2024-10-03

089_ Do atmospheric variables affect psychotic symptoms in patients within the psychotic to affective spectrum?

Research Question and Aims

One of the most significant challenges of the 21st century is the impact of climate change and the intensifying fluctuations in atmospheric variables (Romanello et al. 2023). Mental health is highly relevant in the context of climate change and is recently becoming more recognized in the discourse (Lawrence et al. 2022). Mental illnesses like schizophrenia and bipolar disorder and their hospitalization are correlated with changes in atmospheric variables (e.g., temperature, air pressure, cloud cover/cloudiness, precipitation, humidity, solar radiation, air quality) (Asimakopoulos et al. 2021, Christensen et al 2008, Hu et al. 2023, Tupinier Martin et al. 2024). Higher hospitalization rates suggest an exacerbation of the disease and is often caused by an increase of psychotic symptoms and irritability (Walinski et al. 2023), but analyses are still limited and do not disclose detailed information about disorder specific symptoms. Also the mental wellbeing of individuals without diagnosed major psychiatric disorders can be affected by weather conditions and anomalies (Koppe, Zacharias, and Bernhard 2013).
The main research question is: Are weather conditions and anomalies associated with higher positive psychotic symptoms and irritability across individuals within the affective to psychotic spectrum and healthy controls. And, is this association mediated by the polygenic risk score for schizophrenia (PRS) and sex.
H1a: Weather anomalies and fluctuating weather conditions (temperature, heat waves, cold spells, wet bulb globe temperature (WBGT), precipitation, wind, air pressure, radiation, air quality index) influence the severity of positive psychotic symptoms (Positive and Negative Symptom Scale (PANSS) positive score) in individuals of the affective-to-psychotic spectrum and healthy controls. . H1b: Weather anomalies and fluctuating weather conditions (temperature, heat waves, cold spells, wet bulb globe temperature (WBGT), precipitation, wind, air pressure, radiation, air quality index) influence the grade of irritability (The Inventory of Depressive Symptomatology (IDS-C30): item 6 and Young Mania Rating Scale (YMRS): item 5) in individuals of the affective-to-psychotic spectrum and healthy controls.
H2a: The effects of weather conditions on positive psychotic symptoms (PANSS positive score) are moderated by the PRS for schizophrenia.
H2b: The effects of weather conditions on irritability (IDS-C30: item 6 and YMRS: item 5) are moderated by the PRS for schizophrenia.
H3a: Changes of the atmospheric variables have a higher impact on positive psychotic symptoms (PANSS positive score) in females than in males.
H3b: Changes of the atmospheric variables have a higher impact on irritability (IDS-C30: item 6 and YMRS: item 5) in females than in males.

Analytic Plan

Data:
The primary data source of the atmospheric data will be retained from the gridded ECMWF ERA5 and CAMS reanalyses. Additionally, station data from the German Weather Service as well as from the provincial environmental agencies are available. Besides using the raw atmospheric variables, human biometeorological indices like the WBGT will be calculated as well as air quality indices like the European Air Quality Index.
The phenotypic and genetic data will be used from the PsyCourse study.
Dependent variable: PANSS positive score, irritability (IDS-C30: item 6 and YMRS: item 5)
Independent variables: atmospheric variables (temperature-related, air pressure, precipitation, hours of sunshine, radiation, wind, air quality index)
Modification/Covariate variables: age, sex, PRS for schizophrenia, ancestry principal components, visit number, clinical/control status, diagnostic group, Global assessment of functioning (GAF)

Methods:
For hypothesis 1-3 first exploratory analyses will be conducted. Based on basic statistical methods such as the Spearman correlation to compile a correlation matrix in order to analyse if there are correlations between the predictors themselves and between the predictors and the dependent variables.
Generalized linear (GLM) and generalized additive (GAM) regression models with, depending on the data structure, a normal distribution or (overdispersed) Poisson distribution will be implemented for each hypothesis. Starting with baseline models using only each single atmospheric variable as independent variable, gradually more complex models will be generated which include different atmospheric variables as independent variables as well as age, sex, PRS for schizophrenia, ancestry principal components, clinical/control status, diagnostic group, and GAF as modifiers and/or confounder. Each patient visit is regarded as an independent case. However, the four study visits of an individual in the overall 18-months study period, will be included as a potential modifier. Within the exposure-outcome relationships the influence of two specific time periods will be considered- the direct impact of weather represented as the mean weather and its anomalies in the 1-week period matching the rating scales’ horizons as well as the longerterm weather conditions represented by the conditions in the preceding four weeks to the study visit. For each individual the atmospheric information will be retrieved for the area of the individual’s study centre location.

Resources needed

v1_id
v1_stat
v1_center
v1_interv_date
v1_sex
v1_age
v1_scid_dsm_dx
v1_scid_dsm_dx_cat
v1_panss_p1
v1_panss_p2
v1_panss_p3
v1_panss_p4
v1_panss_p5
v1_panss_p6
v1_panss_p7
v1_panss_n1
v1_panss_n2
v1_panss_n3
v1_panss_n4
v1_panss_n5
v1_panss_n6
v1_panss_n7
v1_panss_g1
v1_panss_g2
v1_panss_g3
v1_panss_g4
v1_panss_g5
v1_panss_g6
v1_panss_g7
v1_panss_g8
v1_panss_g9
v1_panss_g10
v1_panss_g11
v1_panss_g12
v1_panss_g13
v1_panss_g14
v1_panss_g15
v1_panss_g16
v1_panss_sum_pos
v1_panss_sum_neg
v1_panss_sum_gen
v1_panss_sum_tot
v1_idsc_itm1
v1_idsc_itm2
v1_idsc_itm3
v1_idsc_itm4
v1_idsc_itm5
v1_idsc_itm6
v1_idsc_itm7
v1_idsc_itm8
v1_idsc_itm9
v1_idsc_itm9a
v1_idsc_itm9b
v1_idsc_itm10
v1_idsc_itm11
v1_idsc_itm12
v1_idsc_itm13
v1_idsc_itm14
v1_idsc_itm15
v1_idsc_itm16
v1_idsc_itm17
v1_idsc_itm18
v1_idsc_itm19
v1_idsc_itm20
v1_idsc_itm21
v1_idsc_itm22
v1_idsc_itm23
v1_idsc_itm24
v1_idsc_itm25
v1_idsc_itm26
v1_idsc_itm27
v1_idsc_itm28
v1_idsc_itm29
v1_idsc_itm30
v1_idsc_sum
v1_ymrs_itm1
v1_ymrs_itm2
v1_ymrs_itm3
v1_ymrs_itm4
v1_ymrs_itm5
v1_ymrs_itm6
v1_ymrs_itm7
v1_ymrs_itm8
v1_ymrs_itm9
v1_ymrs_itm10
v1_ymrs_itm11
v1_ymrs_sum
v1_gaf
v2_age
v2_interv_date
v2_panss_p1
v2_panss_p2
v2_panss_p3
v2_panss_p4
v2_panss_p5
v2_panss_p6
v2_panss_p7
v2_panss_n1
v2_panss_n2
v2_panss_n3
v2_panss_n4
v2_panss_n5
v2_panss_n6
v2_panss_n7
v2_panss_g1
v2_panss_g2
v2_panss_g3
v2_panss_g4
v2_panss_g5
v2_panss_g6
v2_panss_g7
v2_panss_g8
v2_panss_g9
v2_panss_g10
v2_panss_g11
v2_panss_g12
v2_panss_g13
v2_panss_g14
v2_panss_g15
v2_panss_g16
v2_panss_sum_pos
v2_panss_sum_neg
v2_panss_sum_gen
v2_panss_sum_tot
v2_idsc_itm1
v2_idsc_itm2
v2_idsc_itm3
v2_idsc_itm4
v2_idsc_itm5
v2_idsc_itm6
v2_idsc_itm7
v2_idsc_itm8
v2_idsc_itm9
v2_idsc_itm9a
v2_idsc_itm9b
v2_idsc_itm10
v2_idsc_itm11
v2_idsc_itm12
v2_idsc_itm13
v2_idsc_itm14
v2_idsc_itm15
v2_idsc_itm16
v2_idsc_itm17
v2_idsc_itm18
v2_idsc_itm19
v2_idsc_itm20
v2_idsc_itm21
v2_idsc_itm22
v2_idsc_itm23
v2_idsc_itm24
v2_idsc_itm25
v2_idsc_itm26
v2_idsc_itm27
v2_idsc_itm28
v2_idsc_itm29
v2_idsc_itm30
v2_idsc_sum
v2_ymrs_itm1
v2_ymrs_itm2
v2_ymrs_itm3
v2_ymrs_itm4
v2_ymrs_itm5
v2_ymrs_itm6
v2_ymrs_itm7
v2_ymrs_itm8
v2_ymrs_itm9
v2_ymrs_itm10
v2_ymrs_itm11
v2_ymrs_sum
v2_gaf
v3_age
v3_interv_date
v3_panss_p1
v3_panss_p2
v3_panss_p3
v3_panss_p4
v3_panss_p5
v3_panss_p6
v3_panss_p7
v3_panss_n1
v3_panss_n2
v3_panss_n3
v3_panss_n4
v3_panss_n5
v3_panss_n6
v3_panss_n7
v3_panss_g1
v3_panss_g2
v3_panss_g3
v3_panss_g4
v3_panss_g5
v3_panss_g6
v3_panss_g7
v3_panss_g8
v3_panss_g9
v3_panss_g10
v3_panss_g11
v3_panss_g12
v3_panss_g13
v3_panss_g14
v3_panss_g15
v3_panss_g16
v3_panss_sum_pos
v3_panss_sum_neg
v3_panss_sum_gen
v3_panss_sum_tot
v3_idsc_itm1
v3_idsc_itm2
v3_idsc_itm3
v3_idsc_itm4
v3_idsc_itm5
v3_idsc_itm6
v3_idsc_itm7
v3_idsc_itm8
v3_idsc_itm9
v3_idsc_itm9a
v3_idsc_itm9b
v3_idsc_itm10
v3_idsc_itm11
v3_idsc_itm12
v3_idsc_itm13
v3_idsc_itm14
v3_idsc_itm15
v3_idsc_itm16
v3_idsc_itm17
v3_idsc_itm18
v3_idsc_itm19
v3_idsc_itm20
v3_idsc_itm21
v3_idsc_itm22
v3_idsc_itm23
v3_idsc_itm24
v3_idsc_itm25
v3_idsc_itm26
v3_idsc_itm27
v3_idsc_itm28
v3_idsc_itm29
v3_idsc_itm30
v3_idsc_sum
v3_ymrs_itm1
v3_ymrs_itm2
v3_ymrs_itm3
v3_ymrs_itm4
v3_ymrs_itm5
v3_ymrs_itm6
v3_ymrs_itm7
v3_ymrs_itm8
v3_ymrs_itm9
v3_ymrs_itm10
v3_ymrs_itm11
v3_ymrs_sum
v3_gaf
v4_age
v4_interv_date
v4_panss_p1
v4_panss_p2
v4_panss_p3
v4_panss_p4
v4_panss_p5
v4_panss_p6
v4_panss_p7
v4_panss_n1
v4_panss_n2
v4_panss_n3
v4_panss_n4
v4_panss_n5
v4_panss_n6
v4_panss_n7
v4_panss_g1
v4_panss_g2
v4_panss_g3
v4_panss_g4
v4_panss_g5
v4_panss_g6
v4_panss_g7
v4_panss_g8
v4_panss_g9
v4_panss_g10
v4_panss_g11
v4_panss_g12
v4_panss_g13
v4_panss_g14
v4_panss_g15
v4_panss_g16
v4_panss_sum_pos
v4_panss_sum_neg
v4_panss_sum_gen
v4_panss_sum_tot
v4_idsc_itm1
v4_idsc_itm2
v4_idsc_itm3
v4_idsc_itm4
v4_idsc_itm5
v4_idsc_itm6
v4_idsc_itm7
v4_idsc_itm8
v4_idsc_itm9
v4_idsc_itm9a
v4_idsc_itm9b
v4_idsc_itm10
v4_idsc_itm11
v4_idsc_itm12
v4_idsc_itm13
v4_idsc_itm14
v4_idsc_itm15
v4_idsc_itm16
v4_idsc_itm17
v4_idsc_itm18
v4_idsc_itm19
v4_idsc_itm20
v4_idsc_itm21
v4_idsc_itm22
v4_idsc_itm23
v4_idsc_itm24
v4_idsc_itm25
v4_idsc_itm26
v4_idsc_itm27
v4_idsc_itm28
v4_idsc_itm29
v4_idsc_itm30
v4_idsc_sum
v4_ymrs_itm1
v4_ymrs_itm2
v4_ymrs_itm3
v4_ymrs_itm4
v4_ymrs_itm5
v4_ymrs_itm6
v4_ymrs_itm7
v4_ymrs_itm8
v4_ymrs_itm9
v4_ymrs_itm10
v4_ymrs_itm11
v4_ymrs_sum
v4_gaf
gsa_id