Background:

Fatigue is often reported by colorectal cancer survivors and largely impacts their quality of life. Inflammation has been linked to fatigue mainly in patients with breast cancer. Therefore, we investigated how inflammation is longitudinally associated with fatigue in colorectal cancer survivors, up to 2 years posttreatment.

Methods:

A total of 257 patients from the ongoing Energy for life after ColoRectal cancer cohort study were included in the analysis. Plasma levels of IL6, IL8, IL10, TNFα, high-sensitivity C-reactive protein (hsCRP), and fatigue were measured at 6 weeks, 6, 12, and 24 months posttreatment. Fatigue was measured through the validated Checklist Individual Strength (CIS; total, 20–140), consisting of four subscales – subjective fatigue (8–56), motivation (4–28), physical activity (3–21), and concentration (5–35), and the European Organisation for Research and Treatment of Cancer (EORTC) Quality of Life Questionnaire-Core 30 fatigue subscale (0–100). Linear mixed-models were used to assess the confounder-adjusted longitudinal associations between inflammatory markers and overall fatigue along with the subscales.

Results:

Mean levels of CIS fatigue decreased from 62.9 at 6 weeks to 53.0 at 24 months. In general, levels of inflammatory markers also decreased over time. No statistically significant longitudinal associations were found between IL6, IL8, IL10, TNFα, and fatigue. Higher levels of hsCRP were associated with more CIS fatigue (β per SD 3.21, 95% confidence interval (CI), 1.42–5.01) and EORTC fatigue (β 2.41, 95% CI, 0.72–4.10).

Conclusions:

Increased levels of hsCRP are longitudinally associated with more posttreatment fatigue in colorectal cancer survivors.

Impact:

These findings suggest that low-grade inflammation may play a role in fatigue reported by colorectal cancer survivors up to 2 years posttreatment.

Population ageing, screening programs, early detection, and more effective treatments have led to an increase in the number of colorectal cancer survivors (1). In 2020, worldwide, over 5 million individuals had a colorectal cancer diagnosis in the past 5 years (2). Due to the rising number of colorectal cancer survivors, it becomes increasingly important to address factors that impact their health-related quality of life (HRQoL) posttreatment. There are several chronic or late effects caused by both colorectal cancer and its treatment, such as fatigue, pain, bowel dysfunction, and emotional distress, all of which can affect a patient's HRQoL (3, 4).

Fatigue is a common and debilitating symptom experienced by colorectal cancer survivors during and posttreatment (5, 6). Reported rates of fatigue among colorectal cancer survivors range from 12% to 69.7% depending on the measurement instrument used and time elapsed since treatment (6–10). Results from prospective studies, including ours, and a systematic review showed that fatigue peaked between 6 weeks and 6 months posttreatment but persisted up to 2 years posttreatment (9, 11, 12). Many factors, such as treatment, comorbidities, and physical and psychologic factors, possibly contribute to cancer-related fatigue (5, 13). Furthermore, there is an increasing interest in the underlying biological mechanisms of fatigue (14).

Inflammation has been mainly identified as an underlying mechanism in posttreatment cancer-related fatigue, with the majority of studies performed in breast cancer survivors (13, 15, 16). Current thought is that production of pro-inflammatory cytokines in the periphery stimulates the brain resulting in fatigue, among other sickness behaviors (17, 18). Indeed, elevated circulating levels of pro-inflammatory markers, such as IL6, TNFα, and C-reactive protein (CRP), have been linked to more fatigue in breast cancer survivors (15, 19, 20). In contrast, anti-inflammatory cytokines, such as IL10, may attenuate sickness behavior, but little is known in relation to cancer-related fatigue (18, 21). Most longitudinal studies exploring the association between cancer-related fatigue and inflammation in breast cancer survivors focus on the period during or up to 6 months posttreatment and therefore have not assessed longer-term effects (22, 23). In addition, there are important differences between breast cancer and colorectal cancer survivors regarding several characteristics, namely age, sex, and treatment that can differentially affect fatigue. Some studies point to sex differences in both immune response and reporting of fatigue (24–27). Thus, despite the evidence of links between inflammation and fatigue in breast cancer survivors, a further exploration of this association is needed in colorectal cancer survivors.

Few studies, with differing methodologies, have explored the link between inflammation and fatigue in colorectal cancer survivors (9, 28–30). These methodologic differences include the measurement instruments used to assess fatigue, the start (pre- or posttreatment) and duration of follow-up time, and the availability of repeated measurements for both the inflammatory markers and fatigue. To our knowledge, only one study investigated the association between several inflammatory markers, excluding hsCRP, and fatigue up to 2 years posttreatment with repeated measurements over time (9).

Investigating how posttreatment inflammation is related to posttreatment fatigue over time will help to better understand the role of inflammation in the progression of cancer-related fatigue in colorectal cancer survivors. Therefore, the primary aim of this study is to determine how plasma levels of inflammatory markers, namely IL6, IL8, IL10, TNFα, high-sensitivity C-reactive protein (hsCRP), are longitudinally associated with overall fatigue, as well as different dimensions of fatigue (subjective fatigue, motivation, physical activity, and concentration) in colorectal cancer survivors followed up from 6 weeks until 2 years posttreatment.

Study design and population

Data analysis was performed with longitudinal data collected from April 18, 2012 up until November 1, 2016, from the Energy for life after ColoRectal cancer (EnCoRe) study. The EnCoRe study is an ongoing prospective cohort study with patient recruitment at three participating centers: Maastricht University Medical Center+, VieCuri Medical Center, and Zuyderland Medical Centre (31). Eligible for participation were men and women above the age of 18, diagnosed with stage I to III colorectal cancer. Exclusion criteria were stage IV colorectal cancer, inability to understand and speak Dutch, residential address outside of the Netherlands, or the presence of comorbidities that could impede a successful study participation, including cognitive and visibility/hearing disorders (31).

Patients were enrolled at diagnosis and followed up with repeated measurements at 6 weeks (n = 237), 6 months (n = 184), 12 months (n = 150), and 24 months posttreatment (n = 63). Study measurements were performed during home visits. In case participants were ill (e.g., the flu) or hospitalized, home visits were postponed. Participation rate at diagnosis was 46% and >90% at all posttreatment follow-up visits (Supplementary Fig. S1). The main reason for the decrease in sample size as follow-up time increases was that not all participants included at diagnosis had reached the subsequent follow-up points on November 1, 2016. The EnCoRe study was approved by the Medical Ethics Committee of the Academic Hospital Maastricht and Maastricht University, the Netherlands (Netherlands Trial Register no. NL6904). The study was conducted in accordance with the principles of the Declaration of Helsinki (version 7, October 2008).

Plasma inflammatory markers (exposure)

Fasting blood samples collected during home visits at 6 weeks, 6, 12, and 24 months posttreatment were used to assess plasma levels of inflammatory markers. After collection into EDTA tubes, blood samples were centrifuged, aliquoted into plasma, and stored in a freezer at –80°C until analysis (32). A custom-made multiplex assay using electrochemiluminescence detection (Meso Scale Diagnostics, Rockville, MD) was used to measure plasma concentration (pg/mL) of IL6, IL8, IL10, and TNFα. Assay plates were analyzed on a QuickPlex SQ 120 plate reader (Meso Scale Diagnostics), according to the manufacturer's instructions, at Wageningen University & Research, as described previously (32). Alongside the calibration curve, three quality controls were included per plate. All samples were analyzed in duplicates and the sample mean was accepted if the coefficient of variation (CV) was <40% (32). Inter- and intra-assay CVs were <8%, with reported values deviating less than 15% from target values (32). Levels of hsCRP were measured at 6 weeks, 6 and 12 months posttreatment. Plasma concentration (μg/mL) of hsCRP was determined through an immuno-MALDI (matrix-assisted laser desorption/ionization) mass spectrometry method (BEVITAL, Bergen, Norway) (33). The inter-assay CV ranged from 3% to 6%. hsCRP is used to measure lower levels of CRP which reflect low-grade systemic inflammation (34, 35).

Summary inflammatory z-scores were calculated to group the inflammatory markers and improve statistical efficiency (32, 36). Higher z-scores indicate higher levels of inflammation. First, normalized z-scores from each inflammatory marker were calculated as z =  (xij – μj) / σj, in which x is the participant's (i) inflammatory marker value at a given visit (j), μ is the study population mean, and σ is the study SD, both at given visits (j; ref. 32). Two summary inflammatory z-scores were computed for each participant, at each time point, to use all available data. One was calculated by summing the normalized z-scores of IL6, IL8, TNFα, hsCRP, and subtracting IL10, and thus only includes patients with measurements up to 12 months posttreatment. The other summary inflammatory z-score excluded hsCRP thereby including patients with data available at all posttreatment time points.

Fatigue (outcome)

The validated Checklist Individual Strength (CIS) and the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30 (EORTC QLQ-C30) fatigue subscale were used to measure fatigue at 6 weeks, 6, 12, and 24 months posttreatment. The CIS is a 20-item questionnaire composed of 4 subscales – subjective fatigue (8–56), motivation (4–28), physical activity (3–21), and concentration (5–35; ref. 37). The total fatigue score (20–140) was obtained by summing all item scores. Higher scores represent higher levels of fatigue. The EORTC QLQ-C30 fatigue subscale contains 3 items and ranges from 0 to 100 (38).

Although initially developed for patients with chronic fatigue syndrome (37), the CIS has been used to measure fatigue in cancer survivors (39). In a study among working people, the CIS was able to adequately distinguish fatigued and non-fatigued individuals (40). A recent study assessed the construct validity of the CIS subjective fatigue subscale and the EORTC QLQ-C30 fatigue subscale in cancer survivors (n = 320) and found a high Spearman rank correlation coefficient of 0.77 (41).

Other relevant variables

At the time of diagnosis, patients reported sex and birth date, which was used to calculate the age at each posttreatment time point (11). Data on treatment, such as chemotherapy and radiotherapy, were obtained from clinical records. The number of comorbidities at each posttreatment time point was determined using the 13-item Self-Administered Comorbidity Questionnaire (42). Height and weight, measured by trained dietitians, were used to calculate body mass index (BMI) at every time point. Current smoking status at each time point was self-reported. Information on use of nonsteroidal anti-inflammatory drugs (NSAID) during the 6 months prior to the follow-up time point was collected using self-reported questionnaires (32). Physical activity was evaluated using the validated Short QUestionnaire to ASsess Health-enhancing physical activity (SQUASH; ref. 43). Ainsworth's Compendium of Physical Activities was used to give activities a metabolic equivalent of task (MET) value (44). Activities were categorized as light physical activity (LPA) (<3 MET) or moderate-to-vigorous physical activity (MVPA; ≥3 MET), and total time spent in each activity was calculated as hours/week (11).

Statistical analysis

Descriptive analyses were performed to describe patient characteristics at 6 weeks (i.e., the baseline for longitudinal analyses). Categorical variables were presented as frequencies with percentages, and continuous variables as the mean with SD or medians with interquartile range (IQR) for normally and non-normally distributed data, respectively. Data on inflammatory markers, summary inflammatory z-scores, and fatigue, including the subscales, were presented for all posttreatment time points.

Linear mixed model regression was used to investigate the longitudinal associations between levels of inflammatory markers and fatigue (45). The regression coefficients obtained are a weighted average of the inter-individual (between-subject) differences and intra-individual (within-subject) changes (45). Therefore, separate hybrid models were used to disentangle the intra- and inter-individual components (46). To estimate the intra-individual association, the deviation of an individual's level of inflammatory marker from the person-mean was modelled. The regression coefficient from this model represents changes in fatigue over time in relation to a one-unit change in levels of inflammatory markers over time within individuals. To estimate the inter-individual association, a centered person-mean value of an inflammatory marker—difference between a subject's mean value of the inflammatory marker and the sample mean—was modelled to obtain a regression coefficient which indicates the average difference over time in fatigue between individuals in relation to a one-unit difference in mean levels of inflammatory markers between individuals.

To improve interpretability, levels of inflammatory markers were divided by their SD at 6 weeks to obtain regression coefficients that represented the difference in fatigue per SD increase of the inflammatory marker. The first model included age at measurement (years), sex (men/women), and time since diagnosis (days). The second model included additional potential confounders, selected a priori based on the available literature: NSAIDs (yes/no), BMI (kg/m2), physical activity (LPA and MVPA - hours/week), comorbidities (0, 1, ≥2), treatment (chemotherapy/radiotherapy, yes/no), and smoking status (yes/no). The likelihood ratio test was used to evaluate whether including a random slope improved the model fit (45). The FDR method (q < 0.05) was used to correct for multiple testing of the various exposures with the outcome (47). This was applied separately for each outcome (CIS: total fatigue, subjective fatigue, motivation, physical activity, concentration; EORTC fatigue). No correction was performed for the inflammatory z-scores because they are correlated with the inflammatory markers.

Post hoc subgroup analyses were performed for sex to explore the longitudinal associations in men and women separately; testing for interaction was done by including a product-term for the inflammatory marker and sex in each model. Sensitivity analyses excluding participants with recurrence and participants who died were performed.

To further explore the role of hsCRP, linear mixed model regression was performed with hsCRP categorized to represent normal values (≤3 mg/L), low-grade inflammation (3–10 mg/L), and acute inflammation (>10 mg/L; refs. 35, 48).

Statistical analyses were conducted in STATA (version 15). P values below 0.05 (two-sided) after correction for multiple testing were considered statistically significant.

Data availability

Data analyzed in the manuscript, code book, and analytic code will be made available upon request pending (e.g., application and approval, payment, other) to coauthor M.J.L. Bours.

Participant characteristics

Data on fatigue and IL6, IL8, IL10, and TNFα was available for 237 participants at 6 weeks, 184 at 6 months, 150 at 12 months, and 63 at 24 months posttreatment. Data on fatigue and hsCRP was available for 200 participants at 6 weeks, 148 at 6 months, and 114 at 12 months posttreatment. Participants were on average 67 years old, and the majority were men (68.8%; Table 1). There were some differences between men and women, notably a higher percentage of women had two or more comorbidities compared with men (women, 68.9%; men, 47.9%). Women also reported higher median levels of LPA than men (women: 14.0 hours/week, IQR 7.0–24.5; men: 6.3 IQR 1.2–12.0), but men reported higher median levels of MVPA than women (men: 9.0 hours/week, IQR 3.5–16.3; women: 4.1 IQR 1.5–7.0). The percentage of men who received radiotherapy was higher than that of women (men, 30.7%; women, 18.9%) and more men were diagnosed with rectum cancer (men, 42.3%; women, 29.7%).

Table 1.

Demographic, lifestyle, and clinical characteristics of stage I to III colorectal cancer survivors at 6 weeks posttreatment, overall, and according to sex.

Baseline characteristicsTotal population (n = 237)Men (n = 163)Women (n = 74)
Age, mean (SD) 66.8 (9.2) 66.3 (8.8) 68.1 (9.9) 
BMI (kg/m2)a, median (IQR) 27.3 (24.4–30.3) 27.3 (24.4–30.4) 27.7 (24.6–29.9) 
Use of NSAIDsb (yes), n (%) 20 (9.8) 14 (9.8) 6 (9.8) 
Physical activity (hours/week), median (IQR) 
 LPA 7.5 (2.0–16.5) 6.3 (1.2–12.0) 14.0 (7.0–24.5) 
 MVPA 7.0 (2.7–14.3) 9.0 (3.5–16.3) 4.1 (1.5–7) 
Smoking status (yes), n (%) 22 (9.3) 16 (9.8) 6 (8.1) 
Comorbidities, n (%) 
 0 49 (20.6) 42 (25.8) 7 (9.5) 
 1 59 (24.8) 43 (26.4) 16 (21.6) 
 ≥2 129 (54.4) 78 (47.9) 51 (68.9) 
Chemotherapy (yes), n (%) 89 (37.6) 62 (38.0) 27 (36.5) 
Radiotherapy (yes), n (%) 64 (27.0) 50 (30.7) 14 (18.9) 
Cancer type, n (%) 
 Colon cancer 146 (61.6) 94 (57.7) 52 (70.3) 
 Rectum cancer 91 (38.4) 69 (42.3) 22 (29.7) 
Baseline characteristicsTotal population (n = 237)Men (n = 163)Women (n = 74)
Age, mean (SD) 66.8 (9.2) 66.3 (8.8) 68.1 (9.9) 
BMI (kg/m2)a, median (IQR) 27.3 (24.4–30.3) 27.3 (24.4–30.4) 27.7 (24.6–29.9) 
Use of NSAIDsb (yes), n (%) 20 (9.8) 14 (9.8) 6 (9.8) 
Physical activity (hours/week), median (IQR) 
 LPA 7.5 (2.0–16.5) 6.3 (1.2–12.0) 14.0 (7.0–24.5) 
 MVPA 7.0 (2.7–14.3) 9.0 (3.5–16.3) 4.1 (1.5–7) 
Smoking status (yes), n (%) 22 (9.3) 16 (9.8) 6 (8.1) 
Comorbidities, n (%) 
 0 49 (20.6) 42 (25.8) 7 (9.5) 
 1 59 (24.8) 43 (26.4) 16 (21.6) 
 ≥2 129 (54.4) 78 (47.9) 51 (68.9) 
Chemotherapy (yes), n (%) 89 (37.6) 62 (38.0) 27 (36.5) 
Radiotherapy (yes), n (%) 64 (27.0) 50 (30.7) 14 (18.9) 
Cancer type, n (%) 
 Colon cancer 146 (61.6) 94 (57.7) 52 (70.3) 
 Rectum cancer 91 (38.4) 69 (42.3) 22 (29.7) 

aData on BMI is missing for 1 person.

bThirty-three participants have missing data for use of NSAIDs 6 weeks prior to measurement.

Fatigue and inflammatory markers

Total fatigue was highest at 6 weeks posttreatment (62.9, SD 26.5) and decreased over time, with the largest decrease occurring between 6 and 12 months posttreatment (Fig. 1; Table 2). Across all time points, women reported higher levels of fatigue compared with men. This was also observed in the subjective fatigue subscale, reduced motivation, reduced concentration, being most pronounced for subjective fatigue (Table 2; Supplementary Fig. S2). The CIS total fatigue and EORTC fatigue subscale were significantly correlated at all time points (range, 0.68–0.76). Median levels of IL6, IL10, and TNFα slightly decreased over the course of 24 months posttreatment, and for levels of hsCRP up to 12 months posttreatment, while IL8 increased between the 12 and 24 months time points (Fig. 1; Table 2). Pearson correlation coefficients indicated weak to moderated correlations between the inflammatory markers at 6 weeks (range, –0.02 to 0.45), with similar ranges at following time points (Supplementary Fig. S3).

Figure 1.

Median levels of inflammatory markers (A–E) and fatigue score (F and G) in stage I to III colorectal cancer survivors from 6 weeks to 24 months posttreatment, overall, and according to sex.

Figure 1.

Median levels of inflammatory markers (A–E) and fatigue score (F and G) in stage I to III colorectal cancer survivors from 6 weeks to 24 months posttreatment, overall, and according to sex.

Close modal
Table 2.

Fatigue and plasma inflammatory markers in stage I to III colorectal cancer survivors at 6 weeks, 6, 12, and 24 months posttreatment, overall, and according to sex.

Posttreatment follow-up measurements
6 weeks n = 237a6 months n = 184a12 months n = 150a24 months n = 63a
Total PopulationTotal populationTotal populationTotal population
Men n = 163Women n = 74Men n = 123Women n = 61Men n = 104Women n = 46Men n = 44Women n = 19
CIS (four subscales), mean (SD) 
 Total fatigue 62.9 (26.5) 59.9 (27.3) 53.2 (26.3) 53.0 (25.2) 
20–140 61.5 (25.9) 66.0 (27.9) 57.4 (26.6) 65.1 (28.0) 51.2 (25.4) 57.7 (27.9) 47.9 (23.7) 64.7 (25.3) 
Subjective fatigue 27.3 (13.4) 25.3 (12.9) 22.3 (12.4) 22.2 (13.1) 
8–56 26.4 (13.3) 29.3 (13.6) 24.1 (12.6) 28.0 (13.2) 21.0 (11.8) 25.4 (13.2) 19.0 (11.9) 29.5 (13.0) 
Reduced motivation 12.3 (6.1) 12.1 (6.2) 10.7 (6.1) 10.7 (6.0) 
4–28 11.9 (5.8) 13.0 (6.7) 11.3 (5.7) 13.6 (7.0) 10.1 (5.4) 11.9 (7.5) 9.7 (5.7) 13.1 (6.2) 
Reduced physical activity 10.5 (5.2)  9.6 (5.1) 8.4 (4.9) 8.2 (4.9) 
3–21 10.7 (5.0) 10.2 (5.5) 9.6 (5.2) 9.7 (4.8) 8.5 (4.9) 8.0 (4.9) 7.7 (4.9) 9.5 (4.8) 
Reduced concentration 12.8 (7.2) 12.9 (7.2) 11.8 (6.7) 11.8 (6.3) 
5–35 12.6 (7.2) 13.4 (7.2) 12.4 (7.3) 13.9 (6.9) 11.6 (6.9) 12.4 (6.3) 11.5 (6.0) 12.6 (7.0) 
EORTC QLQ-C30, Fatigue, mean (SD) 29.1 (22.7) 23.6 (22.0) 21.3 (23.6) 20.3 (22.4) 
0–100 28.3 (23.8) 30.9 (20.2) 22.0 (23.6) 27.0 (18.3) 18.2 (22.8) 28.3 (24.3) 14.6 (19.1) 33.3 (24.3) 
Inflammatory markers, median (IQR) 
IL6 (pg/ml) 1.5 (0.8–2.2) 1.3 (0.8–2.1) 0.9 (0.6–1.4) 0.9 (0.5–1.5) 
 1.4 (0.8–2.3) 1.5 (0.8–2.1) 1.3 (0.9–2.3) 1.1 (0.8–1.9) 0.9 (0.6–1.5) 0.8 (0.5–1.2) 0.8 (0.4–1.6) 1.0 (0.6–1.3) 
IL8 (pg/ml) 5.6 (4.4–7.3) 5.3 (4.4–7.0) 3.9 (3.1–4.8) 4.9 (3.8–7.0) 
 5.5 (4.3–6.8) 5.9 (4.5–8.1) 5.1 (4.4–7.1) 5.5 (4.6–7.0) 3.9 (3.2–4.8) 3.9 (2.9–4.7) 4.8 (3.6–5.9) 5.2 (4.3–7.7) 
IL10 (pg/ml) 0.4 (0.3–0.5) 0.4 (0.2–0.5) 0.3 (0.2–0.4) 0.2 (0.1–0.3) 
 0.4 (0.3–0.5) 0.4 (0.3–0.5) 0.4 (0.2–0.5) 0.4 (0.3–0.5) 0.3 (0.2–0.4) 0.2 (0.2–0.3) 0.2 (0.1–0.3) 0.2 (0.1–0.4) 
TNFα (pg/ml) 2.9 (2.4–3.8) 2.8 (2.3–3.6) 2.0 (1.6–2.5) 2.0 (1.6–2.9) 
 2.9 (2.4–3.6) 3.0 (2.2–4.0) 2.8 (2.3–3.7) 2.8 (2.3–3.4) 2.0 (1.6–2.4) 2.0 (1.5–2.6) 2.0 (1.6–2.7) 2.3 (1.6–3.4) 
Summary inflammatory z-scoreb –0.3 (–0.7 to 0.2) –0.4 (–0.8 to 0.2) –0.3 (–0.8 to 0.3) –0.5 (–1.2 to 1.4) 
excluding hsCRP, median (IQR) 0.3 (0.7 to 0.2) 0.3 (0.7 to 0.2) 0.3 (0.8 to 0.3) 0.4 (0.8 to 0.1) 0.4 (0.7 to 0.3) 0.3 (0.8 to 0.1) 0.6 (1.2 to 0.9) 0.1 (1.2 to 1.8) 
Posttreatment follow-up measurements
6 weeks n = 237a6 months n = 184a12 months n = 150a24 months n = 63a
Total PopulationTotal populationTotal populationTotal population
Men n = 163Women n = 74Men n = 123Women n = 61Men n = 104Women n = 46Men n = 44Women n = 19
CIS (four subscales), mean (SD) 
 Total fatigue 62.9 (26.5) 59.9 (27.3) 53.2 (26.3) 53.0 (25.2) 
20–140 61.5 (25.9) 66.0 (27.9) 57.4 (26.6) 65.1 (28.0) 51.2 (25.4) 57.7 (27.9) 47.9 (23.7) 64.7 (25.3) 
Subjective fatigue 27.3 (13.4) 25.3 (12.9) 22.3 (12.4) 22.2 (13.1) 
8–56 26.4 (13.3) 29.3 (13.6) 24.1 (12.6) 28.0 (13.2) 21.0 (11.8) 25.4 (13.2) 19.0 (11.9) 29.5 (13.0) 
Reduced motivation 12.3 (6.1) 12.1 (6.2) 10.7 (6.1) 10.7 (6.0) 
4–28 11.9 (5.8) 13.0 (6.7) 11.3 (5.7) 13.6 (7.0) 10.1 (5.4) 11.9 (7.5) 9.7 (5.7) 13.1 (6.2) 
Reduced physical activity 10.5 (5.2)  9.6 (5.1) 8.4 (4.9) 8.2 (4.9) 
3–21 10.7 (5.0) 10.2 (5.5) 9.6 (5.2) 9.7 (4.8) 8.5 (4.9) 8.0 (4.9) 7.7 (4.9) 9.5 (4.8) 
Reduced concentration 12.8 (7.2) 12.9 (7.2) 11.8 (6.7) 11.8 (6.3) 
5–35 12.6 (7.2) 13.4 (7.2) 12.4 (7.3) 13.9 (6.9) 11.6 (6.9) 12.4 (6.3) 11.5 (6.0) 12.6 (7.0) 
EORTC QLQ-C30, Fatigue, mean (SD) 29.1 (22.7) 23.6 (22.0) 21.3 (23.6) 20.3 (22.4) 
0–100 28.3 (23.8) 30.9 (20.2) 22.0 (23.6) 27.0 (18.3) 18.2 (22.8) 28.3 (24.3) 14.6 (19.1) 33.3 (24.3) 
Inflammatory markers, median (IQR) 
IL6 (pg/ml) 1.5 (0.8–2.2) 1.3 (0.8–2.1) 0.9 (0.6–1.4) 0.9 (0.5–1.5) 
 1.4 (0.8–2.3) 1.5 (0.8–2.1) 1.3 (0.9–2.3) 1.1 (0.8–1.9) 0.9 (0.6–1.5) 0.8 (0.5–1.2) 0.8 (0.4–1.6) 1.0 (0.6–1.3) 
IL8 (pg/ml) 5.6 (4.4–7.3) 5.3 (4.4–7.0) 3.9 (3.1–4.8) 4.9 (3.8–7.0) 
 5.5 (4.3–6.8) 5.9 (4.5–8.1) 5.1 (4.4–7.1) 5.5 (4.6–7.0) 3.9 (3.2–4.8) 3.9 (2.9–4.7) 4.8 (3.6–5.9) 5.2 (4.3–7.7) 
IL10 (pg/ml) 0.4 (0.3–0.5) 0.4 (0.2–0.5) 0.3 (0.2–0.4) 0.2 (0.1–0.3) 
 0.4 (0.3–0.5) 0.4 (0.3–0.5) 0.4 (0.2–0.5) 0.4 (0.3–0.5) 0.3 (0.2–0.4) 0.2 (0.2–0.3) 0.2 (0.1–0.3) 0.2 (0.1–0.4) 
TNFα (pg/ml) 2.9 (2.4–3.8) 2.8 (2.3–3.6) 2.0 (1.6–2.5) 2.0 (1.6–2.9) 
 2.9 (2.4–3.6) 3.0 (2.2–4.0) 2.8 (2.3–3.7) 2.8 (2.3–3.4) 2.0 (1.6–2.4) 2.0 (1.5–2.6) 2.0 (1.6–2.7) 2.3 (1.6–3.4) 
Summary inflammatory z-scoreb –0.3 (–0.7 to 0.2) –0.4 (–0.8 to 0.2) –0.3 (–0.8 to 0.3) –0.5 (–1.2 to 1.4) 
excluding hsCRP, median (IQR) 0.3 (0.7 to 0.2) 0.3 (0.7 to 0.2) 0.3 (0.8 to 0.3) 0.4 (0.8 to 0.1) 0.4 (0.7 to 0.3) 0.3 (0.8 to 0.1) 0.6 (1.2 to 0.9) 0.1 (1.2 to 1.8) 
Patients with data on hsCRP
n = 200cn = 148cn = 114cn = 0c
hsCRP (mg/L), median (IQR) 2.1 (1.1–5.3) 2.1 (0.8–4.6) 1.7 (0.8–5.3)   
 2.0 (1.0–5.5) 2.8 (1.7–5.1) 2.0 (0.8–4.8) 2.2 (0.7–4.4) 1.7 (0.7–5.7) 1.7 (0.8–4.7)   
Summary inflammatory z-score including –0.6 (–1.0 to 0.5) –0.5 (–1.1 to 0.5) –0.7 (–1.3 to 0.5)   
hsCRP d, median (IQR) 0.6 (1.1 to 0.5) 0.6 (0.9 to 0.5) 0.5 (1.1 to 0.5) 0.5 (1.0 to 0.5) 0.7 (1.3 to 0.4) 0.7 (1.3 to 0.5)   
Patients with data on hsCRP
n = 200cn = 148cn = 114cn = 0c
hsCRP (mg/L), median (IQR) 2.1 (1.1–5.3) 2.1 (0.8–4.6) 1.7 (0.8–5.3)   
 2.0 (1.0–5.5) 2.8 (1.7–5.1) 2.0 (0.8–4.8) 2.2 (0.7–4.4) 1.7 (0.7–5.7) 1.7 (0.8–4.7)   
Summary inflammatory z-score including –0.6 (–1.0 to 0.5) –0.5 (–1.1 to 0.5) –0.7 (–1.3 to 0.5)   
hsCRP d, median (IQR) 0.6 (1.1 to 0.5) 0.6 (0.9 to 0.5) 0.5 (1.1 to 0.5) 0.5 (1.0 to 0.5) 0.7 (1.3 to 0.4) 0.7 (1.3 to 0.5)   

aParticipants with data available on fatigue, IL6, IL8, IL10, and TNFα.

bThe inflammatory z-score was calculated as z =  (x – μ)/σ, in which x is the participant's inflammatory marker value at a given visit, μ is the study population mean, and σ is the study SD, both at given visits. The summary inflammatory z-score for each participant was computed by summing the z-scores of IL6, IL8, TNFα and subtracting IL10.

cParticipants with data available on fatigue, IL6, IL8, IL10, TNFα, and hsCRP.

dThe inflammatory z-score was calculated as z =  (x – μ)/σ, in which x is the participant's inflammatory marker value at a given visit, μ is the study population mean, and σ is the study SD, both at given visits. The summary inflammatory z-score for each participant was computed by summing the z-scores of IL6, IL8, TNFα, hsCRP, and subtracting IL10.

Longitudinal associations between inflammatory markers and fatigue

The coefficients presented represent the change in fatigue score for one SD increase of the inflammatory markers (Fig. 2; Table 3; Supplementary Fig. S4). In the fully adjusted models after FDR correction, there were no statistically significant overall, intra- or inter-individual associations between IL6, IL8, IL10, TNFα, and CIS total fatigue, as well as the subscales. Similar results were observed in the analyses with IL6, IL8, IL10, TNFα, and the EORTC fatigue subscale.

Figure 2.

Forest plots demonstrating beta-coefficients and corresponding 95% CI of overall longitudinal associations, including intra- and inter-individual associations, between inflammatory markers and CIS total fatigue (A) and EORTC QLQ-C30 fatigue (B) in colorectal cancer survivors followed-up at 6 weeks, 6, 12, and 24 months after treatment. Asterisk (*) indicates statistically significant associations after FDR correction for multiple testing.

Figure 2.

Forest plots demonstrating beta-coefficients and corresponding 95% CI of overall longitudinal associations, including intra- and inter-individual associations, between inflammatory markers and CIS total fatigue (A) and EORTC QLQ-C30 fatigue (B) in colorectal cancer survivors followed-up at 6 weeks, 6, 12, and 24 months after treatment. Asterisk (*) indicates statistically significant associations after FDR correction for multiple testing.

Close modal
Table 3.

Longitudinal associations between inflammatory markers and fatigue in stage I to III colorectal cancer survivors followed-up from 6 weeks to 2 years posttreatment.

EORTC QLQ-C30, Fatigue β (95% CI)CIS Total fatigue β (95% CI)Subjective fatigue β (95% CI)Reduced motivation β (95% CI)Reduced physical activity β (95% CI)Reduced concentration β (95% CI)
 Model I       
 Overall associationa 1.85 (0.233.47) 2.07 (0.383.76) 0.97 (0.121.83) 0.49 (0.070.90) 0.23 (–0.13 to 0.60) 0.48 (–0.00 to 0.97) 
 Intra-individualb 1.50 (–0.35 to 3.35) 1.40 (–0.44 to 3.24) 0.60 (–0.34 to 1.54) 0.35 (–0.11 to 0.82) 0.08 (–0.34 to 0.50) 0.37 (–0.18 to 0.91) 
IL6 Inter-individualc 3.04 (–0.35 to 6.44) 5.44 (1.319.57) 2.67 (0.664.67)* 1.00 (0.091.92) 0.74 (–0.01 to 1.50) 0.94 (–0.13 to 2.01) 
 Model II       
 Overall associationa 1.69 (0.083.32) 1.67 (–0.03 to 3.36) 0.71 (–0.14 to 1.56) 0.36 (–0.05 to 0.77) 0.12 (–0.24 to 0.49) 0.48 (–0.02 to 0.98) 
 Intra-individualb 1.74 (–0.11 to 3.60) 1.28 (–0.60 to 3.16) 0.46 (–0.49 to 1.41) 0.30 (–0.16 to 0.77) 0.05 (–0.37 to 0.48) 0.44 (–0.13 to 1.01) 
 Inter-individualc 1.56 (–1.74 to 4.86) 3.34 (–0.56 to 7.24) 1.71 (–0.19 to 3.61) 0.58 (–0.29 to 1.45) 0.32 (–0.40 to 1.03) 0.62 (–0.44 to 1.68) 
 Model I       
 Overall associationa 1.85 (–0.05 to 3.76) 0.08 (–1.95 to 2.11) 0.36 (–0.66 to 1.39) –0.24 (–0.73 to 0.25) 0.08 (–0.35 to 0.51) –0.06 (–0.64 to 0.52) 
 Intra-individualb 1.55 (–0.81 to 3.91) –0.37 (–2.73 to 1.98) 0.18 (–1.02 to 1.38) –0.40 (–0.99 to 0.18) 0.13 (–0.40 to 0.66) –0.28 (–0.98 to 0.42) 
IL8 Inter-individualc 2.41 (–0.81 to 5.64) 1.40 (–2.62 to 5.43) 0.84 (–1.11 to 2.79) 0.14 (–0.75 to 1.03) –0.03 (–0.76 to 0.70) 0.44 (–0.59 to 1.47) 
 Model II       
 Overall associationa 3.34 (0.775.90) 0.74 (–2.10 to 3.57) 0.68 (–0.73 to 2.08) –0.18 (–0.84 to 0.49) –0.03 (–0.61 to 0.55) 0.37 (–0.44 to 1.18) 
 Intra-individualb 3.68 (0.367.01) 0.25 (–3.21 to 3.71) 0.54 (–1.20 to 2.28) –0.43 (–1.27 to 0.41) 0.07 (–0.69 to 0.83) 0.15 (–0.87 to 1.18) 
 Inter-individualc 3.01 (–0.24 to 6.26) 1.40 (–2.50 to 5.30) 0.84 (–1.06 to 2.75) 0.10 (–0.77 to 0.97) –0.12 (–0.84 to 0.60) 0.61 (–0.45 to 1.67) 
 Model I       
 Overall associationa –0.12 (–1.87 to 1.62) 0.62 (–1.16 to 2.41) 0.12 (–0.79 to 1.03) 0.13 (–0.31 to 0.57) 0.53 (0.140.92) –0.33 (–0.85 to 0.19) 
 Intra-individualb 0.65 (–1.25 to 2.55) 1.41 (–0.48 to 3.31) 0.47 (–0.50 to 1.44) 0.32 (–0.16 to 0.79) 0.77 (0.351.19)* –0.14 (–0.70 to 0.42) 
IL10 Inter-individualc –4.31 (–8.69 to 0.06) 5.83 (11.20 to –0.44) –2.45 (–5.06 to 0.17) –1.04 (–2.24 to 0.15) –0.77 (–1.76 to 0.22) 1.49 (2.89 to –0.10) 
 Model II       
 Overall associationa –0.07 (–1.80 to 1.66) 0.77 (–1.01 to 2.54) 0.20 (–0.70 to 1.10) 0.14 (–0.30 to 0.57) 0.08 (–0.31 to 0.48) –0.29 (–0.82 to 0.24) 
 Intra-individualb 0.76 (–1.09 to 2.61) 1.53 (–0.34 to 3.39) 0.54 (–0.40 to 1.49) 0.33 (–0.13 to 0.79) 0.80 (0.031.57) –0.13 (–0.69 to 0.44) 
 Inter-individualc 5.90 (10.74 to1.06) 6.71 (12.51 to –0.92) 2.95 (5.79 to –0.11) 1.40 (2.69 to –0.11) –0.81 (–1.87 to 0.25) 1.57 (3.14 to –0.00) 
 Model I       
 Overall associationa 0.32 (–1.90 to 2.54) 0.43 (–1.89 to 2.75) 0.44 (–0.74 to 1.61) 0.39 (–0.17 to 0.95) 0.07 (–0.42 to 0.57) –0.41 (–1.08 to 0.25) 
 Intra-individualb –0.05 (–2.61 to 2.52) 0.17 (–2.39 to 2.73) 0.39 (–0.92 to 1.70) 0.27 (–0.37 to 0.91) –0.01 (–0.59 to 0.56) –0.47 (–1.23 to 0.29) 
TNFα Inter-individualc 1.39 (–2.97 to 5.76) 1.58 (–3.85 to 7.0) 0.63 (–2.00 to 3.25) 0.81 (–0.38 to 2.01) 0.33 (–0.65 to 1.31) –0.23 (–1.62 to 1.16) 
 Model II       
 Overall associationa –0.34 (–2.52 to 1.85) –0.13 (–2.43 to 2.16) 0.16 (–0.99 to 1.31) 0.23 (–0.33 to 0.78) –0.08 (–0.58 to 0.41) –0.45 (–1.13 to 0.22) 
 Intra-individualb 0.10 (–2.45 to 2.65) 0.20 (–2.37 to 2.78) 0.44 (–0.86 to 1.74) 0.28 (–0.36 to 0.91) –0.05 (–0.64 to 0.53) –0.42 (–1.19 to 0.36) 
 Inter-individualc –1.56 (–5.84 to 2.72) –1.45 (–6.57 to 3.67) –0.86 (–3.36 to 1.64) 0.07 (–1.07 to 1.20) –0.16 (–1.09 to 0.77) –0.56 (–1.94 to 0.82) 
 Model I       
 Overall associationa 2.36 (0.474.25) 1.13 (–0.69 to 2.95) 0.73 (–0.19 to 1.65) 0.25 (–0.19 to 0.69) 0.03 (–0.36 to 0.42) 0.33 (–0.19 to 0.85) 
 Intra-individualb 1.71 (–0.47 to 3.89) –0.07 (–2.11 to 1.96) 0.15 (–0.89 to 1.19) –0.05 (–0.56 to 0.46) –0.22 (–0.68 to 0.24) 0.05 (–0.56 to 0.65) 
IZ excluding Inter-individualc 4.28 (0.538.04) 5.89 (1.869.92) 2.76 (0.814.71) 1.18 (0.292.07) 0.68 (–0.06 to 1.41) 1.17 (0.132.21) 
hsCRPd Model II       
 Overall associationa 2.29 (0.344.24) 0.74 (–1.13 to 2.60) 0.42 (–0.52 to 1.36) 0.20 (–0.25 to 0.65) –0.16 (–0.56 to 0.24) 0.40 (–0.15 to 0.95) 
 Intra-individualb 2.01 (–0.23 to 4.24) –0.10 (–2.19 to 2.0) 0.01 (–1.04; 1.07) –0.01 (–0.53 to 0.50) –0.31 (–0.79 to 0.16) 0.23 (–0.40 to 0.86) 
 Inter-individualc 3.14 (–0.63 to 6.91) 3.76 (–0.15 to 7.67) 1.79 (–0.13 to 3.70) 0.82 (–0.05 to 1.69) 0.19 (–0.53 to 0.91) 0.89 (–0.17 to 1.95) 
 Model I       
 Overall associationa 3.11 (1.554.68)* 3.20 (1.634.77)* 1.75 (0.952.54)* 0.98 (0.581.37)* 0.56 (0.210.90)* 0.14 (–0.34 to 0.62) 
 Intra-individualb 2.37 (0.604.13) 2.22 (0.513.93) 1.33 (0.462.20)* 0.77 (0.331.22)* 0.37 (–0.03 to 0.76) –0.25 (–0.79 to 0.30) 
hsCRPe Inter-individualc 5.68 (2.458.92)* 8.25 (4.4012.11)* 3.76 (1.865.66)* 1.73 (0.892.58)* 1.18 (0.471.90)* 1.59 (0.552.62)* 
 Model II       
 Overall associationa 2.41 (0.724.10)* 3.21 (1.425.01)* 1.82 (0.942.70)* 0.85 (0.411.29)* 0.38 (–0.01 to 0.76) 0.24 (–0.31 to 0.78) 
 Intra-individualb 2.08 (0.124.04) 2.57 (0.534.62) 1.62 (0.622.62)* 0.73 (0.221.25)* 0.26 (–0.20 to 0.72) –0.08 (–0.72 to 0.57) 
 Inter-individualc 3.38 (0.046.73) 5.44 (1.619.27)* 2.53 (0.634.43)* 1.15 (0.321.99)* 0.65 (–0.07 to 1.37) 1.10 (0.032.17) 
 Model I       
 Overall associationa 5.24 (2.887.60) 2.82 (0.664.98) 1.72 (0.632.80) 0.84 (0.311.38) 0.35 (–0.11 to 0.81) 0.33 (–0.31 to 0.98) 
 Intra-individualb 4.06 (1.266.87) 0.67 (–1.79 to 3.13) 0.78 (–0.47 to 2.03) 0.30 (–0.35 to 0.94) –0.06 (–0.62 to 0.50) –0.35 (–1.12 to 0.42) 
IZ including Inter-individualc 8.01 (3.7212.29) 9.71 (5.3514.07) 4.47 (2.336.61) 2.03 (1.092.97) 1.20 (0.392.00) 1.90 (0.743.06) 
hsCRPd,f Model II       
 Overall associationa 4.49 (1.977.01) 2.42 (0.064.79) 1.40 (0.242.57) 0.72 (0.161.29) 0.04 (–0.46 to 0.53) 0.46 (–0.24 to 1.17) 
 Intra-individualb 3.92 (0.936.91) 0.69 (–2.07 to 3.45) 0.68 (–0.68 to 2.04) 0.30 (–0.40 to 1.00) –0.27 (–0.89 to 0.36) –0.01 (–0.87 to 0.85) 
 Inter-individualc 5.75 (1.3910.12) 6.71 (2.4311.00) 3.22 (1.095.35) 1.47 (0.552.39) 0.53 (–0.26 to 1.33) 1.39 (0.202.58) 
EORTC QLQ-C30, Fatigue β (95% CI)CIS Total fatigue β (95% CI)Subjective fatigue β (95% CI)Reduced motivation β (95% CI)Reduced physical activity β (95% CI)Reduced concentration β (95% CI)
 Model I       
 Overall associationa 1.85 (0.233.47) 2.07 (0.383.76) 0.97 (0.121.83) 0.49 (0.070.90) 0.23 (–0.13 to 0.60) 0.48 (–0.00 to 0.97) 
 Intra-individualb 1.50 (–0.35 to 3.35) 1.40 (–0.44 to 3.24) 0.60 (–0.34 to 1.54) 0.35 (–0.11 to 0.82) 0.08 (–0.34 to 0.50) 0.37 (–0.18 to 0.91) 
IL6 Inter-individualc 3.04 (–0.35 to 6.44) 5.44 (1.319.57) 2.67 (0.664.67)* 1.00 (0.091.92) 0.74 (–0.01 to 1.50) 0.94 (–0.13 to 2.01) 
 Model II       
 Overall associationa 1.69 (0.083.32) 1.67 (–0.03 to 3.36) 0.71 (–0.14 to 1.56) 0.36 (–0.05 to 0.77) 0.12 (–0.24 to 0.49) 0.48 (–0.02 to 0.98) 
 Intra-individualb 1.74 (–0.11 to 3.60) 1.28 (–0.60 to 3.16) 0.46 (–0.49 to 1.41) 0.30 (–0.16 to 0.77) 0.05 (–0.37 to 0.48) 0.44 (–0.13 to 1.01) 
 Inter-individualc 1.56 (–1.74 to 4.86) 3.34 (–0.56 to 7.24) 1.71 (–0.19 to 3.61) 0.58 (–0.29 to 1.45) 0.32 (–0.40 to 1.03) 0.62 (–0.44 to 1.68) 
 Model I       
 Overall associationa 1.85 (–0.05 to 3.76) 0.08 (–1.95 to 2.11) 0.36 (–0.66 to 1.39) –0.24 (–0.73 to 0.25) 0.08 (–0.35 to 0.51) –0.06 (–0.64 to 0.52) 
 Intra-individualb 1.55 (–0.81 to 3.91) –0.37 (–2.73 to 1.98) 0.18 (–1.02 to 1.38) –0.40 (–0.99 to 0.18) 0.13 (–0.40 to 0.66) –0.28 (–0.98 to 0.42) 
IL8 Inter-individualc 2.41 (–0.81 to 5.64) 1.40 (–2.62 to 5.43) 0.84 (–1.11 to 2.79) 0.14 (–0.75 to 1.03) –0.03 (–0.76 to 0.70) 0.44 (–0.59 to 1.47) 
 Model II       
 Overall associationa 3.34 (0.775.90) 0.74 (–2.10 to 3.57) 0.68 (–0.73 to 2.08) –0.18 (–0.84 to 0.49) –0.03 (–0.61 to 0.55) 0.37 (–0.44 to 1.18) 
 Intra-individualb 3.68 (0.367.01) 0.25 (–3.21 to 3.71) 0.54 (–1.20 to 2.28) –0.43 (–1.27 to 0.41) 0.07 (–0.69 to 0.83) 0.15 (–0.87 to 1.18) 
 Inter-individualc 3.01 (–0.24 to 6.26) 1.40 (–2.50 to 5.30) 0.84 (–1.06 to 2.75) 0.10 (–0.77 to 0.97) –0.12 (–0.84 to 0.60) 0.61 (–0.45 to 1.67) 
 Model I       
 Overall associationa –0.12 (–1.87 to 1.62) 0.62 (–1.16 to 2.41) 0.12 (–0.79 to 1.03) 0.13 (–0.31 to 0.57) 0.53 (0.140.92) –0.33 (–0.85 to 0.19) 
 Intra-individualb 0.65 (–1.25 to 2.55) 1.41 (–0.48 to 3.31) 0.47 (–0.50 to 1.44) 0.32 (–0.16 to 0.79) 0.77 (0.351.19)* –0.14 (–0.70 to 0.42) 
IL10 Inter-individualc –4.31 (–8.69 to 0.06) 5.83 (11.20 to –0.44) –2.45 (–5.06 to 0.17) –1.04 (–2.24 to 0.15) –0.77 (–1.76 to 0.22) 1.49 (2.89 to –0.10) 
 Model II       
 Overall associationa –0.07 (–1.80 to 1.66) 0.77 (–1.01 to 2.54) 0.20 (–0.70 to 1.10) 0.14 (–0.30 to 0.57) 0.08 (–0.31 to 0.48) –0.29 (–0.82 to 0.24) 
 Intra-individualb 0.76 (–1.09 to 2.61) 1.53 (–0.34 to 3.39) 0.54 (–0.40 to 1.49) 0.33 (–0.13 to 0.79) 0.80 (0.031.57) –0.13 (–0.69 to 0.44) 
 Inter-individualc 5.90 (10.74 to1.06) 6.71 (12.51 to –0.92) 2.95 (5.79 to –0.11) 1.40 (2.69 to –0.11) –0.81 (–1.87 to 0.25) 1.57 (3.14 to –0.00) 
 Model I       
 Overall associationa 0.32 (–1.90 to 2.54) 0.43 (–1.89 to 2.75) 0.44 (–0.74 to 1.61) 0.39 (–0.17 to 0.95) 0.07 (–0.42 to 0.57) –0.41 (–1.08 to 0.25) 
 Intra-individualb –0.05 (–2.61 to 2.52) 0.17 (–2.39 to 2.73) 0.39 (–0.92 to 1.70) 0.27 (–0.37 to 0.91) –0.01 (–0.59 to 0.56) –0.47 (–1.23 to 0.29) 
TNFα Inter-individualc 1.39 (–2.97 to 5.76) 1.58 (–3.85 to 7.0) 0.63 (–2.00 to 3.25) 0.81 (–0.38 to 2.01) 0.33 (–0.65 to 1.31) –0.23 (–1.62 to 1.16) 
 Model II       
 Overall associationa –0.34 (–2.52 to 1.85) –0.13 (–2.43 to 2.16) 0.16 (–0.99 to 1.31) 0.23 (–0.33 to 0.78) –0.08 (–0.58 to 0.41) –0.45 (–1.13 to 0.22) 
 Intra-individualb 0.10 (–2.45 to 2.65) 0.20 (–2.37 to 2.78) 0.44 (–0.86 to 1.74) 0.28 (–0.36 to 0.91) –0.05 (–0.64 to 0.53) –0.42 (–1.19 to 0.36) 
 Inter-individualc –1.56 (–5.84 to 2.72) –1.45 (–6.57 to 3.67) –0.86 (–3.36 to 1.64) 0.07 (–1.07 to 1.20) –0.16 (–1.09 to 0.77) –0.56 (–1.94 to 0.82) 
 Model I       
 Overall associationa 2.36 (0.474.25) 1.13 (–0.69 to 2.95) 0.73 (–0.19 to 1.65) 0.25 (–0.19 to 0.69) 0.03 (–0.36 to 0.42) 0.33 (–0.19 to 0.85) 
 Intra-individualb 1.71 (–0.47 to 3.89) –0.07 (–2.11 to 1.96) 0.15 (–0.89 to 1.19) –0.05 (–0.56 to 0.46) –0.22 (–0.68 to 0.24) 0.05 (–0.56 to 0.65) 
IZ excluding Inter-individualc 4.28 (0.538.04) 5.89 (1.869.92) 2.76 (0.814.71) 1.18 (0.292.07) 0.68 (–0.06 to 1.41) 1.17 (0.132.21) 
hsCRPd Model II       
 Overall associationa 2.29 (0.344.24) 0.74 (–1.13 to 2.60) 0.42 (–0.52 to 1.36) 0.20 (–0.25 to 0.65) –0.16 (–0.56 to 0.24) 0.40 (–0.15 to 0.95) 
 Intra-individualb 2.01 (–0.23 to 4.24) –0.10 (–2.19 to 2.0) 0.01 (–1.04; 1.07) –0.01 (–0.53 to 0.50) –0.31 (–0.79 to 0.16) 0.23 (–0.40 to 0.86) 
 Inter-individualc 3.14 (–0.63 to 6.91) 3.76 (–0.15 to 7.67) 1.79 (–0.13 to 3.70) 0.82 (–0.05 to 1.69) 0.19 (–0.53 to 0.91) 0.89 (–0.17 to 1.95) 
 Model I       
 Overall associationa 3.11 (1.554.68)* 3.20 (1.634.77)* 1.75 (0.952.54)* 0.98 (0.581.37)* 0.56 (0.210.90)* 0.14 (–0.34 to 0.62) 
 Intra-individualb 2.37 (0.604.13) 2.22 (0.513.93) 1.33 (0.462.20)* 0.77 (0.331.22)* 0.37 (–0.03 to 0.76) –0.25 (–0.79 to 0.30) 
hsCRPe Inter-individualc 5.68 (2.458.92)* 8.25 (4.4012.11)* 3.76 (1.865.66)* 1.73 (0.892.58)* 1.18 (0.471.90)* 1.59 (0.552.62)* 
 Model II       
 Overall associationa 2.41 (0.724.10)* 3.21 (1.425.01)* 1.82 (0.942.70)* 0.85 (0.411.29)* 0.38 (–0.01 to 0.76) 0.24 (–0.31 to 0.78) 
 Intra-individualb 2.08 (0.124.04) 2.57 (0.534.62) 1.62 (0.622.62)* 0.73 (0.221.25)* 0.26 (–0.20 to 0.72) –0.08 (–0.72 to 0.57) 
 Inter-individualc 3.38 (0.046.73) 5.44 (1.619.27)* 2.53 (0.634.43)* 1.15 (0.321.99)* 0.65 (–0.07 to 1.37) 1.10 (0.032.17) 
 Model I       
 Overall associationa 5.24 (2.887.60) 2.82 (0.664.98) 1.72 (0.632.80) 0.84 (0.311.38) 0.35 (–0.11 to 0.81) 0.33 (–0.31 to 0.98) 
 Intra-individualb 4.06 (1.266.87) 0.67 (–1.79 to 3.13) 0.78 (–0.47 to 2.03) 0.30 (–0.35 to 0.94) –0.06 (–0.62 to 0.50) –0.35 (–1.12 to 0.42) 
IZ including Inter-individualc 8.01 (3.7212.29) 9.71 (5.3514.07) 4.47 (2.336.61) 2.03 (1.092.97) 1.20 (0.392.00) 1.90 (0.743.06) 
hsCRPd,f Model II       
 Overall associationa 4.49 (1.977.01) 2.42 (0.064.79) 1.40 (0.242.57) 0.72 (0.161.29) 0.04 (–0.46 to 0.53) 0.46 (–0.24 to 1.17) 
 Intra-individualb 3.92 (0.936.91) 0.69 (–2.07 to 3.45) 0.68 (–0.68 to 2.04) 0.30 (–0.40 to 1.00) –0.27 (–0.89 to 0.36) –0.01 (–0.87 to 0.85) 
 Inter-individualc 5.75 (1.3910.12) 6.71 (2.4311.00) 3.22 (1.095.35) 1.47 (0.552.39) 0.53 (–0.26 to 1.33) 1.39 (0.202.58) 

Note: Levels of inflammatory markers were divided by their SD at 6 weeks (IL6: SD = 3.15, IL8: SD = 19.57, IL10: SD = 0.86, TNFα: SD = 3.14, IZ excluding hsCRP: SD = 2.08, CRP: SD = 8.19, IZ including hsCRP: SD = 2.62). Model I: adjusted for age (years), sex (men/women), time since diagnosis (days). Model II: adjusted for age (years), sex (men/women), time since diagnosis (days), use of NSAIDs (yes/no), BMI (kg/m2), light physical activity (hours/week), MVPA (hours/week), comorbidities (0, 1, ≥2), chemotherapy (yes/no), radiotherapy (yes/no), and smoking status (yes/no). Asterisk (*) represents a significant association after FDR correction for multiple testing.

Abbreviations: β, beta-coefficient; IZ, summary inflammatory z-score.

aThe beta-coefficient represents the overall longitudinal difference in fatigue score per SD difference of the inflammatory marker. It is a weighted average of the intra- and inter-individual associations.

bThe beta-coefficient represents the change in fatigue score over time within-individuals per SD increase of the inflammatory marker.

cThe beta-coefficient represents the difference in fatigue score between-individuals per SD difference of the inflammatory marker.

dFDR adjustment for multiple testing not performed.

ehsCRP was measured at 6 weeks, 6 and 12 months posttreatment.

fAnalysis includes patients with data available at 6 weeks, 6 and 12 months posttreatment.

After fully adjusting the model and FDR correction, higher levels of hsCRP were longitudinally associated with more CIS total fatigue (β 3.21; 95% confidence interval (CI), 1.42–5.01), subjective fatigue (β 1.82; 95% CI, 0.94–2.70), reduced motivation (β 0.85; 95% CI, 0.41–1.29), and EORTC fatigue (β 2.41; 95% CI, 0.72–4.10). Applying hybrid models revealed a significant inter-individual association between hsCRP and CIS total fatigue (β 5.44; 95% CI, 1.61–9.27). In addition, higher levels of hsCRP were longitudinally associated with higher scores, both between- and within-subjects, in the subjective fatigue and reduced motivation subscales. The sensitivity analyses indicate that associations were similar after excluding participants who had a recurrence or died (Supplementary Table S1).

Analyses with the summary inflammatory z-score including hsCRP indicated that more inflammation was associated with more CIS total fatigue (β 2.42; 95% CI, 0.06–4.79) and EORTC fatigue (β 4.49; 95% CI, 1.97–7.01). For CIS total fatigue, an inter-individual association was observed (β 6.71; 95% CI, 2.43–11.00) while the intra-individual association was small and nonsignificant (β 0.69; 95% CI, –2.07 to 3.45). In the analyses with EORTC fatigue, both the inter-individual (β 5.75; 95% CI, 1.39–10.12) and intra-individual (β 3.92; 95% CI, 0.93–6.91) associations were statistically significant. The summary inflammatory z-score excluding hsCRP was associated with more EORTC fatigue (β 2.29; 95% CI, 0.34–4.24) but not with CIS total fatigue (β 0.74; 95% CI, –1.13 to 2.60). To ensure the 24-month time point was not responsible for the different results between the inflammatory z-scores including and excluding hsCRP, extra analyses excluding the 24-month time point were performed for the inflammatory z-score excluding hsCRP. The results led to the same conclusions with similar effect sizes between the inflammatory score excluding hsCRP and fatigue in which all time points were considered.

Results from the exploratory analysis indicated that survivors with levels of hsCRP between 3 to 10 mg/L, and levels >10 mg/L experienced more fatigue compared with those with levels ≤3 mg/L (Fig. 3; Supplementary Table S2). In subgroup analysis, the statistically significant associations after FDR correction for the individual inflammatory markers were only observed in men (Supplementary Table S3). In addition, only 5 of 84 interaction terms were statistically significant.

Figure 3.

Overall longitudinal associations between levels of hsCRP increments (≤3 mg/L, 3–10 mg/L, and >10 mg/L) with CIS total fatigue (A), EORTC QLQ-C30 fatigue (B), and the CIS subscales – subjective fatigue (C), reduced motivation (D), reduced physical activity level (E), reduced concentration (F), in colorectal cancer survivors followed-up at 6 weeks, 6, and 12 months after treatment. CIS ranges: total fatigue, 20–140; subjective fatigue, 8–56; motivation, 4–28; physical activity level, 3–21; concentration, 5–35. EORTC QLQ-C30 ranges from 0 to 100.

Figure 3.

Overall longitudinal associations between levels of hsCRP increments (≤3 mg/L, 3–10 mg/L, and >10 mg/L) with CIS total fatigue (A), EORTC QLQ-C30 fatigue (B), and the CIS subscales – subjective fatigue (C), reduced motivation (D), reduced physical activity level (E), reduced concentration (F), in colorectal cancer survivors followed-up at 6 weeks, 6, and 12 months after treatment. CIS ranges: total fatigue, 20–140; subjective fatigue, 8–56; motivation, 4–28; physical activity level, 3–21; concentration, 5–35. EORTC QLQ-C30 ranges from 0 to 100.

Close modal

No statistically significant associations were found between IL6, IL8, IL10, TNFα, and CIS and EORTC fatigue after FDR correction. Higher levels of hsCRP were longitudinally associated with more fatigue from 6 weeks to 12 months posttreatment. Statistically significant inter-individual associations were observed, indicating that colorectal cancer survivors with higher mean levels of hsCRP over time reported higher scores of total fatigue. Similar trends were observed in the subjective and reduced motivation subscales, where both inter- and intra-individual associations for hsCRP were statistically significant. In addition, statistically significant associations were found between the summary inflammatory score including hsCRP and both CIS and EORTC fatigue. Together these findings suggest that higher levels of low-grade inflammation are associated with more fatigue in colorectal cancer survivors.

Findings from longitudinal colorectal cancer studies are scarce and inconsistent, the latter likely due to methodologic differences in the timing and frequency of measurements for the inflammatory markers and fatigue, the duration of follow-up time, and the types of measurement instruments used to assess the inflammatory markers and fatigue (22, 23). A recent study of 236 stage I to IV colorectal cancer survivors did not find statistically significant associations between levels of IL6, IL8, TNFα, CRP measured pre-surgery, and fatigue measured pre-surgery and at 6 and 12 months post-surgery (30). However, unlike this study, only preoperative inflammatory markers were used, and fatigue was only measured using the EORTC QLQ-C30 fatigue subscale, which mainly measures physical fatigue. A study in patients with localized colorectal cancer found weak correlations of IL6, IL8, and IL10 with fatigue at 6 (ρ, –0.16 to –0.20) and 24 months (ρ, –0.16 to –0.30) after treatment, but not with TNFα (9). From the inflammatory markers we investigated, excluding hsCRP which was not measured, only IL8 was longitudinally inversely associated with more fatigue. Another study on patients with colorectal cancer (n = 50) and esophageal cancer (n = 53) found a significant association between IL6 and a component score of fatigue-centered symptom cluster, but not between IL6, IL10, and fatigue severity (29). In the latter study, fatigue was measured weekly for 13 weeks after treatment initiated and the inflammatory markers were measured pretreatment, during the 5 to 6 weeks of treatment, and 1 month posttreatment. Because IL10 is considered to have anti-inflammatory properties, it was expected to be inversely associated with fatigue (49, 50). We observed an inverse association between patients, for both CIS and EORTC fatigue, but this was nonsignificant after FDR correction. Although evidence is still scarce and inconsistent, higher levels of pro-inflammatory markers seem to be associated with more fatigue in colorectal cancer survivors.

A cross-sectional study in 299 disease-free breast cancer survivors, at 4 years post-diagnosis on average, reported a significant association between levels of hsCRP and fatigue (20). Other inflammatory markers, such as IL6, were analyzed but no statistically significant associations were found. Similar results were found in a longitudinal study in breast (n = 28) and prostate cancer (n = 20) survivors during radiotherapy (51). Both studies argued that pro-inflammatory cytokines, such as IL6, are produced in low quantities and thus harder to detect, possibly explaining the lack of association, as seen in our study (20, 51).

No other studies have used summary inflammatory z-scores to assess the association between inflammatory markers and fatigue. In our study, a significant association was found between the inflammatory z-score excluding hsCRP and EORTC fatigue. This association was not observed in the analysis with CIS fatigue and this difference is possibly explained by the weaker association between IL8 and CIS fatigue compared with EORTC fatigue. Higher levels of the inflammatory z-score including hsCRP were statistically significantly associated with more fatigue. This association is likely driven by levels of hsCRP because the association between the inflammatory z-score excluding hsCRP and fatigue remained nonsignificant, and with similar effect sizes, after excluding the 24-month time point.

In summary, results from the main analyses add to the existing body of literature on inflammation and fatigue and suggest a link between hsCRP and fatigue. hsCRP can detect low CRP in the blood, and thus can be used to evaluate low-grade inflammation (34, 52). Low-grade inflammation can reduce cellular energy availability and increase energy expenditure, creating an imbalance, which possibly explains persistent fatigue (53). CRP is an acute-phase protein mainly upregulated by IL6, and therefore considered a downstream marker for IL6 activity (52). Other cytokines such as TNFα, IL1, IL1β are also involved in the production of acute-phase proteins (54, 55), and thus require further research as to whether they could be potential targets for intervention (52, 56–58).

In terms of clinical relevance, the observed effect sizes from the fully adjusted models were smaller than the minimal clinically important difference (MCID) defined as 9.3 points for CIS total fatigue (59) and 9 points for EORTC fatigue (60, 61). The largest effect sizes were observed in the analysis with categories of hsCRP where levels >10 mg/L were associated with a 6.14 point (95% CI, 0.10–12.19) increase in EORTC fatigue score, compared with levels ≤3 mg/L. Results from these analyses provide a better comparison with the MCID as the cut-off values chosen are more clinically relevant than the SD increments used in the main analysis (62, 63). Despite not reaching the MCID, the results provide evidence for a longitudinal association between higher levels of hsCRP and an increase in posttreatment fatigue in colorectal cancer survivors.

Results from subgroup analysis indicated that the association between hsCRP and fatigue was only present in men. However, this should be interpreted with caution as the analysis in men had twice the sample size as the women's analysis, rendering the associations in women less stable. Furthermore, most of the interaction terms were nonsignificant.

One of the strengths of this study was the availability of repeated measurements for both inflammatory markers and fatigue, as well as potential confounders. In addition, the use of hybrid models to disentangle between- and within-individual associations was important to understand how changes in inflammation within-individuals are, on average, related to fatigue over time. To date, this approach has not been attempted by any of the studies exploring an association between inflammatory markers and fatigue.

A limitation of the current study is its observational nature, which does not allow for any causal inference. Moreover, patients with higher levels of fatigue at time of diagnosis may view the measurements involved (i.e., filling out questionnaires and blood collection) as being too burdensome. Thus, patients with higher levels of fatigue could be underrepresented in the study, in part explaining the 45% participation rate at diagnosis and potentially causing an underestimation of the true association. Although the participation rate at diagnosis was 45%, our interest was in the association between inflammation and fatigue, specifically in the posttreatment phase, and all follow-up participation rates were high (≥90%). The decrease in sample size as follow-up time increased was mainly due to patients not reaching those time points at the time of data-freeze. Therefore, most participants with missing data are likely missing at random. The smaller sample sizes decrease the power to detect true associations and provide less information on the long-term posttreatment associations of inflammation and fatigue. In addition, to minimize the potential impact of time of sampling on hsCRP values, which exhibits diurnal variations (64, 65), all samples were collected in fasting individuals during the morning period before breakfast after an overnight fast.

In conclusion, the current study found that higher levels of hsCRP were longitudinally associated with more fatigue in colorectal cancer survivors up to 12 months posttreatment. Further longitudinal studies with larger sample sizes will help provide stronger evidence on the long-term association between low-grade inflammation and fatigue posttreatment.

E.H. van Roekel reports grants from Wereld Kanker Onderzoek Fonds (WKOF), as part of the World Cancer Research Fund International during the conduct of the study. P.M. Ueland is a member of the steering board of the nonprofit Foundation, which owns Bevital, and R&D director of Bevital, the company that carried out biochemical analyses. No other potential conflict of interest relevant to this article was reported. M.P. Weijenberg reports grants from Dutch Cancer Society, World Cancer Research Fund International/WKOF; and grants from Health Foundation Limburg/Kanker Onderzoek Fonds Limburg (KOFL) during the conduct of the study. No disclosures were reported by the other authors.

N.R. Querido: Conceptualization, formal analysis, methodology, writing–original draft. M.F. Kenkhuis: Formal analysis, investigation, writing–review and editing. E.H. van Roekel: Conceptualization, investigation, writing–review and editing. S.O. Breukink: Writing–review and editing. F.J.B. van Duijnhoven: Funding acquisition, writing–review and editing. M.L.G. Janssen-Heijnen: Writing–review and editing. E.T.P. Keulen: Writing–review and editing. P.M. Ueland: Writing–review and editing. F.J. Vogelaar: Writing–review and editing. E. Wesselink: Writing–review and editing. M.J.L. Bours: Conceptualization, supervision, funding acquisition, methodology, writing–review and editing. M.P. Weijenberg: Conceptualization, supervision, funding acquisition, methodology, writing–original draft.

The EnCoRe study was supported by a grant from Kankeronderzoekfonds Limburg as part of Health Foundation Limburg (grant 00005739, to M.P. Weijenberg), from Stichting Alpe d'Huzes within the research program “Leven met kanker” of the Dutch Cancer Society grants UM 2010-4867 (to M.P. Weijenberg), and UM 2012-5653 (to M.P. Weijenberg), and by ERA-NET on Translational Cancer Research [TRANSCAN: Dutch Cancer Society (UM 2014–6877, to M.P. Weijenberg)]. The measurement of inflammatory markers was funded by a grant from Alpe d'Huzes/ Dutch Cancer Society (UW2013-6397, to F.J.B. van Duijnhoven). M.F. Kenkhuis is supported by a grant from WKOF/World Cancer Research Fund International (WCRF) (grant number 2017/1619, to M.J.L. Bours). E.H. van Roekel is supported by the WKOF, as part of the World Cancer Research Fund International grant program (grant number 2016/1620, to M.P. Weijenberg).

We would like to thank all participants of the EnCoRe study and the health professionals in the three hospitals involved in the recruitment of participants of the study: Maastricht University Medical Centre+, VieCuri Medical Centre, and Zuyderland Medical Centre. We would also like to thank the MEMIC centre for data and information management for facilitating the logistic processes and data management of our study. We thank Michiel Balvers and Nhien Ly at Wageningen University & Research for their work on the inflammation markers. Finally, we would like to thank the research dietitians and research assistant who are responsible for patient inclusion and follow-up, performing home visits, as well as data collection and processing.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Ait Ouakrim D,
Pizot
C
,
Boniol
M
,
Malvezzi
M
,
Boniol
M
,
Negri
E
, et al
.
Trends in colorectal cancer mortality in Europe: retrospective analysis of the WHO mortality database
.
BMJ
2015
;
351
:
h4970
.
2.
IARC
WHO
.
Colorectal cancer fact sheet 2020
.
Available from
: https://gco.iarc.fr/today/data/factsheets/cancers/10_8_9-Colorectum-fact-sheet.pdf.
3.
Bours
MJ
,
van der Linden
BW
,
Winkels
RM
,
van Duijnhoven
FJ
,
Mols
F
,
van Roekel
EH
, et al
.
Candidate predictors of health-related quality of life of colorectal cancer survivors: a systematic review
.
Oncologist
2016
;
21
:
433
52
.
4.
Marventano
S
,
Forjaz
M
,
Grosso
G
,
Mistretta
A
,
Giorgianni
G
,
Platania
A
, et al
.
Health related quality of life in colorectal cancer patients: state of the art
.
BMC Surg
2013
;
13
Suppl 2
:
S15
.
5.
Aapro
M
,
Scotte
F
,
Bouillet
T
,
Currow
D
,
Vigano
A
.
A practical approach to fatigue management in colorectal cancer
.
Clin Colorectal Cancer
2017
;
16
:
275
85
.
6.
Thong
MSY
,
Mols
F
,
Wang
XS
,
Lemmens
VEPP
,
Smilde
TJ
,
van de Poll-Franse
LV
.
Quantifying fatigue in (long-term) colorectal cancer survivors: a study from the population-based patient reported outcomes following initial treatment and long-term evaluation of survivorship registry
.
Eur J Cancer
2013
;
49
:
1957
66
.
7.
Wang
XS
,
Zhao
F
,
Fisch
MJ
,
O'Mara
AM
,
Cella
D
,
Mendoza
TR
, et al
.
Prevalence and characteristics of moderate to severe fatigue: a multicenter study in cancer patients and survivors
.
Cancer
2014
;
120
:
425
32
.
8.
Husson
O
,
Mols
F
,
van de Poll-Franse
LV
,
Thong
MSY
.
The course of fatigue and its correlates in colorectal cancer survivors: a prospective cohort study of the PROFILES registry
.
Support Care Cancer
2015
;
23
:
3361
71
.
9.
Vardy
JL
,
Dhillon
HM
,
Pond
GR
,
Renton
C
,
Dodd
A
,
Zhang
H
, et al
.
Fatigue in people with localized colorectal cancer who do and do not receive chemotherapy: a longitudinal prospective study
.
Ann Oncol
2016
;
27
:
1761
7
.
10.
van Baar
H
,
Bours
MJL
,
Beijer
S
,
van Zutphen
M
,
van Duijnhoven
FJB
,
Kok
DE
, et al
.
Body composition and its association with fatigue in the first 2 years after colorectal cancer diagnosis
.
J Cancer Surviv
2021
;
15
:
597
606
.
11.
van Roekel
EH
,
Duchateau
J
,
Bours
MJL
,
van Delden
L
,
Breedveld-Peters
JJL
,
Koole
JL
, et al
.
Longitudinal associations of light-intensity physical activity with quality of life, functioning and fatigue after colorectal cancer
.
Qual Life Res
2020
;
29
:
2987
98
.
12.
Rutherford
C
,
Müller
F
,
Faiz
N
,
King
MT
,
White
K
.
Patient-reported outcomes and experiences from the perspective of colorectal cancer survivors: meta-synthesis of qualitative studies
.
J Patient Rep Outcomes
2020
;
4
:
27
.
13.
Yang
S
,
Chu
S
,
Gao
Y
,
Ai
Q
,
Liu
Y
,
Li
X
, et al
.
A narrative review of cancer-related fatigue (CRF) and its possible pathogenesis
.
Cells
2019
;
8
:
738
.
14.
Bower
JE
.
Cancer-related fatigue–mechanisms, risk factors, and treatments
.
Nat Rev Clin Oncol
2014
;
11
:
597
609
.
15.
Bower
JE
,
Lamkin
DM
.
Inflammation and cancer-related fatigue: mechanisms, contributing factors, and treatment implications
.
Brain Behav Immun
2013
;
30
Suppl
:
S48
57
.
16.
LaVoy
ECP
,
Fagundes
CP
,
Dantzer
R
.
Exercise, inflammation, and fatigue in cancer survivors
.
Exerc Immunol Rev
2016
;
22
:
82
93
.
17.
Bower
JE
.
The role of neuro-immune interactions in cancer-related fatigue: biobehavioral risk factors and mechanisms
.
Cancer
2019
;
125
:
353
64
.
18.
Dantzer
R
,
O'Connor
JC
,
Freund
GG
,
Johnson
RW
,
Kelley
KW
.
From inflammation to sickness and depression: when the immune system subjugates the brain
.
Nat Rev Neurosci
2008
;
9
:
46
56
.
19.
Collado-Hidalgo
A
,
Bower
JE
,
Ganz
PA
,
Cole
SW
,
Irwin
MR
.
Inflammatory biomarkers for persistent fatigue in breast cancer survivors
.
Clin Cancer Res
2006
;
12
:
2759
66
.
20.
Orre
IJ
,
Reinertsen
KV
,
Aukrust
P
,
Dahl
AA
,
Fosså
SD
,
Ueland
T
, et al
.
Higher levels of fatigue are associated with higher CRP levels in disease-free breast cancer survivors
.
J Psychosom Res
2011
;
71
:
136
41
.
21.
Dantzer
R
.
Cytokine-induced sickness behavior: a neuroimmune response to activation of innate immunity
.
Eur J Pharmacol
2004
;
500
:
399
411
.
22.
Saligan
LN
,
Kim
HS
.
A systematic review of the association between immunogenomic markers and cancer-related fatigue
.
Brain Behav Immun
2012
;
26
:
830
48
.
23.
Schubert
C
,
Hong
S
,
Natarajan
L
,
Mills
PJ
,
Dimsdale
JE
.
The association between fatigue and inflammatory marker levels in cancer patients: a quantitative review
.
Brain Behav Immun
2007
;
21
:
413
27
.
24.
Husain
AF
,
Stewart
K
,
Arseneault
R
,
Moineddin
R
,
Cellarius
V
,
Librach
SL
, et al
.
Women experience higher levels of fatigue than men at the end of life: a longitudinal home palliative care study
.
J Pain Symptom Manage
2007
;
33
:
389
97
.
25.
Klein
SL
,
Flanagan
KL
.
Sex differences in immune responses
.
Nat Rev Immunol
2016
;
16
:
626
38
.
26.
Lasselin
J
,
Lekander
M
,
Axelsson
J
,
Karshikoff
B
.
Sex differences in how inflammation affects behavior: what we can learn from experimental inflammatory models in humans
.
Front Neuroendocrinol
2018
;
50
:
91
106
.
27.
Valentine
RJ
,
McAuley
E
,
Vieira
VJ
,
Baynard
T
,
Hu
L
,
Evans
EM
, et al
.
Sex differences in the relationship between obesity, C-reactive protein, physical activity, depression, sleep quality, and fatigue in older adults
.
Brain Behav Immun
2009
;
23
:
643
8
.
28.
Wesselink
E
,
van Baar
H
,
van Zutphen
M
,
Tibosch
M
,
Kouwenhoven
EA
,
Keulen
ETP
, et al
.
Inflammation is a mediating factor in the association between lifestyle and fatigue in colorectal cancer patients
.
Cancers
2020
;
12
:
3701
.
29.
Wang
XS
,
Williams
LA
,
Krishnan
S
,
Liao
Z
,
Liu
P
,
Mao
L
, et al
.
Serum sTNF-R1, IL6, and the development of fatigue in patients with gastrointestinal cancer undergoing chemoradiation therapy
.
Brain Behav Immun
2012
;
26
:
699
705
.
30.
Himbert
C
,
Ose
J
,
Lin
T
,
Warby
CA
,
Gigic
B
,
Steindorf
K
, et al
.
Inflammation- and angiogenesis-related biomarkers are correlated with cancer-related fatigue in colorectal cancer patients: results from the ColoCare Study
.
Eur J Cancer Care
2019
;
28
:
e13055
.
31.
van Roekel
EH
,
Bours
MJ
,
de Brouwer
CP
,
Ten Napel
H
,
Sanduleanu
S
,
Beets
GL
, et al
.
The applicability of the international classification of functioning, disability, and health to study lifestyle and quality of life of colorectal cancer survivors
.
Cancer Epidemiol Biomarkers Prev
2014
;
23
:
1394
405
.
32.
Wesselink
E
,
Balvers
M
,
Bours
MJL
,
de Wilt
JHW
,
Witkamp
RF
,
van Baar
H
, et al
.
The association between circulating levels of vitamin D and inflammatory markers in the first 2 years after colorectal cancer diagnosis
.
Therap Adv Gastroenterol
2020
;
13
:
1756284820923922
.
33.
Meyer
K
,
Ueland
PM
.
Targeted quantification of C-reactive protein and cystatin c and its variants by immuno-MALDI-MS
.
Anal Chem
2014
;
86
:
5807
14
.
34.
Dinh
KM
,
Kaspersen
KA
,
Mikkelsen
S
,
Pedersen
OB
,
Petersen
MS
,
Thørner
LW
, et al
.
Low-grade inflammation is negatively associated with physical health-related quality of life in healthy individuals: results from the danish blood donor study (DBDS)
.
PLoS One
2019
;
14
:
e0214468
.
35.
Kushner
I
,
Samols
D
,
Magrey
M
.
A unifying biologic explanation for "high-sensitivity" C-reactive protein and "low-grade" inflammation
.
Arthritis Care Res
2010
;
62
:
442
6
.
36.
Hopkins
MH
,
Owen
J
,
Ahearn
T
,
Fedirko
V
,
Flanders
WD
,
Jones
DP
, et al
.
Effects of supplemental vitamin D and calcium on biomarkers of inflammation in colorectal adenoma patients: a randomized, controlled clinical trial
.
Cancer Prev Res
2011
;
4
:
1645
54
.
37.
Vercoulen
JH
,
Swanink
CM
,
Fennis
JF
,
Galama
JM
,
van der Meer
JW
,
Bleijenberg
G
.
Dimensional assessment of chronic fatigue syndrome
.
J Psychosom Res
1994
;
38
:
383
92
.
38.
Aaronson
NK
,
Ahmedzai
S
,
Bergman
B
,
Bullinger
M
,
Cull
A
,
Duez
NJ
, et al
.
The European Organisation for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology
.
J Natl Cancer Inst
1993
;
85
:
365
76
.
39.
Koole
JL
,
Bours
MJL
,
van Roekel
EH
,
Breedveld-Peters
JJL
,
van Duijnhoven
FJB
,
van den Ouweland
J
, et al
.
Higher serum vitamin D concentrations are longitudinally associated with better global quality of life and less fatigue in colorectal cancer survivors up to 2 years after treatment
.
Cancer Epidemiol Biomarkers Prev
2020
;
29
:
1135
44
.
40.
Beurskens
AJ
,
Bültmann
U
,
Kant
I
,
Vercoulen
JH
,
Bleijenberg
G
,
Swaen
GM
.
Fatigue among working people: validity of a questionnaire measure
.
Occup Environ Med
2000
;
57
:
353
7
.
41.
Worm-Smeitink
M
,
Gielissen
M
,
Bloot
L
,
van Laarhoven
HWM
,
van Engelen
BGM
,
van Riel
P
, et al
.
The assessment of fatigue: Psychometric qualities and norms for the Checklist individual strength
.
J Psychosom Res
2017
;
98
:
40
6
.
42.
Sangha
O
,
Stucki
G
,
Liang
MH
,
Fossel
AH
,
Katz
JN
.
The self-administered comorbidity questionnaire: a new method to assess comorbidity for clinical and health services research
.
Arthritis Rheum
2003
;
49
:
156
63
.
43.
Wendel-Vos
GC
,
Schuit
AJ
,
Saris
WH
,
Kromhout
D
.
Reproducibility and relative validity of the short questionnaire to assess health-enhancing physical activity
.
J Clin Epidemiol
2003
;
56
:
1163
9
.
44.
Ainsworth
BE
,
Haskell
WL
,
Leon
AS
,
Jacobs
DR
, Jr.,
Montoye
HJ
,
Sallis
JF
, et al
.
Compendium of physical activities: classification of energy costs of human physical activities
.
Med Sci Sports Exerc
1993
;
25
:
71
80
.
45.
Twisk
JWR
.
Applied Longitudinal Data Analysis for Epidemiology: A Practical Guide
:
Cambridge University Press
;
2013
.
46.
Twisk
JWR
,
de Vente
W
.
Hybrid models were found to be very elegant to disentangle longitudinal within- and between-subject relationships
.
J Clin Epidemiol
2019
;
107
:
66
70
.
47.
Benjamini
Y
,
Hochberg
Y
.
Controlling the false discovery rate: a practical and powerful approach to multiple testing
.
Journal of the Royal Statistical Society Series B (Methodological)
1995
;
57
:
289
300
.
48.
Imhof
A
,
Fröhlich
M
,
Loewel
H
,
Helbecque
N
,
Woodward
M
,
Amouyel
P
, et al
.
Distributions of C-reactive protein measured by high-sensitivity assays in apparently healthy men and women from different populations in Europe
.
Clin Chem
2003
;
49
:
669
72
.
49.
Couper
KN
,
Blount
DG
,
Riley
EM
.
IL10: the master regulator of immunity to infection
.
J Immunol
2008
;
180
:
5771
7
.
50.
Iyer
SS
,
Cheng
G
.
Role of interleukin 10 transcriptional regulation in inflammation and autoimmune disease
.
Crit Rev Immunol
2012
;
32
:
23
63
.
51.
Bower
JE
,
Ganz
PA
,
Tao
ML
,
Hu
W
,
Belin
TR
,
Sepah
S
, et al
.
Inflammatory biomarkers and fatigue during radiation therapy for breast and prostate cancer
.
Clin Cancer Res
2009
;
15
:
5534
40
.
52.
Pepys
MB
,
Hirschfield
GM
.
C-reactive protein: a critical update
.
J Clin Invest
2003
;
111
:
1805
12
.
53.
Lacourt
TE
,
Vichaya
EG
,
Chiu
GS
,
Dantzer
R
,
Heijnen
CJ
.
The high costs of low-grade inflammation: persistent fatigue as a consequence of reduced cellular-energy availability and nonadaptive energy expenditure
.
Front Behav Neurosci
2018
;
12
:
78
-.
54.
Ansar
W
,
Ghosh
S
.
Inflammation and inflammatory diseases, markers, and mediators: role of CRP in some inflammatory diseases
. In:
Biology of C-Reactive Protein in Health and Disease
2016
,
New Delhi
:
Springer India
; p.
67
107
.
55.
Ferrucci
L
,
Ble
A
,
Bandinelli
S
,
Lauretani
F
,
Suthers
K
,
Guralnik
JM
.
A flame burning within
.
Aging Clin Exp Res
2004
;
16
:
240
3
.
56.
Sproston
NR
,
Ashworth
JJ
.
Role of C-reactive protein at sites of inflammation and infection
.
Front Immunol
2018
;
9
:
754
.
57.
Weinhold
B
,
Bader
A
,
Poli
V
,
Rüther
U
.
Interleukin 6 is necessary, but not sufficient, for induction of the human C-reactive protein gene in vivo
.
Biochem J
1997
;
325
(
Pt 3
):
617
21
.
58.
Ridker
PM
.
From C-reactive protein to interleukin 6 to interleukin 1: moving upstream to identify novel targets for atheroprotection
.
Circ Res
2016
;
118
:
145
56
.
59.
Rebelo
P
,
Oliveira
A
,
Andrade
L
,
Valente
C
,
Marques
A
.
Minimal clinically important differences for patient-reported outcome measures of fatigue in patients with COPD following pulmonary rehabilitation
.
Chest
2020
;
158
:
550
61
.
60.
Cocks
K
,
King
MT
,
Velikova
G
,
Martyn St-James
M
,
Fayers
PM
,
Brown
JM
.
Evidence-based guidelines for determination of sample size and interpretation of the European Organisation for the Research and Treatment of Cancer Quality of Life Questionnaire-Core 30
.
J Clin Oncol
2011
;
29
:
89
96
.
61.
Jayadevappa
R
,
Cook
R
,
Chhatre
S
.
Minimal important difference to infer changes in health-related quality of life: a systematic review
.
J Clin Epidemiol
2017
;
89
:
188
98
.
62.
Carrero
JJ
,
Andersson Franko
M
,
Obergfell
A
,
Gabrielsen
A
,
Jernberg
T
.
hsCRP level and the risk of death or recurrent cardiovascular events in patients with myocardial infarction: a healthcare-based study
.
J Am Heart Assoc
2019
;
8
:
e012638
.
63.
Salazar
J
,
Martínez
MS
,
Chávez
M
,
Toledo
A
,
Añez
R
,
Torres
Y
, et al
.
C-reactive protein: clinical and epidemiologic perspectives
.
Cardiol Res Pract
2014
;
2014
:
605810
.
64.
Bogaty
P
,
Dagenais
GR
,
Joseph
L
,
Boyer
L
,
Leblanc
A
,
Bélisle
P
, et al
.
Time variability of C-reactive protein: implications for clinical risk stratification
.
PLoS One
2013
;
8
:
e60759
.
65.
Koc
M
,
Karaarslan
O
,
Abali
G
,
Batur
MK
.
Variation in high-sensitivity C-reactive protein levels over 24 hours in patients with stable coronary artery disease
.
Tex Heart Inst J
2010
;
37
:
42
8
.

Supplementary data