**1. Introduction**

As defined by the Word Health Organization in its lexicon of alcohol and drug terms, Dual Disorders are defined as the comorbidity of at least one Substance Use Disorder (SUD) and one Severe Mental Illness (SMI) in the same person [1], being the most frequent psychotic, bipolar and depressive spectrum disorders [2,3]. Given the heterogeneous nature, high prevalence and clinical and functional implications of dual disorders, in recent years interest in its study has increased with the aim of improving both the detection and the therapeutic approach [2–4].

Different studies have shown that addictive behavior has negative effects on circadian rhythmic expression [5,6] which can persist weeks or months after starting substance

**Citation:** Serrano-Serrano, A.B.; Marquez-Arrico, J.E.; Navarro, J.F.; Martinez-Nicolas, A.; Adan, A. Circadian Characteristics in Patients under Treatment for Substance Use Disorders and Severe Mental Illness (Schizophrenia, Major Depression and Bipolar Disorder). *J. Clin. Med.* **2021**, *10*, 4388. https://doi.org/ 10.3390/jcm10194388

Academic Editor: Nuri B. Farber

Received: 30 July 2021 Accepted: 23 September 2021 Published: 25 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

withdrawal [7,8] and they do not always respond to treatment with medication [9,10] Rhythmic alterations commonly observed in patients with SUD are amplitude reduction and phase delay, which in severe cases can lead to chronodisruption or disappearance of rhythmicity [8]. This affectation is related both to the type of substance used and to the person's metabolism and tolerance (i.e., sensitivity to reward) [11].

The relationship between circadian rhythms and SUD is bidirectional, with eveningtype as a precipitating factor for drug use and with drug use generating chronodisruption [8,9]. Moreover, sleep disturbances have been associated with a higher risk of drug use and relapses [12,13], while the magnitude of the phase delay with the degree of dependence [14]. Thus, the exploration of the affectation and recovery of circadian rhythmicity seems to be of special relevance in patients under treatment for SUD [7,15]. In relation to the circadian typology, there seem to be rhythmic differences depending on the chronotype. The evening typology has been identified as a probable risk factor for the development of SUD, while the morning typology has been identified as a protective factor [16,17]. The reinforcing effect of the substances, mediated by clock genes, is greater in evening typology, especially during adolescence and early youth [16,18].

Studies in patients with SUD who have completed a detoxification phase indicated they have a higher prevalence of the morning and intermediate typologies, compared with those who have not initiated treatment, which seems to be a positive characteristic linked to better adherence to treatment [7,15]. Likewise, these patients exhibited a more robust circadian pattern of distal skin temperature (DST) than healthy controls, which is associated with longer abstinence periods and related to some of the treatment strategies used in therapeutic communities (e.g., strong daytime activity and morning habits). Furthermore, poorer sleep quality in patients with SUD is linked to their age and to a greater fragility of the circadian rhythm and chronodisruption [15].

There is a grea<sup>t</sup> lack of knowledge about the circadian characteristics in patients with SUD and comorbid SMI. Thus, to date no study has been conducted in dual patients with the diagnoses of schizophrenia (SZ+), bipolar disorder (BD+) and major depressive disorder (MDD+). In this sense, the available evidence on circadian rhythmicity in psychiatric conditions without SUD is limited and heterogeneous, indicating disruptions in prodromal phases in patients without medication [19,20] and in those with remission symptoms or in a sustained withdrawal phase [15,21–23]. All of this evidence suggests that circadian alterations could not only be symptoms but a significant clinical characteristic that affects the appearance and development of SMI [24,25].

Patients with schizophrenia show phase delays, free-running rhythms of 48 h or less than 24 h in the sleep-wake cycle or in melatonin secretion, flat amplitude and fragmentation of the activity-rest pattern [26–29]. All these alterations could be related to poor endogenous control and/or inadequate exposure to external synchronizers [28]. On the other hand, disruptions in the sleep-wake rhythm in schizophrenia have been associated with both clinical and functional prognosis. Deficits in sleep quality have been related with the predominance of positive symptoms [28,30,31] and a lower amplitude and inter-daily stability in the activity-rest rhythm, with poorer neurocognitive performance [27]. Some authors have observed an evening typology predominance in these patients compared to healthy controls [32,33], while in other work no differences have been appreciated [34]. None of these studies found a relationship between circadian typology and the age of the patient, as observed in healthy control subjects. Overall, no previous work has evaluated the circadian rhythm of DST in patients with schizophrenia.

Regarding patients with bipolar disorder, the findings indicate a reduced amplitude activity-rest rhythm, with phase delays (for example, melatonin secretion), greater rhythm fragmentation and more prevalent evening-type [35,36]. Furthermore, although depressive and maniac mood episodes change the rest and activity patterns, circadian alterations persist in euthymic phases [22], which seems to be relevant for the differential diagnosis and for non-psychiatric groups at risk of bipolar disorder [37,38].

Stabilized bipolar patients do not differ in nocturnal activation but they do show less intradaily variability compared to control participants, which may be due to depressive symptoms dependent on variable mood [23,39]. Two systematic reviews that examined activation and energy patterns found that both the euthymic and depressed bipolar groups differ from the controls by a lower mean activity mediated by mood, also concluding that it may be a consequence of bipolar disorder itself [21,40]. It is remarkable that the results that point out a delayed circadian phase in these patients are associated with a younger age, a shorter duration of bipolar disorder and more frequent depressive episodes [41], while those patients who show an advanced phase presented manic episodes and more suicide attempts [42].

Finally, in relation to circadian characteristics in MDD+ patients, only one published study [15] compared these dual patients with SUD ones and explored the possible influence of outpatient vs. residential treatment in therapeutic community. Such work described that the SUD group in therapeutic community presented a better adjustment to the light-dark cycle and a better DST pattern (greater amplitude, relative amplitude and percentage of rhythm and lower minimum temperature average) compared with MDD+ and with patients under outpatient treatment. Furthermore, the therapeutic community patients had the highest prevalence of morning-type regardless of their psychiatric diagnosis. These observations contrast with studies that have described an association of the evening typology with SUD [6,14] and with depression [43,44], although in none of these studies the participants were under residential treatment.

Even though the circadian rhythmic alteration or even its chronodisruption are not precipitating factors for mental disorders, they are related to a greater clinical symptomatology, more difficulties for remission, worse clinical prognosis, lesser healthy habits and worse quality of life [6,45]. All of this can be applicable to both SUD and the three comorbid diagnoses (SZ+, BD+ and MDD+) that have focused our attention on this study.

Therefore, the main goal of this work is to explore the possible differences in circadian rhythmicity in a sample of under treatment patients with SUD taking into account their comorbid SMI. Additionally, we aim to elucidate the possible relationship among circadian rhythmicity with epidemiological and clinical characteristics. This research could provide data of interest and applicability at the therapeutic level, especially when it comes to improve treatment adherence and recovery of dual patients with different SMI.

#### **2. Materials and Methods**

#### *2.1. Study Design and Participants*

The present study has a cross-sectional multicenter design, with a sample of 114 male patients with a diagnosis of dual disorder undergoing treatment for SUD (outpatient or residential) in different public and private specialized services located in the province of Barcelona. All patients with SUD were divided into three groups according to the comorbid psychiatric diagnosis: SZ+ (*n* = 38), BD+ (*n* = 36) and MDD+ (*n* = 40).

The inclusion criteria for the study were: (a) men between 20 and 50 years of age; (b) diagnosis of SUD (dependence) in initial remission phase according to Diagnostic and Statistical Manual of Mental Disorders (DSM-5) criteria [46] (c) comorbid psychiatric diagnosis of Schizophrenia, Bipolar Disorder or Major Depressive Disorder, not induced by substances or due to medical condition and (d) a minimum abstinence period of three months up to one year. The consideration of including only men patients in our sample is based on the higher prevalence of SUD in this gender and due to significant greater proportion of men in treatment facilities [2]. Moreover, only males were included to avoid possible biases in the circadian characteristics generated by the differential consumption patterns observed in men vs. women [2]. On the other hand, the exclusion criteria were: (a) patients with unstable or uncontrolled psychiatric symptoms or (b) inability (intellectual, cognitive, developmental or physical) to complete the assessment. Disorders related to caffeine and nicotine consumption were not considered as SUD, although data related to the consumption of both substances were recorded.

For the comparison of temperature data, a group of 40 male healthy control (HC) volunteers (mean age 36.50 yrs.; SD = 8.83; age range 21–51 yrs. old) recruited and assessed at the University of Murcia was also included. Regarding these participants 62.5% were married/with a stable partner, 25% were single and 12.5% were separated/divorced; the majority of them were active (working 88%), and very few were unemployed (10%) or with a disability pension (2%). None of the participants in the HC group had a medical or psychiatric diagnosis, nor any past or present SUD, and they were not under any kind of pharmacological treatment.

This study was approved by the Ethics Committee of the University of Barcelona (registration number: IRB00003099) and complied with the ethical principles of the Declaration of Helsinki. Participation in the study was voluntary and the patients did not receive any compensation except for individualized verbal feedback of their results.

#### *2.2. Sociodemographic and Clinical Assessment Instruments*

Sociodemographic and clinical data were collected through the Structured Clinical Interview for Axis I Disorders of the DSM-IV (SCID-I) [47], together with a structured interview specifically designed for our study. All collected data were corroborated by the psychologist/psychiatrist in charge of patient's treatment, as well as checked in the clinical records of each treatment center.

The Positive and Negative Syndrome Scale (PANSS) [48] in its Spanish version was used [49] for the assessment of psychiatric symptoms in the SZ+ group. The PANSS scale yields scores in four areas related to different symptomatology: positive syndrome, negative syndrome, composite scale and general psychopathology. In the BD+ group, we used the Young Mania Rating Scale (YMRS) [50] to measure the severity of maniac symptoms as well as the Hamilton Depression Rating Scale (HDRS) [51]. The YMRS in its Spanish version [52] gives a total score from 0 to 60 understood as it follows: 0–6 euthymic, 7–20 mixed episode and >20 possible maniac episode. On the other hand, the HDRS was used to assess depressive symptoms for the BD+ and MDD+ groups, with its 17-item Spanish version [53] and cut-off points being: 0–7, no current depression; 8–13, low; 14–18, mild; 19–22, severe; and >23, very severe depressive symptoms.

#### *2.3. Circadian Assessment Instruments*

To evaluate the circadian typology, the Spanish version of the Composite Morning Scale (CSM) [54] was used, consisting of 13 items and a total score from 0 to 55. Its interpretation considers the following cut-off points: 13–25 as an evening typology, 26–36 as an intermediate typology and 37–55 as a morning typology. The sleep-wake schedules were recorded using the structured interview designed for our study.

DST was recorded using the Thermochron iButton® DS1921H (Maxim Integrated Products, San Jose, CA, USA), previously programmed to take measurements every 2 min for 48 consecutive hours with an accuracy of ±0.125 ◦C. The sensor, which is attached to a strap similar to that of a wristwatch, was placed on the wrist of the non-dominant hand over the temporal artery [55].

### *2.4. Statistical Analysis*

For the sociodemographic, clinical and SUD data, descriptive statistics were calculated for the three groups of patients (mean, standard deviation, frequencies and percentages) and subsequent contrasts were performed with ANOVA and Chi-square, depending on the data were parametric or non-parametric.

For the analysis of the DST data, the CircadianwareTM software version 7.1.1 [56] was used. The parametric analyses of cosinor (maximum and minimum temperature, mesor, amplitude, acrophase and percentage of variance explained by the cosine wave), and the analysis of the Rayleigh vector and the Fourier analysis with the first 12 harmonics were made to characterize the circadian rhythm of the DST. The circadianity index was calculated as detailed in previous publications [57]. Non-parametric analyses were performed [55,58]

to obtain the values of interdaily stability (IS), intradaily variability (IV), relative amplitude (RA), maximum mean temperature in 5 consecutive hours (M5), temperature minimum average in 2 and 10 consecutive hours (L2 and L10).

DST values, both parametric and non-parametric, and sleep schedules (after transformation to the centesimal system) were evaluated using MANCOVA, while IS and CSM scores were analyzed with ANCOVA. In all cases, age was considered as a covariate, the analyses were performed with the diagnostic group as a factor (SZ+, BD+ and MDD+) and they were repeated considering treatment modality (outpatient/residential). Furthermore, in the parametric analyses for the DST the HC group was also incorporated together with the three clinical groups. Correlational analyses were also performed among DST and clinical variables. Subsequently, a linear regression analysis was carried out with significant correlations at the level of *p* = 0.01. The effect size was calculated as an estimate of the risk of committing type I error with the partial square index of Eta (*ηp*2), assuming values of 0.01 as low, 0.06 as moderate and 0.14 as high [59]. Bonferroni test was applied in all the post-hoc contrasts. The data of the present study have been analyzed using the Statistical Package for the Social Sciences program (IBM SPSS Statistics 25.0, Armonk, New York, United States). Two-sided statistical significance was established with a predefined type I error of 5% (*p* < 0.05).
