Introduction

Inflammatory bowel disease (IBD) is a term for nonspecific inflammatory diseases of the digestive tract. Canonical types include ulcerative colitis (UC) and Crohn disease (CD), with distinct clinical manifestations, prognosis, and treatment.

IBD carries a psychological burden, leading to a relatively high rate of sleep and emotional disorders in the affected patients. Insomnia and low sleep quality might afflict every other individual with IBD; even higher rates are observed in those with the disease exacerbation.1-4 Depression is associated with disturbed sleep; its prevalence in IBD ranges from about 20% in the disease remission to 35% during its exacerbation.2

The causes of poor mental health in IBD are elusive. They might be associated with treatment (steroids, immunosuppressive therapy) and the disease symptoms (pain, gastrointestinal symptoms). The lesser-known mechanism driving the psychiatric aspect of IBD is related to neurotrophins, in particular brain-derived neurotrophic factor (BDNF).5 This protein and its precursor, proBDNF, exert a pleiotropic effect on the organism, mediating the interactions between the central nervous system and the periphery.6,7

Circadian rhythm has been gaining notoriety as a factor in developing immune-mediated diseases and psychiatric comorbidities on a molecular level.8 This rhythm regulates a majority of biological processes, from metabolism on a cellular level to sleep / wake cycle. The genes of the circadian rhythm are known as “clock genes”; they comprise, among others, brain and muscle ARNT-Like 1 (BMAL1), circadian locomotor output cycles kaput (CLOCK), neuronal PAS domain protein 2 (NPAS2), or nuclear receptor subfamily 1 group D member 1 (NR1D1) gene. Their expression is ubiquitous in the tissues and shows some autonomy from the central master clock, that is, the suprachiasmatic nucleus. The major feedback loop that drives the circadian rhythm is the interaction between CLOCK/BMAL1 heterodimer, period circadian protein homolog (PER), and cryptochrome (CRY).9 It is mediated by a transcription regulator E-box. In mammals, the CLOCK/BMAL1 heterodimer activates the transcription of the PER and CRY by binding to the E-box; next, the PER and CRY repress their own expression.10,11 The NR1D1 protein binds to receptor-related orphan receptor response element inhibiting the expression of the BMAL1 and CLOCK genes.12,13 The NPAS2 gene is a CLOCK paralog. It is capable of dimerizing with BMAL1, substituting CLOCK.

The exact role of the clock genes in IBD pathogenesis is unclear. It seems that their expression is important for the functioning of the immune system cells. Thus, alterations to the circadian rhythm might alter the immune response, potentially changing the IBD course. It has been shown that BMAL1-deficient mice had arrhythmic and decreased cell proliferation in the intestines, resulting in impaired tissue regeneration.14 Another report showed that the expression of the clock genes, such as the CLOCK and the BMAL1 genes, was downregulated in IBD, as compared with healthy controls (HCs).10 These findings indicate that the level of the circadian rhythm gene expression might contribute to the pathomechanisms of IBD. Complex and bilateral relationships between the inflammation and the clock genes are well evidenced.13,15-18 It might be suspected that increased levels of some proinflammatory cytokines, also resulting from disrupted sleep, further enhance changes in the clock gene expression caused by IBD, creating a vicious circle.19 Furthermore, there might be a positive feedback loop between the sleep disorders and the circadian rhythm disruptions.20 However, there are few studies investigating alterations of the circadian rhythm in relation to sleep disturbances and depression symptoms in this group of patients.

Antitumor necrosis factor (TNF) therapy is effective and has a good safety profile.21-23 It does not appear to cause sleep disruptions, with some studies showing it may even improve sleep quality.3,24,25 Its influence on the clock gene expression in IBD has not been extensively studied. As mentioned above, the circadian rhythm interacts with the immune system; cytokines such as TNF or interferon-γ might modify it directly in the suprachiasmatic nucleus.26 In the macrophages, those cytokines were shown to inhibit the expression of the PER2 gene independently of the canonical JAK/STAT pathway.27 On the other hand, anti-inflammatory interleukin-4 promoted the expression of the aforementioned gene.27

Interactions between the circadian rhythm and biologic treatment are an important subject, as they help explain the links between the clock genes and the immune system. Additionally, their understanding might aid in developing a chronotherapeutic approach to IBD.

Therefore, the aim of the study was to compare the expressions levels of the BMAL1, CLOCK, NPAS2, and NR1D1 mRNA between the IBD patients and HCs in peripheral blood leukocytes (PBLs). The association between the gene expression and the disease severity, anti-TNF therapy, sleep quality, insomnia, and depression was also evaluated.

Patients and Methods

Recruitment and patient eligibility

The study participants (n = 125) were recruited at the Department of Digestive Tract Diseases (Medical University of Lodz, Poland). HCs (n = 44) were chosen to match the IBD patients (n = 80) as closely as possible in terms of age, sex, ethnicity, and body mass index (BMI).

The inclusion criteria comprised signing the informed consent, age between 18 and 65 years, and either CD or UC diagnosis based on the endoscopic, clinical, and histopathologic criteria. The exclusion criteria included a diagnosis of a chronic inflammatory disease, abdominal surgery in the last 6 months, use of psychoactive substances, diagnosed psychotic disorders, and malignancy except for basal cell carcinoma.

The Bioethical Committee of the Medical University of Lodz, Poland approved the study protocol (KE/1139/20).

Material collection

Venous blood samples (9 ml) were collected from all participants. The participants also filled out the following questionnaires: the Athens Insomnia Scale (AIS), the Pittsburgh Sleep Quality Index (PSQI), the Epworth Sleepiness Scale (ESS), the Beck Depression Inventory (BDI), the Harvey–Bradshaw Index (HBI), and the Partial Mayo Score (PMS). From the participants who qualified for the biologic treatment, venous blood (9 ml) was drawn at 2 time points, that is, before starting the anti-TNF therapy and after the induction therapy (after 14 weeks). The blood was collected within fixed time interval (between 9.00 and 11.00 AM)

Assessment of the disease severity and psychological variables

The disease severity was assessed using the HBI in CD and the PMS in UC patients. The HBI assesses 5 domains: the patient’s well-being, the intensity of abdominal pain, the number of liquid or soft stools the day before the examination, the occurrence of complications, such as arthralgia, uveitis, erythema nodosum, and the appearance of a new fistula or a new abscess.28 Steroid-free individuals who scored less than 5 points were assigned to the CD remission group. The PMS has 3 components: stool frequency, rectal bleeding, and physician’s global assessment evaluated on a 4-point scale.29 The individuals who scored less than 2 points were qualified for the UC remission group. The same person evaluated the HBI and the PMS scores in all participants.

Apart from the disease severity, the participants were interviewed about their past medical history and divided into 2 groups based on the presence of comorbid chronic diseases. For the purpose of the study, these included any noninflammatory chronic conditions (eg, type 2 diabetes, hypertension, atherosclerosis).

The participants filled out questionnaires assessing their sleep quality, severity of insomnia, depression symptoms, and excessive daytime sleepiness. The BDI is a screening tool for depression consisting of 21 items.30 A patient can assess the severity and frequency of symptoms on a 4-point scale. An outcome of 11 points or above indicates the possibility of depression. The AIS assesses symptoms of insomnia.31,32 It comprises 8 domains, such as sleep continuity or daily functioning; each item might be rated from 0 to 3 points. A score above 5 points indicates insomnia. The PSQI assesses general sleep quality.33 It includes various subjective domains (ie, trouble falling asleep, sleep continuity, daily functioning). The cutoff score of above 5 points indicates poor sleep quality. In the ESS, an individual assesses the probability of falling asleep in 8 different situations on a scale from 0 to 3.34 A score greater than 10 is considered abnormal and suggests the need for a more thorough diagnostics.

Antitumor necrosis factor therapy

A subset of patients (n = 26) in the exacerbation (according to the criteria established by the national insurer) was included in the anti-TNF therapy (infliximab or adalimumab) and treated in accordance with the guidelines of the European Crohn’s and Colitis Organization.35,36 The patients received the anti-TNF therapy under the Minister of Health’s treatment program. The study did not affect the recruitment of patients for the treatment or the dosage method. Infliximab was administered intravenously at weeks 0, 2, 6, and 14, at a dose of 5 mg/kg body weight. Adalimumab was administered subcutaneously at week 0 (160 mg), 2 (80 mg), then every 2 weeks, at a dose of 40 mg. The follow-up visit took place after 14 weeks of the therapy and involved the collection of peripheral blood, the assessment of the disease clinical severity, as well as completion of the questionnaires.

Evaluation of gene expression

mRNA isolation from PBLs was performed using the TRIzol reagent (Invitrogen, Waltham, Massachusetts, United States). The RNA Integrity Number and the concentration of the isolated RNA were assessed using a Nanodrop Colibri microvolume spectrometer (Titertek Berthold, Pforzheim, Germany). The obtained material was reversely transcribed with a dedicated kit according to the protocol provided by the manufacturer (SuperScript IV First-Strand Synthesis System, Thermo Fisher Scientific Inc., California, United States). The process comprised 3 steps in which the assays underwent annealing at 60 °C for 60 seconds. The expression levels of the selected genes was determined by quantitative real-time polymerase chain reaction; the applied mixture consisted of nuclease-free water, master mix TaqMan Universal, cDNA, gene specific probes (TaqMan assays for the BMAL1, CLOCK, NPAS2, NR1D1 genes; reference gene, β-actin) (Invitrogen). Three reactions were performed for each sample and the reference gene. For each sample, the cycle threshold (CT) was calculated. The results were presented as ΔCT and analyzed using the Livak method.37

Statistical analysis

Statistical analysis was performed using Statistica 13.1 PL (StatSoft, Tulsa, Oklahoma, United States). A value of P below 0.05 was considered significant. The distribution of continuous variables (normal / non-normal) was assessed using the Shapiro–Wilk test. Data were presented as mean with SD or median with interquartile range for normal and non-normal distribution, respectively. The differences between the independent groups with non-normal distribution were analyzed with the Mann–Whitney test. The Wilcoxon signed-rank test was used for repeated measures analysis of differences between the dependent groups with distribution other than normal. For the independent groups with non-normal distribution, the t test was applied. Correlations were calculated with the Spearman correlation test.

Results

The baseline characteristics of all study participants are summarized in Table 1. There were no significant differences in age, sex, BMI, smoking status, or presence of concomitant chronic diseases between the control and the study groups. There were also no significant differences between the IBD patients in remission and exacerbation regarding all the abovementioned parameters.

Table 1. Baseline characteristics of the study participants

Parameter

IBD

HC

P value

All

Exacerbation

Remission

N

80

47

33

44

–

CD, n

49

29

20

–

–

UC, n

31

18

13

–

–

Women, n (%)

45 (56.3)

25 (53.2)

20 (60.6)

22 (50)

0.51a

0.5b

Age, median (IQR)

34.5 (28–41)

34 (29–41)

36 (25–42)

31.5 (25–45.5)

0.61a

0.81b

BMI, kg/m2

23.2 (20.6–25.8)

23.5 (20.8–27.1)

22.5 (19.7–25)

23.9 (20.6–26.8)

0.13a

0.58b

Smoker, n (%)

11 (13.8)

6 (12.8)

5 (15.2)

5 (11.4)

0.5a

0.79b

Chronic diseases, n (%)

19 (23.8)

15 (31.9)

4 (12.1)

6 (13.6)

0.06a

0.24b

Steroidsc, n (%)

26 (32.5)

24 (51.1)

2 (6.1)

0

<⁠0.001a

Thiopurinesd, n (%)

28 (35)

15 (31.9)

13 (39.4)

0

0.49a

a Exacerbation vs remission

b IBD vs HC

c Budesonide (n = 2), systemic corticosteroids (n = 24)

d Mercaptopurine (n = 1), azathioprine (n = 28)

Abbreviations: BMI, body mass index; CD, Crohn disease; HC, healthy controls; IBD, inflammatory bowel disease; IQR, interquartile range; UC, ulcerative colitis

The IBD group had a decreased expression of the CLOCK (P = 0.001), NPAS2 (P = 0.001), and NR1D1 (P = 0.001) genes, as compared with the HCs. In the case of the BMAL1 gene, the difference was nonsignificant (P = 0.96; Table 2). In the UC group, the individuals with exacerbation had decreased expression of the CLOCK (P = 0.05) and NPAS2 (P = 0.04; Table 2) genes, as compared with those in remission; such a relationship was not observed in the CD group. Moreover, the severity of UC symptoms, as measured by the PMS, negatively correlated with the expression of the CLOCK (R = –0.389; P = 0.03; Supplementary material, Figure S1), NPAS2 (R = –0.483; P = 0.008; Supplementary material, Figure S1), and NR1D1 (R = –0.491; P = 0.006; Supplementary material, Figure S1; Table 3) genes. Such a correlation was not observed in the CD group.

Table 2. Clock gene expression in the study participants

Parameter

n

BMAL1 mRNA

n

CLOCK mRNA

n

NPAS2 mRNA

n

NR1D1 mRNA

IBD

79

9.868 (3.347–16.086)

77

7.609 (3.441–12.975)

74

0.428 (0.218–0.757)

78

12.086 (5.908–20.861)

HC

44

9.337 (6.802–13.674)

44

14.386 (11.498–20.56)

42

0.845 (0.526–1.264)

43

25.951 (16.254–36.321)

P value

–

0.96

–

<⁠0.001

–

0.001

–

<⁠0.001

CD

48

9.035 (3.756–15.278)

47

7.324 (3.752–13.02)

45

0.355 (0.233–0.757)

48

10.999 (6.221–18.167)

UC

31

10.503 (2.626–16.538)

30

8.410 (2.728–12.975)

29

0.581 (0.129–0.744)

30

12.150 (5.628–25.861)

P value

–

0.6

–

0.93

–

0.97

–

0.69

Ex

47

9.239 (2.342–14.802)

46

5.94 (2.208–10.576)

44

0.402 (0.137–0.732)

46

9.582 (4.7–17.664)

R

32

10.613 (4.447–17.465)

31

10.145 (4.985–13.433)

30

0.655 (0.233–0.971)

32

12.819 (7.505–21.262)

P value

–

0.21

–

0.08

–

0.23

–

0.24

UC Ex

18

10.095 (1.482–15.324)

18

5.084 (1.87–8.893)

17

0.334 (0.128–0.621)

18

8.672 (4.635–17.42)

UC R

13

11.613 (8.924–16.769)

12

11.93 (7.971–13.181)

12

0.741 (0.381–1.052)

12

16.93 (10.752–26.132)

P value

–

0.18

–

0.047

–

0.04

–

0.06

CD Ex

29

8.832 (4.509–13.573)

28

7.467 (3.413–10.706)

27

0.45 (0.215–0.887)

28

9.732 (6.553–19.262)

CD R

19

10.613 (3.188–18.16)

19

7.274 (3.752–14.649)

18

0.324 (0.233–0.728)

20

12.732 (5.683–17.926)

P value

–

0.78

–

0.62

–

0.95

–

0.97

Women

44

12.797 (5.5–18.383)

44

8.817 (4.985–14.182)

43

0.569 (0.298–0.837)

45

12.931 (7.375–21.522)

Men

35

7.583 (2.342–10.613)

33

6.053 (1.87–12.447)

31

0.279 (0.128–0.739)

33

8.801 (4.635–17.664)

P value

–

0.004

–

0.11

–

0.07

–

0.16

With chronic diseases

18

8.299 (3.925–16.769)

18

5.042 (2.728–8.832)

18

0.299 (0.118–0.666)

19

7.427 (4.635–12.233)

Without chronic diseases

61

10.145 (3.347–15.755)

59

7.987 (4.416–13.433)

56

0.536 (0.241–0.862)

59

12.931 (7.375–21.522)

P value

–

0.64

–

0.09

–

0.11

–

0.02

Smoking

11

8.472 (3.188–10.503)

11

8.127 (4.416–10.911)

11

0.244 (0.128–0.407)

11

12.533 (3.324–28.104)

Nonsmoking

68

10.058 (3.467–16.567)

66

7.299 (3.255–13.02)

63

0.581 (0.233–0.887)

67

10.287 (5.908–20.861)

P value

–

0.38

–

0.83

–

0.03

–

0.97

BDI >10

19

8.832 (1.085–10.503)

19

4.416 (1.388–10.503)

18

0.350 (0.218–0.621)

19

4.832 (2.011–12.708)

BDI ≤10

60

10.468 (4.764–16.624)

59

8.057 (4.985–13.433)

56

0.573 (0.22–0.894)

59

12.708 (7.583–21.747)

P value

–

0.07

–

0.04

–

0.21

–

0.003

AIS >5

38

9.435 (3.188–16.086)

38

8.419 (2.208–13.433)

36

0.402 (0.174–0.718)

39

10.287 (4.295–20.861)

AIS ≤5

41

10.503 (3.587–15.484)

39

7.274 (4.917–12.975)

38

0.573 (0.226–0.971)

39

12.233 (7.375–21.373)

P value

–

0.62

–

0.77

–

0.39

–

0.49

ESS >10

17

8.127 (4.509–16.653)

17

5.867 (1.806–10.075)

15

0.503 (0.118–0.887)

17

9.698 (2.824–17.664)

ESS ≤10

62

10.145 (3.347–15.484)

60

8.419 (3.812–13.41)

59

0.407 (0.233–0.757)

61

12.404 (7.199–20.861)

P value

–

0.68

–

0.11

–

0.56

–

0.36

PSQI >5

37

8.832 (3.188–14.15)

36

6.873 (2.189–10.743)

35

0.397 (0.218–0.68)

37

9.207 (3.324–16.769)

PSQI ≤5

42

10.613 (3.587–16.595)

41

7.742 (4.985–13.433)

39

0.632 (0.215–1.048)

41

14.052 (8.269–25.861)

P value

–

0.19

–

0.26

–

0.15

0.02

Pre anti-TNF

25

10.145 (8.127–16.595)

24

7.868 (4.951–10.873)

24

0.381 (0.226–0.894)

24

11.260 (7.25–21.634)

Post anti-TNF

25

5.343 (1.002–7.796)

24

9.261 (3.569–12.276)

22

0.372 (0.086–0.786)

25

15.864 (2.493–19.464)

P value

–

<⁠0.001

–

0.67

–

0.88

–

0.61

Data are presented as median (IQR).

P values <⁠0.05 were considered significant.

Abbreviations: AIS, Athens Insomnia Scale; anti-TNF, antitumor necrosis factor therapy; BDI, Beck Depression Inventory; BMAL1, brain and muscle ARNT-Like 1; CLOCK, circadian locomotor output cycles kaput; ESS, Epworth Sleepiness Scale; Ex, exacerbation; NPAS2, neuronal PAS domain protein 2; NR1D1, nuclear receptor subfamily 1 group D member 1; PSQI, Pittsburgh Sleep Quality Index; R, remission; others, see Table 1

Table 3. Correlations between the clock gene expression and the disease activity

mRNA

HBI

PMS

CLOCK

R

–0.151

–0.389

P value

0.31

0.03

BMAL1

R

0.01

–0.352

P value

0.95

0.05

NPAS2

R

–0.059

–0.483

P value

0.7

0.008

NR1D1

R

–0.051

–0.491

P value

0.73

0.006

P values <⁠0.05 were considered significant.

Abbreviations: HBI, Harvey Bradshaw Index; PMS, Partial Mayo Score; others, see Table 2

In the IBD participants in remission, the expression of the BMAL1 (R = 0.39; P = 0.03; Supplementary material, Figure S2) and the CLOCK (R = 0.469; P = 0.008; Supplementary material, Figure S2; Table 4) genes correlated with the disease duration, as measured by years from diagnosis. Sex did not affect the expression of the studied circadian rhythm genes except for BMAL1, which was higher in women (P = 0.004; Table 1) than in men. Comorbid chronic diseases in the IBD patients were related to a decrease in the expression of the NR1D1 gene (P = 0.02; Table 1). In the IBD group, smoking was correlated with a decline of the NPAS2 expression (P = 0.03; Table 1), but not the other studied clock genes.

Table 4. Correlations between the clock gene expression, questionnaire scores, and disease duration

IBD

Exacerbation

Remission

R

P value

R

P value

R

P value

BMAL1 mRNA

YFD

0.192

0.09

0.086

0.56

0.385

0.03

PSQI

–0.109

0.34

–0.086

0.57

0.009

0.96

ESS

–0.058

0.61

–0.032

0.83

–0.015

0.94

AIS

–0.085

0.46

–0.093

0.54

0.019

0.92

BDI

–0.182

0.11

–0.07

0.64

–0.29

0.11

CLOCK mRNA

YFD

0.204

0.08

0.075

0.62

0.469

0.008

PSQI

–0.122

0.29

–0.013

0.93

–0.166

0.37

ESS

–0.15

0.19

0.092

0.54

–0.397

0.03

AIS

–0.051

0.66

0.009

0.96

–0.047

0.8

BDI

–0.15

0.19

–0.052

0.73

–0.213

0.25

NPAS2 mRNA

YFD

–0.022

0.85

–0.067

0.67

0.079

0.68

PSQI

–0.089

0.45

–0.091

0.56

0.053

0.78

ESS

–0.116

0.32

–0.076

0.62

–0.094

0.62

AIS

–0.065

0.58

–0.177

0.25

0.19

0.31

BDI

–0.117

0.32

–0.149

0.34

–0.022

0.91

NR1D1 mRNA

YFD

0.182

0.11

0.127

0.4

0.32

0.07

PSQI

–0.205

0.07

–0.074

0.62

–0.34

0.06

ESS

–0.125

0.28

0.033

0.83

–0.28

0.12

AIS

–0.097

0.4

0.001

0.99

–0.22

0.22

BDI

–0.37

0.001

–0.307

0.04

–0.44

0.01

P values <⁠0.05 were considered significant.

Abbreviations: YFD, years from diagnosis; others, see Tables 1 and 2

The IBD individuals suffering from pronounced depressive symptoms (BDI >10) had decreased expression of the CLOCK (P = 0.04) and NR1D1 (P = 0.003; Table 2) genes, as compared with those with less severe symptoms. Moreover, there was a negative correlation between the BDI score and the expression of the NR1D1 gene in all IBD patients, those in exacerbation, and those in remission (R = –0.37, R = –0.307, and R = –0.436; P = 0.001, P = 0.04, and P = 0.01, respectively; Supplementary material, Figure S3A–S3C). Excessive daytime sleepiness (ESS >10) or insomnia symptoms (AIS >5) did not correlate with the expression of the circadian clock genes in the IBD group (P = 0.05; Table 2).

Nevertheless, poor sleep quality (PSQI >5) in those patients was associated with decreased expression of the NR1D1 gene (P = 0.02; Table 2), as compared with those satisfied with their sleep quality. The ESS score also negatively correlated with the CLOCK gene expression in the patients in remission (R = –0.397; P = 0.03; Table 4; Supplementary material, Figure S3D), but not those in exacerbation.

Out of 25 patients, 8 did not respond to the biologic treatment with anti-TNF medications. This therapy did not affect the gene expression except for the BMAL1 gene, which decreased after the induction treatment (P = 0.001; Table 2). The therapy brought improvements in the BDI, AIS, ESS, and PSQI scores (P = 0.002, P = 0.04, P = 0.04, and P = 0.003, respectively; Table 5).

Table 5. The influence of treatment on sleep quality, insomnia, excessive daytime sleepiness, and symptoms of depression

Parameter

Value

P value

AIS before treatment

6 (3–10)

0.04

AIS after treatment

3.75 (3–5)

BDI before treatment

8 (5–10)

0.002

BDI after treatment

3 (3–6)

ESS before treatment

6 (5–9)

0.04

ESS after treatment

5 (3–7)

PSQI before treatment

6 (5–11)

0.003

PSQI after treatment

4 (3–5)

Data are presented as median (IQR).

P values <⁠0.05 were considered significant.

Abbreviations: see Tables 1 and 2

Discussion

The role of the circadian rhythm in IBD is poorly understood. Due to its regulatory role in the body metabolism and immune system, it might be suspected to modulate the pathophysiology and course of this disease.38 Three similar studies were conducted on the expression of the clock genes in IBD. All of them used intestinal biopsy samples to measure the level of the gene expression; only 2 assessed their expression in PBLs. The choice of the material is important, as the outcomes obtained with intestinal biopsies are similar, but not identical, to the pattern seen in PBLs. The reasons for these discrepancies are not clear; locally secreted proinflammatory cytokines as well as products of cell necrosis could contribute to this effect. In general, the expression of the circadian rhythm genes decreases in tissues of the IBD patients as compared with healthy individuals.10,12 This reduction might persist even in noninflamed biopsies.10,39 Considering that colonoscopy is an invasive procedure, studying the expression of the clock genes in PBLs seems a viable alternative. Moreover, changes in their expression may better explain the alterations in mental health, as the expression of the clock genes in PBLs is not affected by regional proinflammatory mediators.

In this study, we observed lower expression of the CLOCK, NR1D1, and NPAS1 genes, but not the BMAL1 gene in the peripheral blood of IBD patients than of the HCs. These findings differed from those published by Weintraub et al10 and Liu et al.12 In their studies, the expression of both the CLOCK and BMAL1 genes decreased in the IBD individuals as compared with the HCs. They also noted lowered expression of other circadian clock genes (CRY1/2, PER1/2). It is unlikely that the sample size contributed to this discrepancy, as our study group (80 IBD patients, 44 HCs) was larger than in the first study10 (32 IBD patients, 18 HCs) but similar to that in the second one12 (90 IBD patients, 42 HCs). The difference in the BMAL1 expression between the studies is difficult to explain, as its protein product closely cooperates with that of the CLOCK gene, maintaining the vital loop of the circadian rhythm. This could indicate some sort of an interaction between the BMAL1 gene and its paralog, BMAL2.40 In mice, BMAL2 is dependent on BMAL1, but this link has not been thoroughly investigated in humans.40,41 However, it was demonstrated that these genes might show different levels of expression, suggesting the role of some external factors.41 It is not implausible that in our patients, the function of BMAL1 was partially fulfilled by BMAL2; thus the decrease in the BMAL2 expression was sufficient to disrupt the circadian rhythm without any changes to the BMAL1 expression. This hypothesis, considering rather sparse available research in the field, remains tenuous; more studies on BMAL2 and other paralogs of the core clock rhythm genes, as well as their respective loops are desirable to elucidate the compensational mechanisms of the circadian rhythm. Moreover, this study assesses only the mRNA expression; any further processes modulating the protein functionality were beyond our scope.

A novel finding of our study is that the expression of the NR1D1 and NPAS2 genes is reduced in PBLs of the IBD patients, as compared with the HCs. The NR1D1 expression was only studied in the intestinal biopsies by Wang et al,39 who also noted its decrease in the inflamed vs noninflamed samples. The expression of the NPAS2 gene has not been investigated in this context. As NPAS2 is a CLOCK paralog with the ability to act as a CLOCK substitute, its decreased expression further shows the extent of the circadian rhythm disruption and impairment of the potential compensational mechanisms.

While UC and CD belong to the same group of disorders, they present with many differences, such as expansion area (whole gastrointestinal system in CD vs colon in UC), or the type of inflammation (transmural in CD, mucosal in UC, etc).42-44 In this study, no differences were found in terms of the circadian gene expression between the CD and the UC groups. Nevertheless, we noted a stronger trend to the circadian rhythm disruption in the UC group. The individuals in exacerbation had significantly decreased NPAS2 and CLOCK expression, as compared with the UC patients in remission. We also obtained negative correlations for the expression of the CLOCK, NPAS2, and NR1D1 genes and the PMS. There were no such associations in the CD group. Liu et al12 also found no differences in the CLOCK and BMAL1 expression between the CD and the UC individuals; the CRY1/2 and PER1/2 mRNA levels were similar in these groups. They also observed a positive correlation between the expression of the PER1 and erythrocyte sedimentation rate (ESR) in the UC, but not CD or IBD group. Moreover, the PER1/2 levels also showed a positive correlation with C-reactive protein (CRP) only in the UC group.12 Both ESR and CRP are indicators of inflammation. The mentioned correlations suggest that interactions between these parameters and inflammatory processes are more pronounced in the UC than the CD patients. Further analyses involving clinical severity of the disease were not conducted. In contrary, Weintraub et al10 found the CLOCK, BMAL1, CRY1, and PER2 expression to be more reduced in the UC than the CD individuals. In their UC vs HCs comparison, the expression of the CLOCK, BMAL, CRY1/2, and PER1/2 was reduced in the UC group, and interestingly, for CD vs HCs comparison lowered expression was only confirmed for the CLOCK gene. Surprisingly, there was no correlation between CRP and gene expression. However, this could be due to a relatively large discrepancy in CRP levels between the CD and UC patients.10 However, a small sample size (UC, n = 5; CD, n = 8) could account for those unexpected outcomes. None of the mentioned researchers investigated the influence of the disease severity on the circadian clock genes. Differences in the results of the discussed studies could be partially ascribed to the demographic characteristics of the study group, namely age. The circadian rhythm changes throughout the lifetime and disintegrates at old age; moreover, younger individuals could have a different course of the disease (eg, shorter disease duration, more frequent extraintestinal manifestations).45-47 The patients studied by Weintraub et al10 were newly-diagnosed individuals aged 8–21 years, whereas only older adults participated in our study and that by Liu et al.12 Another reason for the variability in the findings might be a different pathophysiological background of UC and CD. This study does not include evaluation of selected inflammatory mediators, however, the results of the abovementioned studies10,12 indicate a complex relationship between the circadian rhythm and immunity in the UC patients. More research on this subject would be desirable.

A novel element of our study is a comprehensive assessment of sleep, symptoms of insomnia, and depression in relation to the circadian rhythm in IBD patients. There is a well-established relationship between the circadian rhythm and psychiatric disorders, mood disturbances in particular. Individuals afflicted with these conditions tend to experience a variety of anomalies related to sleep time, its quality, etc.48 Moreover, depression symptoms in most cases have a diurnal rhythm, being the most severe in the morning and decreasing throughout the day.48 However, studies on the clock gene expression in peripheral blood are lacking.

In this study, abnormally high severity of depressive symptoms was associated with decreased expression of the CLOCK and NR1D1 genes, with the latter being negatively correlated with the BDI score. Surprisingly, as revealed by all the methods we used for sleep evaluation, only poor sleep quality was associated with reduced NR1D1 expression; neither excessive daytime sleepiness nor symptoms of insomnia were significant in this regard. However, a subjective opinion on sleep might not necessarily mirror objective parameters; perhaps data obtained using objective methods (eg, actigraphy) would provide more insight into the influence of sleep disturbances and sleep time on the clock genes.

Another new finding is that the expression of the clock genes was not influenced by biologic treatment, except for BMAL1, which showed a remarkable decrease after 14 weeks of therapy. This indirectly confirms the findings of Yoshida et al,49 who observed that TNF promotes transcription of the BMAL1 gene. However, that observation also suggests that this form of therapy (ie, strong attenuation of inflammation via cytokine blocking) does not lead to more profound improvements in disrupted circadian rhythm. This finding additionally confirms that the clock genes do not correlate with the disease activity. A separate analysis of the UC and CD group before and after the treatment would be desirable; our study groups were too small to conduct such an investigation.

Study limitations

This study has several limitations. One of them is a lack of endoscopic assessment of the disease severity, routinely performed in other studies of this kind. Instead, we conducted a clinical evaluation of the disease severity, which should reflect the disease burden. It is difficult to decide which biological material is the most suitable for this type of study. PBLs have certain limitations, however, our results obtained with this material do not considerably differ from those in biopsy samples. A comprehensive comparison of the clock gene expression in the tissues and PBLs would be an interesting direction of future research.

Furthermore, in this and previous studies, the gene expression was assessed only at 1 time point (except for the patients receiving anti-TNF therapy). Due to the time-dependent expression of this group of genes, utilizing more time points in future projects would broaden the current knowledge on molecular aspects of the circadian rhythm.

We also have not investigated the expression of all core circadian rhythm genes. Thus, our results might not entirely reflect important alterations occurring in IBD. Moreover, only the changes regarding the level of gene expression were studied, and a lack of assessment of actual protein concentrations limits our insight into functional alterations of the circadian rhythm.

As previously mentioned, we only used subjective sleep assessment methods. However, in certain cases they might be a better diagnostic tool than objective instruments. Poor sleep and insomnia diagnosis are primarily based on a subjective perception of one’s sleep; polysomnography evaluates only a single night, whereas actigraphy does not take into account the patient’s state or physiological need for sleep. A combination of subjective and objective methods would probably yield the best results.

Our sample was relatively large; however, few patients were eligible for the biologic treatment, which did not allow us to compare the response to the treatment in the CD and the UC groups.

In summary, the expression of the clock genes is disrupted in the IBD patients. The IBD type does not appear to determine any specific pattern of changes in the expression of the discussed genes. Clinical activity of the disease, evaluated by questionnaires as either exacerbation or remission, is associated with alterations in the clock gene expression in the UC, but not the CD patients, and endoscopic assessment might provide more insight into ongoing disturbances. Poor mental health might be associated with decreased expression of the core genes of the circadian rhythm, similarly to poor sleep quality. However, more specific subjective sleep characteristics, such as symptoms of insomnia or increased daytime sleepiness, are not significant in this regard. The anti-TNF therapy does not significantly affect the expression of the clock genes, apart from BMAL1. However, as already discussed, our results regarding the BMAL1 expression differ from those in other studies; more research on this subject would be desirable to explain those inconsistencies.

To conclude, the circadian rhythm alterations are not specific to the IBD type, thus, they are rather related to some underlying inflammatory processes. However, the disease severity or anti-TNF therapy might not heavily alter the clock gene expression either, suggesting that those abnormalities are maintained throughout the disease course and do not depend on up- or downregulation of the proinflammatory mediators in the peripheral circulation. Nonetheless, there are significant differences between the CD and the UC patients regarding immunity and circadian rhythm alterations, which strongly suggest discrepancies in the pathomechanisms driving these 2 conditions. The outcomes of this study regarding interactions between the circadian rhythm and sleep are unexpected; perhaps an objective insight into basic sleep parameters would help elucidate them.

Further studies on the circadian rhythm could not only aid the treatment of the underlying immune-mediated conditions but could also facilitate development of a more comprehensive approach to therapy of its psychiatric comorbidities.