Evaluation of seasonal variations for the seasonal

pattern assessment in mood disorder patients and

healthy controls

1,4

Chiao-Erh Chang ,ꢀCheng-Dien Hsu ,ꢀHis-ChungꢀChen ,ꢀWen-YinꢀChen ,ꢀChih-MingꢀLiu ,ꢀ

5,6

1,3

Ming-HsienꢀHsieh ,ꢀChun-Hsin Chen ,ꢀTsung-YangꢀWang ,ꢀandꢀPo-HsiuꢀKuo

DepartmentꢀofꢀPublicꢀHealthꢀ&ꢀInstituteꢀofꢀEpidemiologyꢀandꢀPreventiveꢀMedicine,ꢀCollegeꢀofꢀPublicꢀHealth,ꢀNationalꢀTaiwanꢀ

University,ꢀTaipei,ꢀTaiwan, DepartmentꢀofꢀPsychiatry,ꢀTaiwanꢀAdventistꢀHospital,ꢀTaipei,ꢀTaiwan, DepartmentꢀofꢀPsychiatry,ꢀNationalꢀ

TaiwanꢀUniversityꢀHospital,ꢀTaipei,ꢀTaiwan,ꢀ DepartmentꢀofꢀPsychiatry,ꢀTaipeiꢀCityꢀPsychiatricꢀCenter,ꢀTaipeiꢀCityꢀHospital,ꢀTaipei,ꢀ

Taiwan,ꢀ DepartmentꢀofꢀPsychiatry,ꢀWang-FangꢀHospital,ꢀTaipeiꢀMedicalꢀUniversity,ꢀTaipei,ꢀTaiwan,ꢀ DepartmentꢀofꢀPsychiatry,ꢀSchoolꢀ

ofꢀMedicine,ꢀCollegeꢀofꢀMedicine,ꢀTaipeiꢀMedicalꢀUniversity,ꢀTaipei,ꢀTaiwan

Introduction

The seasonal pattern of mood episodes in patients with mood disorders has long been discussed. The Seasonal Pattern Assessment

Questionnaire (SPAQ) was most used to assess seasonality in mood and related symptoms. The global severity of seasonality (GSS) score

derived from SPAQ contains six dimensions: sleep, activity, mood, weight, appetite, and energy level. It is unclear whether the season in

which SPAQ was administered impacts the magnitude of GSS score. The present study first established the structure of the Chinese

version of SPAQ and further examined the reliability of self-reported seasonality among participants in the four seasons.

Methods

We recruited 597 mood disorder (MD) patients and 138 healthy controls (HC). 121 MD patients and 37 HC had at least twice follow-up

measurements among them. A GSS score ≥ 11 and MD patients who suffered from greater than moderate levels of disturbances due to

seasonal fluctuations were defined as the seasonal affective disorder (SAD). Spearman correlation coefficient was calculated between GSS

score and mood dimensions. We performed statistical analysis using intraclass correlation coefficient reliability (ICCR) to evaluate the

reliability of mood dimension across four seasons. All statistical analyses were conducted using the SAS version 9.4.

Results

MD patients were older (44.11 v. 37.84 years old, p < 0.001) and exhibited more severe seasonality (GSS score: 7.59 v. 3.79, p < 0.001)

than the HC group (Table 1). A quarter of MD patients met the criteria of SAD, and 13.74% filled for S-SAD. In terms of multiple

measurements, the average repeat measurement was 3.66 and 3.59 for MD patients and the HC group, respectively.

Table1.ꢀDemographicꢀandꢀclinicalꢀcharacteristicsꢀofꢀMDꢀpatientsꢀandꢀHC.

Moodꢀdisorderꢀpatientsꢀ Healthyꢀcontrols

n=597)^a(n=138)^b

multiple

(n=121)

b.ꢀHealthyꢀcontrolsꢀ

a.ꢀMoodꢀdisorderꢀpatientsꢀ

(

Variables

once

n=476)

once

(n=101)

multiple

(n=37)

(

43.76ꢀ(13.74) 45.47ꢀ(13.40) 37.65ꢀ(13.21) 38.38ꢀ(13.13)

Age

329ꢀ(69.12)

78ꢀ(64.46)

69ꢀ(68.32)

23ꢀ(62.16)

Femaleꢀgender

Diagnosis

29ꢀ(49.11)

47ꢀ(51.89)

70ꢀ(57.85)

51ꢀ(42.15)

Bipolarꢀdisorder

Majorꢀdepressiveꢀdisorder

Averageꢀtimesꢀofꢀrepeatꢀ

measurement

3.66ꢀ(1.03)

3.59ꢀ(1.07)

.51ꢀ(5.63)

7.93ꢀ(5.04)

44ꢀ(36.36)

22ꢀ(18.18)

3.66ꢀ(3.20)

4.14ꢀ(3.25)

GSSꢀscore

SAD^c

Subsyndromal-SAD^d

107ꢀ(22.48)

0ꢀ(12.61)

GSS, Global severity of seasonality; SAD, Seasonal affective disorder.

A total of 919 records of 597 mood disorder patients.

A total of 234 records of 138 healthy participants.1

Patients with multiple seasonal pattern assessments met the SAD criteria at least once.

Patients with multiple seasonal pattern assessments met the Subsyndromal-SAD criteria at least once.

Figureꢀ1.ꢀTheꢀcorrelationꢀbetweenꢀglobalꢀseasonalityꢀscoreꢀ(GSS),ꢀsixꢀmoodꢀdimensions,ꢀandꢀtheꢀ

degreeꢀofꢀseasonalityꢀproblems.

The energy level (r = 0.84), mood (r = 0.79), and social activity (r = 0.76) were the top three significantly correlated with the GSS scores

among MD patients (Figure 1). In the aspect of the HC group, the energy level and sleep were significantly correlated with the GSS score

(r > 0.75). We found good reliability (ICCR > 0.40) of the GSS score and mood and energy level dimensions in MD patients across the four

seasons (Table 2). GSS dimensions among the HC group displayed good reliability (ICCR > 0.4), expected for social activity and appetite.

Figure 2 illustrates the monthly distribution of MD patients and HC group who declared “feeling worse,” in distinct seasons.

Tableꢀ2.ꢀEvaluationꢀofꢀreliabilityꢀforꢀself-assessedꢀ

seasonalityꢀinꢀmoodꢀdimensions,ꢀseasonalityꢀ

Conclusion

severity,ꢀandꢀspecificꢀseasonꢀpatterns.ꢀ

Our results revealed the seasonal

MDꢀ(n=121) HCꢀ(n=37)

variations in six mood dimensions of

ICCR

seasonal pattern assessment, whereas

the GSS score was relatively stable in

both MD patients and HC.

Sixꢀmoodꢀdimensionsꢀꢀ

Sleep

Socialꢀactivity

Mood

Weight

Appetite

.356

.278

.429

.331

.397

.464

0.428

0.164

0.418

0.437

0.262

0.406

Acknowledgements

Energyꢀlevel

Seasonality evaluation

GSSꢀscore

This research was supported by grants from

the Ministry of Science and Technology (MST

105-2628-B-002-028-MY3), National Health

Research Institutes(NHRI-EX107-10627NI),

and National Taiwan University (108L7860).

.521

0.449

Specificꢀseasonalꢀpatternsꢀinꢀmoodꢀsymptoms

0.316

Depressiveꢀsymptoms

Manicꢀsymptoms

ICCR, Intraclass correlation coefficient reliability; GSS, Global

severity of seasonality.

0.437

Figureꢀ2.ꢀMonthlyꢀprevalenceꢀofꢀmoodꢀsymptomsꢀforꢀfourꢀseasonꢀfollow-upꢀofꢀMDꢀ

patientꢀandꢀhealthyꢀcontrol.