Psychiatry Research 215 (2014) 624–627
Contents lists available at ScienceDirect
Psychiatry Research journal homepage: www.elsevier.com/locate/psychres
Association between seasonal affective disorder and subjective quality of the sleep/wake cycle in adolescents Lorenzo Tonetti a,n, Marco Fabbri b, Alex Erbacci a, Monica Martoni c, Vincenzo Natale a a
Department of Psychology, University of Bologna, Viale Berti Pichat 5, 40127 Bologna, Italy Department of Psychology, Second University of Naples, Caserta, Italy c Department of Experimental, Diagnostic and Specialty Medicine, University of Bologna, Bologna, Italy b
art ic l e i nf o
a b s t r a c t
Article history: Received 1 August 2012 Received in revised form 11 December 2013 Accepted 15 December 2013 Available online 19 December 2013
The relationship between seasonal affective disorder (SAD) and subjective quality of sleep/wake cycle in adolescents was explored. The Seasonal Pattern Assessment Questionnaire for Children and Adolescents (SPAQ-CA) and Mini Sleep Questionnaire (MSQ) were administered to 345 adolescents living in the city of Cesena (Emilia-Romagna region, Italy) (299 females; age range: 14–18 years), to determine SAD and perceived quality of the sleep/wake cycle. The response rate was 92% for females and 90.2% for males. The MSQ includes two factors, sleep and wake, with lower scores corresponding to a lower quality of sleep and wake. The MSQ includes cut-off criteria to detect a good or bad sleep and wake quality. Adolescents with SAD (16 7 5.7) scored significantly lower than those not affected on wake factor (19.5 74.3), while no effect has been observed on sleep factor. SAD was the only one significant predictor of good/bad wake quality, while it did not reach significant level with reference to good/bad sleep quality. Present results are indications of a possible influence of SAD on wake quality and further studies are necessary to confirm them. & 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Adolescents Seasonal affective disorder Sleep/wake quality
1. Introduction The relationship between sleep and depression has been deeply explored in the last decades, as reported in three reviews (Riemann et al., 2001; Ivanenko et al., 2005; Tsuno et al., 2005), showing a bi-directional relationship between them. Specifically, some sleep markers of depression are early morning awakening and shorten REM sleep latency (Srinivasan et al., 2009). On the contrary, hypersomnia is related to some sub-types of depression, as the seasonal affective disorder (SAD) (Thompson and Isaacs, 1988; Allen et al., 1993; Tam et al., 1997), which is characterized by recurrent depressive episodes that occur annually in the fall and winter (Rosenthal et al., 1984), with a summer variant of SAD characterized by periodic depressive episodes in this season (Wehr et al., 1987). SAD patients also present awakening difficulties (Avery et al., 2002). It is well known that the prevalence of SAD changes with age. In particular, during adolescence, the frequency of SAD raises compared to childhood (Swedo et al., 1995; Tonetti et al., 2007). This increase has been linked to the strong hormonal changes, occurring during this period (Swedo et al., 1995). To date only two studies (Swedo et al., 1995; Tonetti et al., 2007) have been carried
n
Corresponding author. Tel.: þ 39 051 2091877; fax: þ39 051 243086. E-mail address: [email protected] (L. Tonetti).
0165-1781/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.psychres.2013.12.023
out on the features of SAD during adolescence and the relationship between this disorder and the features of the sleep/wake cycle in this age stage has been neglected. Our study aimed to fill this lack of knowledge, investigating the association between SAD and the perceived quality of the sleep– wake cycle in adolescents. Since adult SAD patients mainly present hypersomnia and awakening difficulties (Thompson and Isaacs, 1988; Allen et al., 1993; Tam et al., 1997; Avery et al., 2002), we could expect a lower wake quality in SAD adolescents compared to those not affected by this disorder. 2. Methods 2.1. Participants Overall 325 females and 51 males were invited to participate. The final sample was composed of 345 participants (299 females), with a response rate of the 92% for females and 90.2% for males. The mean age of the overall sample was 16.88 þ 1.03 (median 17 years and mode 18 years), with age ranging between 14 and 18 years. The males (17.02 7 1.13) and females (16.86 71.02) mean ages were not significantly different (p ¼ 0.33). All participants attended the same high school in the city of Cesena (EmiliaRomagna region, Italy) and were tested during a school meeting in April 2010, aimed to increase students0 knowledge about sleep and well-being. The attending of the school meeting was not mandatory but students received a school credit for their participation. The difference in the number of females and males is due to the fact the investigated high school is mainly attended by females. The time of compilation of both questionnaires was about 15 min. If 18-years-old, participants
L. Tonetti et al. / Psychiatry Research 215 (2014) 624–627 provided written informed consent prior to their participation in the school meeting; if they were underage, the written informed consent was given by parents. The local ethic committee approved the protocol and the study complied with the tenets of the Declaration of Helsinki. 2.2. Measures 2.2.1. Seasonal Pattern Assessment Questionnaire for Children and Adolescents (SPAQ-CA) The Italian version (Tonetti et al., 2008) of the SPAQ-CA (Swedo et al., 1995) which is a self-reporting questionnaire for normal and clinical seasonal mood variations, was administered. Eleven items on a 5-point Likert scale (no change¼ 0, slight change¼1, moderate change¼2, marked change¼3, and extremely marked change¼ 4) evaluate the seasonal changes in the following areas: sleep length, mood, abuse of stimulants (e.g., alcohol, nicotine, and drugs), social activity, weight, energy, appetite, school performance (troubles and marks), irritability and conduct problems. Adding up the scores of those items, we obtain the Global Seasonality Score (GSS), with higher values corresponding to higher mood seasonality (total range score: 0–44). Another question evaluates the extent to which seasonal changes are felt as a problem. If they are not felt as a problem, the corresponding scoring was 0; on the contrary, if they are felt as a problem, participants indicated if it was no relevant (1), moderate (2), marked (3) or disabling (4). These two scales are used together to classify whether the person suffers from Seasonal Affective Disorder (SAD) or not (NO-SAD). SPAQ-CA criteria (Swedo et al., 1995) for a possible diagnosis of SAD include a GSS of 18 or more with the assessment that these seasonal changes represent at least a moderate problem. The Cronbach Alfa coefficient was 0.75. 2.2.2. Mini Sleep Questionnaire Participants also completed the Italian version (Fabbri et al., 2006) of the MSQ (Zomer et al., 1985) to subjectively evaluate the quality of the sleep/wake cycle. This questionnaire is composed of 10 items on a 7-point Likert scale (1 means always, 7 means never), referring to the past week. MSQ comprises two main factors: sleep and wake. The MSQ sleep factor includes six items, with score ranging from 6 to 42, which investigate the following features: trouble getting asleep, too early morning awakening, use of sleeping pills, snoring during the night, nighttime awakening and unresting sleep. The MSQ wake factor comprises four items, with score ranging from 4 to 28, that cover the following areas: daily sleepiness, feeling tired at Table 1 Actual values (means 7 S.D.) of the sleep and wake MSQ (Mini Sleep Questionnaire) factors by gender and SAD (Seasonal Affective Disorder) category. Higher scores on these two factors correspond to a higher quality of sleep and wake.
MSQ sleep factor Males (N ¼46) Females (N ¼299) Mean value for total sample MSQ Wake factor Males (N ¼46) Females (N ¼299) Mean value for total sample
SAD (N ¼83)
NO-SAD (N ¼262)
Mean value for total sample
31.9 7 6.7 31.0 7 5.8
32.6 7 5.5 33.27 5.5
32.3 7 6.1 32.1 75.7
31.5 7 6.3
32.9 7 5.5
15.7 7 6.3 16.3 7 5.1
19.8 7 4.0 19.2 7 4.6
16.0 7 5.7
19.5 7 4.3
17.8 7 5.2 17.8 7 4.9
625
the morning awakening, headache at the morning awakening, and extended tiredness without a specific reason. Higher scores on these two factors correspond to a higher quality of sleep and wake. Moreover the Italian version of the MSQ includes the following cut-off criteria to discriminate a good sleep quality from a bad sleep quality on one hand and a good wake quality from a bad wake quality on the other: scores higher than 27 indicate a good sleep quality, while scores higher than 14 show a good wake quality. The Cronbach Alfa coefficient was 0.65 for the MSQ sleep factor and 0.75 for the MSQ wake factor. When the Kolmogorov–Smirnov test was performed, the distribution of the MSQ sleep factor scores resulted normal (d ¼0.05; p4 0.05) (kurtosis¼ 0.12; skewness ¼ 0.55), as well as that of the MSQ wake factor scores (d ¼ 0.03; p 40.05) (kurtosis¼ 0.15; skewness¼0 0.39).
2.3. Data analysis We performed an ANCOVA, with gender (males and females) and SAD category (SAD and NO-SAD) as independent variables, and age as a covariate, to separately analyze their effects on the sleep and wake MSQ factors. If ANCOVA gave significant results, Tukey0 s post-hoc test for unequal samples was performed. The results of the ANCOVAs were explored through a set of independent sample t tests aiming to compare the scores reported by SAD and healthy adolescents to the single items of the sleep and wake MSQ factors. Furthermore we performed a multiple regression analysis with the forward stepwise method separately for the sleep and wake MSQ scores, with the following independent variables: gender, age and SAD category. Finally we carried out two separate logistic regression analyses with gender, age and SAD category as independent variables and sleep quality (good and bad) or wake quality (good and bad) categories as dependent variable. Since multiple comparisons were performed, the Bonferroni correction was applied, considering as significant a p value less than 0.003.
3. Results On the basis of SPAQ-CA cut-off criteria, our study included 83 adolescents (of which 73 were females) with possible SAD (24.06%) and 262 (of which 226 were females) with no SAD (75.94%). With reference to the sleep MSQ factor, at the ANCOVA we did not observe any significant effects of gender (p ¼0.87), SAD category (p ¼0.17) and interaction between them (p ¼0.47) (Table 1). As regards the wake MSQ factor, at the ANCOVA, SAD category showed a significant effect (F1,339 ¼15.28; po0.001) (Table 1). Performing post-hoc comparisons, adolescents with SAD scored lower than those not affected (p o0.001). Gender (p ¼0.90) and interaction between the two factors (p ¼0.49) were not significant (Table 1). Table 2 shows the scores reported by SAD and healthy adolescents to the single items of the sleep and wake MSQ factors, compared through a set of independent sample t tests. With reference to the sleep factor, no significant differences were observed, with only one tendency to significance. As regards the wake factor, the scores of three items on four were significantly lower in SAD compared to healthy adolescents.
Table 2 Actual values (means 7 S.D.) of the replies to the items of the MSQ (Mini Sleep Questionnaire) sleep and wake factors, reported by SAD (Seasonal Affective Disorder) and NOSAD adolescents. Statistics of the independent sample t tests are also shown. The Bonferroni correction was applied, considering as significant a p value less than 0.003. Item
SAD N ¼ 83
NO-SAD N ¼262
Statistics
Sleep MSQ factor
Number Number Number Number Number Number
1_ trouble getting asleep 2_ too early morning awakening 3_ use of sleeping pills 6_ snoring during the night 7_ nighttime awakening 10_unresting sleep
4.3 71.8 5.1 71.7 6.6 71.3 6.0 71.6 4.6 71.8 4.5 71.8
4.8 7 1.7 5.4 7 1.6 6.8 7 0.9 6.17 1.4 4.9 7 1.6 5.17 1.7
p ¼0.04 p ¼0.22 p ¼0.19 p ¼0.57 p ¼0.12 p ¼0.004
Wake MSQ factor
Number Number Number Number
4_ 5_ 8_ 9_
3.5 71.5 3.1 71.4 5.2 71.9 4.4 72.0
4.2 7 1.5 3.9 7 1.5 5.7 7 1.6 5.5 7 1.5
t343 ¼ 3.74; po 0.001 t343 ¼ 3.75; p o 0.001 p ¼0.007 t343 ¼ 5.32; p o 0.001
daily sleepiness feeling tired at the morning awakening headache at the morning awakening extended tiredness without a specific reason
626
L. Tonetti et al. / Psychiatry Research 215 (2014) 624–627
Table 3 Contingency tables between the SAD (Seasonal Affective Disorder) category and the category good/bad sleep quality one hand and good/bad wake quality on the other. Percentages and absolute numbers are shown. Statistics of the logistic regression analyses, referred to the SAD category as independent variable, are also shown.
SAD NO-SAD
Good sleep quality
Bad sleep quality
Statistics
Good wake quality
Bad wake quality
Statistics
79.52% (66)
20.48% (17)
B¼ 0.56; E.S.¼ 0.33; Wald¼ 2.90; d.f. ¼ 1; p¼ 0.09
73.49% (61)
26.51% (22)
B¼ 1.11; E.S. ¼0.32; Wald¼ 11.89; d.f. ¼1; p o0.005; exact p-value¼ 0.001
87.02% (228)
12.98% (34)
89.31% (234)
10.69% (28)
The multiple regression analysis showed that gender (p¼ 0.77) and age (p¼ 0.48) were not significant predictors of the sleep MSQ factor, while SAD category (p¼ 0.004) showed a tendency to significance. With regards to the wake MSQ factor, the multiple regression analysis highlighted that SAD category was a significant predictor (Beta ¼ 0.26; t340 ¼ 5.11; p o0.001), while gender (p ¼0.54) did not reach the significant level. Age (p ¼0.003) showed a tendency to significance. The model with one significant predictor accounted for 10% of variance. With reference to the category of good/bad sleep quality, at the logistic regression analysis none of the predictors (gender, p¼0.92; age, p¼ 0.83; SAD category, p¼0.09) reached the significant level. As regards the good/bad wake quality category, performing another logistic regression analysis, SAD category resulted as a significant predictor (po0.005; exact p-value¼ 0.001), contrary to gender (p¼0.45) and age (p¼0.28). The contingency tables (Table 3) showed a higher prevalence of SAD participants with bad wake quality than that of NO-SAD adolescents.
4. Discussion With reference to the gender effect on the perceived quality of the sleep/wake cycle, verified through the ANCOVA, multiple and logistic regression analyses, we did not detect any significant differences between males and females. The present data are in line with previous studies (e.g., Giannotti et al., 2002). Nevertheless we have to bear in mind that the discrepancy between the number of males and females in our sample, could have masked a possible gender difference. The lack of a significant difference in the subjective quality of sleep between SAD and NO-SAD adolescents, observed in the present study using both continuum (ANCOVA and multiple regression analysis) and categorical approaches (logistic regression analysis), is not surprising since a recent book review published by Partonen and Pandi-Perumal (2010) showed no polysomnographic alterations of sleep in SAD patients. On the contrary SAD adolescents presented a significantly worse wake quality, conceived both as a continuum (ANCOVA and multiple regression analysis) and a category (logistic regression analysis), compared to those without this disorder, confirming our expectations. These results mean that SAD students felt tired at the morning awakening, experienced daily sleepiness and extended tiredness without a specific reason (Table 2). Since the quality of sleep is not affected in SAD adolescents, the altered wake cannot be due to a bad sleep quality, but on the contrary it can be considered as a specific feature of SAD. These data are in line with those previously reported by Avery et al. (2002) which showed drowsiness and awakening difficulty in adult patients with SAD, that decreased following the administration of light gradually increasing in intensity before the awakening (i.e., dawn simulation). Furthermore the present results are in agreement with those of previous studies (Thompson and Isaacs, 1988; Allen et al., 1993; Tam et al., 1997; Merikanto et al., 2012) which highlighted that adult SAD patients frequently reported hypersomnia.
The low wake quality in SAD adolescents could be explained on the basis of the phase shift hypothesis (Lewy et al., 1988) that states that SAD symptoms are due to a phase delay of circadian rhythms (i.e., delay of bedtime and get-up time). Since adolescents of the present study get up at a fixed time because they have to go to school to attend classes, it is possible that those affected by SAD sleep less than the needed, resulting more sleepy and tired during the day. Bearing in mind the detrimental effects of sleepiness on scholastic performance (Dewald et al., 2010), future interventions should firstly focus on the screening of adolescents with possible SAD through the administration of the SPAQ-CA, principally developed as a screening tool (Tonetti et al., 2007). Afterwards, therapeutic interventions could help SAD adolescents administering light in the morning, through dawn simulation or for the more severe cases by means of bright light therapy. Finally, some limitations of the present study have to be acknowledged. One major limitation of the present study is the imbalance between the number of males and females. Moreover this work is cross-sectional and thus no causal effects of SAD can be detected on the quality of the sleep/wake cycle. Other major limits are that the present study is based only on self-reported information on seasonal changes and sleep–wake quality and that both SPAQ-CA and MSQ have been administered only once. It would be more informative to assess both mood seasonality and sleep wake quality each season aiming to better understand the influence of mood seasonality on the quality of the sleep/wake cycle. As a consequence, the results of the present study are indications of a possible influence of SAD on wake quality that should be confirmed by further longitudinal studies. In addition, we are aware that the SPAQ-CA diagnostic criteria for SAD may be overinclusive (Tonetti et al., 2007) and thus future studies should confirm the presence of this disorder through a clinical interview.
5. Conclusions To sum up, this is the first study that explored the relationship between SAD and the sleep/wake cycle quality in a sample of adolescents. Despite the above-mentioned limitations, the present work showed that adolescents with a possible diagnosis of SAD had a worse perceived wake quality compared to those not affected, but not a significantly different subjective sleep quality. References Allen, J.M., Lam, R.W., Remick, R.A., Sadovnick, A.D., 1993. Depressive symptoms and family history in seasonal and nonseasonal mood disorders. American Journal of Psychiatry 150, 443–448. Avery, D.H., Kouri, M.E., Monaghan, K., Bolte, M.A., Hellekson, C., Eder, D., 2002. Is dawn simulation effective in ameliorating the difficulty awakening in seasonal affective disorder associated with hypersomnia? Journal of Affective Disorders 69, 231–236. Dewald, J.F., Meijer, A.M., Oort, F.J., Kerkhof, G.A., Bögels, S.M., 2010. The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: a meta-analytic review. Sleep Medicine Reviews 14, 179–189. Fabbri, M., Tonetti, L., Martoni, M., Natale, V., 2006. Uno strumento per la valutazione della qualità del sonno: la versione italiana del Mini-Sleep Questionnaire. Medicina Psicosomatica 51, 37.
L. Tonetti et al. / Psychiatry Research 215 (2014) 624–627
Giannotti, F., Cortesi, F., Sebastiani, T., Ottaviano, S., 2002. Circadian preference, sleep and daytime behaviour in adolescence. Journal of Sleep Research 11, 191–199. Ivanenko, A., McLaughlin Crabtree, V., Gozal, D., 2005. Sleep and depression in children and adolescents. Sleep Medicine Reviews 9, 115–129. Lewy, A.J., Sack, R.L., Singer, C.M., White, D.M., Hoban, T.M., 1988. Winter depression and the phase-shift hypothesis for bright light0 s therapeutic effects: history, theory, and experimental evidence. Journal of Biological Rhythms 3, 121–134. Merikanto, I., Lahti, T., Castaneda, A.E., Tuulio-Henriksson, A., Aalto-Setälä, T., Suvisaari, J., Partonen, T., 2012. Influence of seasonal variation in mood and behavior on cognitive test performance among young adults. Nordic Journal of Psychiatry 66, 303–310. Partonen, T., Pandi-Perumal, S.R., 2010. Seasonal Affective Disorder. Practice and Research, 2nd Ed. Oxford University Press, Oxford. Riemann, D., Berger, M., Voderholzer, U., 2001. Sleep and depression – results from psychobiological studies: an overview. Biological Psychology 57, 67–103. Rosenthal, N.E., Sack, D.A., Gillin, J.C., Lewy, A.J., Goodwin, F.K., Davenport, Y., Mueller, P.S., Newsome, D.A., Wehr, T.A., 1984. Seasonal affective disorder: a description of the syndrome and preliminary findings with light therapy. Archives of General Psychiatry 41, 72–80. Srinivasan, V., Pandi-Perumal, S.R., Trakht, I., Spence, D.W., Hardeland, R., Poeggeler, B., Cardinali, D.P., 2009. Pathophysiology of depression: role of sleep and the melatonergic system. Psychiatry Research 165, 201–214.
627
Swedo, S.E., Pleeter, J.D., Richter, D.M., Hoffman, C.L., Allen, A.J., Hamburger, S.D., Turner, E.H., Yamada, E.M., Rosenthal, N.E., 1995. Rates of seasonal affective disorder in children and adolescents. American Journal of Psychiatry 152, 1016–1019. Tam, E.M., Lam, R.W., Robertson, H.A., Stewart, J.N., Yatham, L.N., Zis, A.P., 1997. Atypical depressive symptoms in seasonal and non-seasonal mood disorders. Journal of Affective Disorders 44, 39–44. Thompson, C., Isaacs, G., 1988. Seasonal affective disorder – a British sample symptomatology in relation to mode of referral and diagnostic subtype. Journal of Affective Disorders 14, 1–12. Tonetti, L., Barbato, G., Fabbri, M., Adan, A., Natale, V., 2007. Mood seasonality: a cross-sectional study of subjects aged between 10 and 25 years. Journal of Affective Disorders 97, 155–160. Tonetti, L., Fabbri, M., D0 Alessandro, I., Barbato, G., Natale, V., 2008. Valutazione della stagionalità in un campione di pre-adolescenti ed adolescenti italiani. Psicologia Clinica dello Sviluppo 22, 529–544. Tsuno, N., Besset, A., Ritchie, K., 2005. Sleep and depression. Journal of Clinical Psychiatry 66, 1254–1269. Wehr, T.A., Sack, D.A., Rosenthal, N.E., 1987. Seasonal affective disorder with summer depression and winter hypomania. American Journal of Psychiatry 144, 1602–1603. Zomer, J., Peder, R., Rubin, A.H., Lavie, P., 1985. Mini Sleep Questionnaire for screening large populations for EDS complaints. In: Koella, W.P., Ruther, E., Schulz, H. (Eds.), Sleep 0 84. Gustav Fischer Verlag, Stuttgart, pp. 467–470.
Contents lists available at ScienceDirect
Psychiatry Research journal homepage: www.elsevier.com/locate/psychres
Association between seasonal affective disorder and subjective quality of the sleep/wake cycle in adolescents Lorenzo Tonetti a,n, Marco Fabbri b, Alex Erbacci a, Monica Martoni c, Vincenzo Natale a a
Department of Psychology, University of Bologna, Viale Berti Pichat 5, 40127 Bologna, Italy Department of Psychology, Second University of Naples, Caserta, Italy c Department of Experimental, Diagnostic and Specialty Medicine, University of Bologna, Bologna, Italy b
art ic l e i nf o
a b s t r a c t
Article history: Received 1 August 2012 Received in revised form 11 December 2013 Accepted 15 December 2013 Available online 19 December 2013
The relationship between seasonal affective disorder (SAD) and subjective quality of sleep/wake cycle in adolescents was explored. The Seasonal Pattern Assessment Questionnaire for Children and Adolescents (SPAQ-CA) and Mini Sleep Questionnaire (MSQ) were administered to 345 adolescents living in the city of Cesena (Emilia-Romagna region, Italy) (299 females; age range: 14–18 years), to determine SAD and perceived quality of the sleep/wake cycle. The response rate was 92% for females and 90.2% for males. The MSQ includes two factors, sleep and wake, with lower scores corresponding to a lower quality of sleep and wake. The MSQ includes cut-off criteria to detect a good or bad sleep and wake quality. Adolescents with SAD (16 7 5.7) scored significantly lower than those not affected on wake factor (19.5 74.3), while no effect has been observed on sleep factor. SAD was the only one significant predictor of good/bad wake quality, while it did not reach significant level with reference to good/bad sleep quality. Present results are indications of a possible influence of SAD on wake quality and further studies are necessary to confirm them. & 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Adolescents Seasonal affective disorder Sleep/wake quality
1. Introduction The relationship between sleep and depression has been deeply explored in the last decades, as reported in three reviews (Riemann et al., 2001; Ivanenko et al., 2005; Tsuno et al., 2005), showing a bi-directional relationship between them. Specifically, some sleep markers of depression are early morning awakening and shorten REM sleep latency (Srinivasan et al., 2009). On the contrary, hypersomnia is related to some sub-types of depression, as the seasonal affective disorder (SAD) (Thompson and Isaacs, 1988; Allen et al., 1993; Tam et al., 1997), which is characterized by recurrent depressive episodes that occur annually in the fall and winter (Rosenthal et al., 1984), with a summer variant of SAD characterized by periodic depressive episodes in this season (Wehr et al., 1987). SAD patients also present awakening difficulties (Avery et al., 2002). It is well known that the prevalence of SAD changes with age. In particular, during adolescence, the frequency of SAD raises compared to childhood (Swedo et al., 1995; Tonetti et al., 2007). This increase has been linked to the strong hormonal changes, occurring during this period (Swedo et al., 1995). To date only two studies (Swedo et al., 1995; Tonetti et al., 2007) have been carried
n
Corresponding author. Tel.: þ 39 051 2091877; fax: þ39 051 243086. E-mail address: [email protected] (L. Tonetti).
0165-1781/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.psychres.2013.12.023
out on the features of SAD during adolescence and the relationship between this disorder and the features of the sleep/wake cycle in this age stage has been neglected. Our study aimed to fill this lack of knowledge, investigating the association between SAD and the perceived quality of the sleep– wake cycle in adolescents. Since adult SAD patients mainly present hypersomnia and awakening difficulties (Thompson and Isaacs, 1988; Allen et al., 1993; Tam et al., 1997; Avery et al., 2002), we could expect a lower wake quality in SAD adolescents compared to those not affected by this disorder. 2. Methods 2.1. Participants Overall 325 females and 51 males were invited to participate. The final sample was composed of 345 participants (299 females), with a response rate of the 92% for females and 90.2% for males. The mean age of the overall sample was 16.88 þ 1.03 (median 17 years and mode 18 years), with age ranging between 14 and 18 years. The males (17.02 7 1.13) and females (16.86 71.02) mean ages were not significantly different (p ¼ 0.33). All participants attended the same high school in the city of Cesena (EmiliaRomagna region, Italy) and were tested during a school meeting in April 2010, aimed to increase students0 knowledge about sleep and well-being. The attending of the school meeting was not mandatory but students received a school credit for their participation. The difference in the number of females and males is due to the fact the investigated high school is mainly attended by females. The time of compilation of both questionnaires was about 15 min. If 18-years-old, participants
L. Tonetti et al. / Psychiatry Research 215 (2014) 624–627 provided written informed consent prior to their participation in the school meeting; if they were underage, the written informed consent was given by parents. The local ethic committee approved the protocol and the study complied with the tenets of the Declaration of Helsinki. 2.2. Measures 2.2.1. Seasonal Pattern Assessment Questionnaire for Children and Adolescents (SPAQ-CA) The Italian version (Tonetti et al., 2008) of the SPAQ-CA (Swedo et al., 1995) which is a self-reporting questionnaire for normal and clinical seasonal mood variations, was administered. Eleven items on a 5-point Likert scale (no change¼ 0, slight change¼1, moderate change¼2, marked change¼3, and extremely marked change¼ 4) evaluate the seasonal changes in the following areas: sleep length, mood, abuse of stimulants (e.g., alcohol, nicotine, and drugs), social activity, weight, energy, appetite, school performance (troubles and marks), irritability and conduct problems. Adding up the scores of those items, we obtain the Global Seasonality Score (GSS), with higher values corresponding to higher mood seasonality (total range score: 0–44). Another question evaluates the extent to which seasonal changes are felt as a problem. If they are not felt as a problem, the corresponding scoring was 0; on the contrary, if they are felt as a problem, participants indicated if it was no relevant (1), moderate (2), marked (3) or disabling (4). These two scales are used together to classify whether the person suffers from Seasonal Affective Disorder (SAD) or not (NO-SAD). SPAQ-CA criteria (Swedo et al., 1995) for a possible diagnosis of SAD include a GSS of 18 or more with the assessment that these seasonal changes represent at least a moderate problem. The Cronbach Alfa coefficient was 0.75. 2.2.2. Mini Sleep Questionnaire Participants also completed the Italian version (Fabbri et al., 2006) of the MSQ (Zomer et al., 1985) to subjectively evaluate the quality of the sleep/wake cycle. This questionnaire is composed of 10 items on a 7-point Likert scale (1 means always, 7 means never), referring to the past week. MSQ comprises two main factors: sleep and wake. The MSQ sleep factor includes six items, with score ranging from 6 to 42, which investigate the following features: trouble getting asleep, too early morning awakening, use of sleeping pills, snoring during the night, nighttime awakening and unresting sleep. The MSQ wake factor comprises four items, with score ranging from 4 to 28, that cover the following areas: daily sleepiness, feeling tired at Table 1 Actual values (means 7 S.D.) of the sleep and wake MSQ (Mini Sleep Questionnaire) factors by gender and SAD (Seasonal Affective Disorder) category. Higher scores on these two factors correspond to a higher quality of sleep and wake.
MSQ sleep factor Males (N ¼46) Females (N ¼299) Mean value for total sample MSQ Wake factor Males (N ¼46) Females (N ¼299) Mean value for total sample
SAD (N ¼83)
NO-SAD (N ¼262)
Mean value for total sample
31.9 7 6.7 31.0 7 5.8
32.6 7 5.5 33.27 5.5
32.3 7 6.1 32.1 75.7
31.5 7 6.3
32.9 7 5.5
15.7 7 6.3 16.3 7 5.1
19.8 7 4.0 19.2 7 4.6
16.0 7 5.7
19.5 7 4.3
17.8 7 5.2 17.8 7 4.9
625
the morning awakening, headache at the morning awakening, and extended tiredness without a specific reason. Higher scores on these two factors correspond to a higher quality of sleep and wake. Moreover the Italian version of the MSQ includes the following cut-off criteria to discriminate a good sleep quality from a bad sleep quality on one hand and a good wake quality from a bad wake quality on the other: scores higher than 27 indicate a good sleep quality, while scores higher than 14 show a good wake quality. The Cronbach Alfa coefficient was 0.65 for the MSQ sleep factor and 0.75 for the MSQ wake factor. When the Kolmogorov–Smirnov test was performed, the distribution of the MSQ sleep factor scores resulted normal (d ¼0.05; p4 0.05) (kurtosis¼ 0.12; skewness ¼ 0.55), as well as that of the MSQ wake factor scores (d ¼ 0.03; p 40.05) (kurtosis¼ 0.15; skewness¼0 0.39).
2.3. Data analysis We performed an ANCOVA, with gender (males and females) and SAD category (SAD and NO-SAD) as independent variables, and age as a covariate, to separately analyze their effects on the sleep and wake MSQ factors. If ANCOVA gave significant results, Tukey0 s post-hoc test for unequal samples was performed. The results of the ANCOVAs were explored through a set of independent sample t tests aiming to compare the scores reported by SAD and healthy adolescents to the single items of the sleep and wake MSQ factors. Furthermore we performed a multiple regression analysis with the forward stepwise method separately for the sleep and wake MSQ scores, with the following independent variables: gender, age and SAD category. Finally we carried out two separate logistic regression analyses with gender, age and SAD category as independent variables and sleep quality (good and bad) or wake quality (good and bad) categories as dependent variable. Since multiple comparisons were performed, the Bonferroni correction was applied, considering as significant a p value less than 0.003.
3. Results On the basis of SPAQ-CA cut-off criteria, our study included 83 adolescents (of which 73 were females) with possible SAD (24.06%) and 262 (of which 226 were females) with no SAD (75.94%). With reference to the sleep MSQ factor, at the ANCOVA we did not observe any significant effects of gender (p ¼0.87), SAD category (p ¼0.17) and interaction between them (p ¼0.47) (Table 1). As regards the wake MSQ factor, at the ANCOVA, SAD category showed a significant effect (F1,339 ¼15.28; po0.001) (Table 1). Performing post-hoc comparisons, adolescents with SAD scored lower than those not affected (p o0.001). Gender (p ¼0.90) and interaction between the two factors (p ¼0.49) were not significant (Table 1). Table 2 shows the scores reported by SAD and healthy adolescents to the single items of the sleep and wake MSQ factors, compared through a set of independent sample t tests. With reference to the sleep factor, no significant differences were observed, with only one tendency to significance. As regards the wake factor, the scores of three items on four were significantly lower in SAD compared to healthy adolescents.
Table 2 Actual values (means 7 S.D.) of the replies to the items of the MSQ (Mini Sleep Questionnaire) sleep and wake factors, reported by SAD (Seasonal Affective Disorder) and NOSAD adolescents. Statistics of the independent sample t tests are also shown. The Bonferroni correction was applied, considering as significant a p value less than 0.003. Item
SAD N ¼ 83
NO-SAD N ¼262
Statistics
Sleep MSQ factor
Number Number Number Number Number Number
1_ trouble getting asleep 2_ too early morning awakening 3_ use of sleeping pills 6_ snoring during the night 7_ nighttime awakening 10_unresting sleep
4.3 71.8 5.1 71.7 6.6 71.3 6.0 71.6 4.6 71.8 4.5 71.8
4.8 7 1.7 5.4 7 1.6 6.8 7 0.9 6.17 1.4 4.9 7 1.6 5.17 1.7
p ¼0.04 p ¼0.22 p ¼0.19 p ¼0.57 p ¼0.12 p ¼0.004
Wake MSQ factor
Number Number Number Number
4_ 5_ 8_ 9_
3.5 71.5 3.1 71.4 5.2 71.9 4.4 72.0
4.2 7 1.5 3.9 7 1.5 5.7 7 1.6 5.5 7 1.5
t343 ¼ 3.74; po 0.001 t343 ¼ 3.75; p o 0.001 p ¼0.007 t343 ¼ 5.32; p o 0.001
daily sleepiness feeling tired at the morning awakening headache at the morning awakening extended tiredness without a specific reason
626
L. Tonetti et al. / Psychiatry Research 215 (2014) 624–627
Table 3 Contingency tables between the SAD (Seasonal Affective Disorder) category and the category good/bad sleep quality one hand and good/bad wake quality on the other. Percentages and absolute numbers are shown. Statistics of the logistic regression analyses, referred to the SAD category as independent variable, are also shown.
SAD NO-SAD
Good sleep quality
Bad sleep quality
Statistics
Good wake quality
Bad wake quality
Statistics
79.52% (66)
20.48% (17)
B¼ 0.56; E.S.¼ 0.33; Wald¼ 2.90; d.f. ¼ 1; p¼ 0.09
73.49% (61)
26.51% (22)
B¼ 1.11; E.S. ¼0.32; Wald¼ 11.89; d.f. ¼1; p o0.005; exact p-value¼ 0.001
87.02% (228)
12.98% (34)
89.31% (234)
10.69% (28)
The multiple regression analysis showed that gender (p¼ 0.77) and age (p¼ 0.48) were not significant predictors of the sleep MSQ factor, while SAD category (p¼ 0.004) showed a tendency to significance. With regards to the wake MSQ factor, the multiple regression analysis highlighted that SAD category was a significant predictor (Beta ¼ 0.26; t340 ¼ 5.11; p o0.001), while gender (p ¼0.54) did not reach the significant level. Age (p ¼0.003) showed a tendency to significance. The model with one significant predictor accounted for 10% of variance. With reference to the category of good/bad sleep quality, at the logistic regression analysis none of the predictors (gender, p¼0.92; age, p¼ 0.83; SAD category, p¼0.09) reached the significant level. As regards the good/bad wake quality category, performing another logistic regression analysis, SAD category resulted as a significant predictor (po0.005; exact p-value¼ 0.001), contrary to gender (p¼0.45) and age (p¼0.28). The contingency tables (Table 3) showed a higher prevalence of SAD participants with bad wake quality than that of NO-SAD adolescents.
4. Discussion With reference to the gender effect on the perceived quality of the sleep/wake cycle, verified through the ANCOVA, multiple and logistic regression analyses, we did not detect any significant differences between males and females. The present data are in line with previous studies (e.g., Giannotti et al., 2002). Nevertheless we have to bear in mind that the discrepancy between the number of males and females in our sample, could have masked a possible gender difference. The lack of a significant difference in the subjective quality of sleep between SAD and NO-SAD adolescents, observed in the present study using both continuum (ANCOVA and multiple regression analysis) and categorical approaches (logistic regression analysis), is not surprising since a recent book review published by Partonen and Pandi-Perumal (2010) showed no polysomnographic alterations of sleep in SAD patients. On the contrary SAD adolescents presented a significantly worse wake quality, conceived both as a continuum (ANCOVA and multiple regression analysis) and a category (logistic regression analysis), compared to those without this disorder, confirming our expectations. These results mean that SAD students felt tired at the morning awakening, experienced daily sleepiness and extended tiredness without a specific reason (Table 2). Since the quality of sleep is not affected in SAD adolescents, the altered wake cannot be due to a bad sleep quality, but on the contrary it can be considered as a specific feature of SAD. These data are in line with those previously reported by Avery et al. (2002) which showed drowsiness and awakening difficulty in adult patients with SAD, that decreased following the administration of light gradually increasing in intensity before the awakening (i.e., dawn simulation). Furthermore the present results are in agreement with those of previous studies (Thompson and Isaacs, 1988; Allen et al., 1993; Tam et al., 1997; Merikanto et al., 2012) which highlighted that adult SAD patients frequently reported hypersomnia.
The low wake quality in SAD adolescents could be explained on the basis of the phase shift hypothesis (Lewy et al., 1988) that states that SAD symptoms are due to a phase delay of circadian rhythms (i.e., delay of bedtime and get-up time). Since adolescents of the present study get up at a fixed time because they have to go to school to attend classes, it is possible that those affected by SAD sleep less than the needed, resulting more sleepy and tired during the day. Bearing in mind the detrimental effects of sleepiness on scholastic performance (Dewald et al., 2010), future interventions should firstly focus on the screening of adolescents with possible SAD through the administration of the SPAQ-CA, principally developed as a screening tool (Tonetti et al., 2007). Afterwards, therapeutic interventions could help SAD adolescents administering light in the morning, through dawn simulation or for the more severe cases by means of bright light therapy. Finally, some limitations of the present study have to be acknowledged. One major limitation of the present study is the imbalance between the number of males and females. Moreover this work is cross-sectional and thus no causal effects of SAD can be detected on the quality of the sleep/wake cycle. Other major limits are that the present study is based only on self-reported information on seasonal changes and sleep–wake quality and that both SPAQ-CA and MSQ have been administered only once. It would be more informative to assess both mood seasonality and sleep wake quality each season aiming to better understand the influence of mood seasonality on the quality of the sleep/wake cycle. As a consequence, the results of the present study are indications of a possible influence of SAD on wake quality that should be confirmed by further longitudinal studies. In addition, we are aware that the SPAQ-CA diagnostic criteria for SAD may be overinclusive (Tonetti et al., 2007) and thus future studies should confirm the presence of this disorder through a clinical interview.
5. Conclusions To sum up, this is the first study that explored the relationship between SAD and the sleep/wake cycle quality in a sample of adolescents. Despite the above-mentioned limitations, the present work showed that adolescents with a possible diagnosis of SAD had a worse perceived wake quality compared to those not affected, but not a significantly different subjective sleep quality. References Allen, J.M., Lam, R.W., Remick, R.A., Sadovnick, A.D., 1993. Depressive symptoms and family history in seasonal and nonseasonal mood disorders. American Journal of Psychiatry 150, 443–448. Avery, D.H., Kouri, M.E., Monaghan, K., Bolte, M.A., Hellekson, C., Eder, D., 2002. Is dawn simulation effective in ameliorating the difficulty awakening in seasonal affective disorder associated with hypersomnia? Journal of Affective Disorders 69, 231–236. Dewald, J.F., Meijer, A.M., Oort, F.J., Kerkhof, G.A., Bögels, S.M., 2010. The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: a meta-analytic review. Sleep Medicine Reviews 14, 179–189. Fabbri, M., Tonetti, L., Martoni, M., Natale, V., 2006. Uno strumento per la valutazione della qualità del sonno: la versione italiana del Mini-Sleep Questionnaire. Medicina Psicosomatica 51, 37.
L. Tonetti et al. / Psychiatry Research 215 (2014) 624–627
Giannotti, F., Cortesi, F., Sebastiani, T., Ottaviano, S., 2002. Circadian preference, sleep and daytime behaviour in adolescence. Journal of Sleep Research 11, 191–199. Ivanenko, A., McLaughlin Crabtree, V., Gozal, D., 2005. Sleep and depression in children and adolescents. Sleep Medicine Reviews 9, 115–129. Lewy, A.J., Sack, R.L., Singer, C.M., White, D.M., Hoban, T.M., 1988. Winter depression and the phase-shift hypothesis for bright light0 s therapeutic effects: history, theory, and experimental evidence. Journal of Biological Rhythms 3, 121–134. Merikanto, I., Lahti, T., Castaneda, A.E., Tuulio-Henriksson, A., Aalto-Setälä, T., Suvisaari, J., Partonen, T., 2012. Influence of seasonal variation in mood and behavior on cognitive test performance among young adults. Nordic Journal of Psychiatry 66, 303–310. Partonen, T., Pandi-Perumal, S.R., 2010. Seasonal Affective Disorder. Practice and Research, 2nd Ed. Oxford University Press, Oxford. Riemann, D., Berger, M., Voderholzer, U., 2001. Sleep and depression – results from psychobiological studies: an overview. Biological Psychology 57, 67–103. Rosenthal, N.E., Sack, D.A., Gillin, J.C., Lewy, A.J., Goodwin, F.K., Davenport, Y., Mueller, P.S., Newsome, D.A., Wehr, T.A., 1984. Seasonal affective disorder: a description of the syndrome and preliminary findings with light therapy. Archives of General Psychiatry 41, 72–80. Srinivasan, V., Pandi-Perumal, S.R., Trakht, I., Spence, D.W., Hardeland, R., Poeggeler, B., Cardinali, D.P., 2009. Pathophysiology of depression: role of sleep and the melatonergic system. Psychiatry Research 165, 201–214.
627
Swedo, S.E., Pleeter, J.D., Richter, D.M., Hoffman, C.L., Allen, A.J., Hamburger, S.D., Turner, E.H., Yamada, E.M., Rosenthal, N.E., 1995. Rates of seasonal affective disorder in children and adolescents. American Journal of Psychiatry 152, 1016–1019. Tam, E.M., Lam, R.W., Robertson, H.A., Stewart, J.N., Yatham, L.N., Zis, A.P., 1997. Atypical depressive symptoms in seasonal and non-seasonal mood disorders. Journal of Affective Disorders 44, 39–44. Thompson, C., Isaacs, G., 1988. Seasonal affective disorder – a British sample symptomatology in relation to mode of referral and diagnostic subtype. Journal of Affective Disorders 14, 1–12. Tonetti, L., Barbato, G., Fabbri, M., Adan, A., Natale, V., 2007. Mood seasonality: a cross-sectional study of subjects aged between 10 and 25 years. Journal of Affective Disorders 97, 155–160. Tonetti, L., Fabbri, M., D0 Alessandro, I., Barbato, G., Natale, V., 2008. Valutazione della stagionalità in un campione di pre-adolescenti ed adolescenti italiani. Psicologia Clinica dello Sviluppo 22, 529–544. Tsuno, N., Besset, A., Ritchie, K., 2005. Sleep and depression. Journal of Clinical Psychiatry 66, 1254–1269. Wehr, T.A., Sack, D.A., Rosenthal, N.E., 1987. Seasonal affective disorder with summer depression and winter hypomania. American Journal of Psychiatry 144, 1602–1603. Zomer, J., Peder, R., Rubin, A.H., Lavie, P., 1985. Mini Sleep Questionnaire for screening large populations for EDS complaints. In: Koella, W.P., Ruther, E., Schulz, H. (Eds.), Sleep 0 84. Gustav Fischer Verlag, Stuttgart, pp. 467–470.
