is associated with significant psychosocial adversity and impairment. Journal of Child Psychology and Psychiatry, 52, 686–695.
(1984). A Dutch-language questionnaire for the selection of morning and evening type individuals. Nederlandse Tijdschrift voor de Psychologie, 39, 281–294.
(2002). Children’s time of day preference: Age, gender and ethnic differences. Personality and Individual Differences, 33, 1083–1090.
(2000). Comparing the German versions of the Strengths and Difficulties Questionnaire (SDQ-Deu) and the Child Behavior Checklist. European Child and Adolescent Psychiatry, 9, 271–276.
(2011). Morningness-eveningness and behavioural problems in adolescents. Sleep and Biological Rhythms, 9, 12–18.
(2012). The affective storms of school children during nighttime: Do affective dysregulated school children show a specific pattern of sleep disturbances? Journal of Neural Transmission, 119, 989–998.
(2003). Length of light treatment trial: Does it influence outcome. Chronobiology International, 20, 1213–1214.
(2008). Parent and child agreement on reports of problem behaviour obtained from a screening questionnaire, the SDQ. European Child and Adolescent Psychiatry, 17, 491–197.
(2006). Correlates of sleep and pediatric bipolar disorder. Sleep, 29, 193–197.
(2010). Longitudinal relations between sleep quality, time in bed and adolescent problem behaviour. Journal of Child Psychology and Psychiatry, 51, 1278–1286.
(2005). Misperceptions of sleep can adversely affect daytime functioning in insomnia: A meta-analysis of treatment efficacy. Behaviour Research and Therapy, 43, 843–856.
(2008). The impact of analytic method on interpretation of outcomes in longitudinal clinical trials. International Journal of Clinical Practice, 62, 1147–1158.
(2006). An open trial of light therapy in adult attention deficit/hyperactivity disorder. Journal of Clinical Psychiatry, 67, 1527–1535.
(2012). Peeking into the minds of troubled adolescents: The utility of polysomnography sleep studies in an inpatient psychiatric unit. Journal of Affective Disorders, 139, 66–74.
(2009). Chronotherapeutics for affective disorders. A clinicians’ manual for light and wake therapy. Basel: Karger.
(2004). Normative data and scale properties of the German parent SDQ. European Child and Adolescent Psychiatry, 13, 3–11.
1 In this subsample, no significant group differences of depression scores over time during the course of BLT were found. Therefore, depression scores were not considered in secondary analyses (F(2, 48) = 0.335; p = .72).
