Impact of sleep deprivation on monocyte subclasses and function

Abstract

The relationship between sleep deprivation, obesity, and systemic inflammation is a critical area of investigation due to its significant impact on health. While it is established that poor sleep adversely affects obesity and metabolic syndromes, the specific mechanisms, particularly subclinical inflammation independent of obesity, remain unclear. This study investigates how sleep quality influences monocyte subclass distribution and its association with systemic inflammation across a spectrum of body mass index categories. In our cohort study, 237 healthy participants were categorized by body mass index. Participants' dietary intake, physical activity, and sleep patterns were objectively tracked through wearable ActiGraph GT3X accelerometer. The data showed that obese individuals had significantly lower sleep quality and higher chronic low-grade inflammation. Nonclassical monocytes increased significantly in obesity, correlating with reduced sleep quality and elevated proinflammatory cytokines. Although body mass index emerged as a significant factor in driving inflammation, mediation analyses further defined that sleep disruption independently contributes to inflammation, regardless of obesity status. Controlled sleep deprivation experiments confirmed these findings, demonstrating reversible increases in nonclassical monocytes expression. This study highlights the importance of sleep quality in regulating immune responses and inflammation in obesity, suggesting that improving sleep quality could reduce inflammation and improve health outcomes.

CD16, inflammation, macrophages, NCM, sleep deprivation

Introduction

The intricate relationship between sleep deprivation and obesity, recognized as a critical factor in the onset of obesity-induced systemic inflammation, has acquired considerable attention in contemporary medical research. Substantial evidence now links sleep disorders and disturbances to an array of chronic conditions and morbidity such as type 2 diabetes mellitus and cardiovascular diseases.^1,2 Interestingly, this observed association is also identified in the opposite direction, as conditions such as type 2 diabetes mellitus and rheumatoid arthritis were found to contribute significantly to poor sleep health, further complicating this multifaceted issue.^3,4 Despite these associations, the direct correlation between sleep efficiency and its influence on circulatory immune cells and their responses remains an area of scientific ambiguity.

Monocytes, derived mainly from the bone marrow's hematopoietic stem cells, represent a vital component of the body's innate immune system. They play a pivotal role in immune surveillance and response to pathogens. These monocytes can be categorized into 3 distinct subsets based on the expression of surface markers CD14 and CD16: (i) classical monocytes (CMs), characterized by high expression of CD14 and low expression of CD16 (CD14⁺⁺CD16⁻), are typically regarded as the “patrolling” subset involved in routine immune surveillance, and they possess phagocytic capabilities and play a key role in initiating immune responses against invading pathogens; (ii) intermediate monocytes (IMs), identified with double presentation of (CD14⁺CD16⁺), exhibit an enhanced capacity for antigen presentation and proinflammatory cytokine production, making them potent effectors in immune activation and inflammation; and (iii) nonclassical monocytes (NCMs), characterized by low expression of CD14 and high expression of CD16 (CD14^lowCD16⁺), are specialized in patrolling the vasculature and extravascular tissues, and they exhibit a unique ability to sense and respond to inflammatory cues, contributing to the maintenance of immune surveillance and tissue homeostasis. Importantly, NCMs demonstrate heightened proinflammatory responses, suggesting their involvement in immune regulation and modulation of inflammatory processes.^5–7

In a recent study conducted by Polasky et al.,⁸ the impact of obstructive sleep apnea syndrome (OSA) on monocyte subsets was investigated. OSA, characterized by respiratory obstruction and oxygen desaturation during sleep, is associated with oxidative stress and increased risk of chronic diseases such as obesity. In their work, the group has shown that in mice models with OSA exhibit a decrease in CMs and an increase in CD16⁺ subsets, with obesity exacerbating these changes due to reduced physical activity (PA) and elevated proinflammatory mediators such as tumor necrosis factor α (TNF-α) and monocyte chemoattractant protein-1 (MCP-1).⁸ However, the complexity of OSA and the overall prevailing proinflammatory milieu of the condition further obscure the relationship between sleep and the differentiation patterns of circulating monocytes with regard to the 3 previously described subsets.

To explore the association between sleep health and immunological alterations within monocyte subclasses, and their potential impact on inflammatory responses, particularly in the context of obesity, independent of chronic sleep conditions such as OSA, we evaluated the role of sleep on inflammation, hypothesizing that poor sleep quality can exacerbate inflammatory responses through alterations in monocyte subsets. The study examined the complex relationships between sleep patterns, monocyte subsets, and inflammation. This research sought to deepen our knowledge of the fundamental processes that connect sleep, immune function, and inflammatory reactions. To clarify these pathways, we proposed a model that emphasizes the evaluation of the association between sleep quality and inflammation, with body mass index (BMI) considered first as a potential confounder and then as an effect modifier. The model illustrated how sleep disturbances may influence monocyte subsets and contribute to a proinflammatory state, while maintaining a focus on the relationship between sleep and inflammation within the context of varying BMI.

Methods

Study design, participants, inclusion/exclusion criteria, and anthropometric measurements

In this comprehensive cross-sectional analysis, we targeted a cohort of 350 adult Kuwaiti individuals (18 yr of age and older). Recruitment strategies included randomized outreach through word of mouth, distribution of flyers, and social media campaigns. Of these, 327 volunteers consented to participate by signing an informed consent form and completing a detailed health screening questionnaire. This screening process was crucial to ensure the inclusion of individuals exhibiting normal liver, cardiopulmonary, and kidney functions based on self-reported health status. The study's exclusion criteria were stringent, excluding individuals with physician-diagnosed diabetes, OSA, or hypertension (blood pressure >160/90 mm Hg), and those undergoing antihypertensive treatment. Participants with a history of significant coronary heart disease events (e.g., myocardial infarction, coronary artery bypass graft surgery, coronary angioplasty), or with a family history of premature cardiac death (before the age of 40 yr) were also excluded. Additionally, individuals with clinically diagnosed depression or those taken medications known to affect sleep quality or body weight by altering lipid or carbohydrate metabolism were not included. Ultimately, 276 participants satisfied the inclusion and exclusion criteria. However, only 237 individuals completed the study. The dropout was due to 45 participants either withdrawing consent during the study or failing to wear the ActiGraph accelerometer for the minimum required duration of 4 nights—a threshold established in the literature to accurately determine reliable sleep-wake patterns. A flowchart detailing the recruitment process for study participants is provided (see Fig. S1). The study population was stratified based on BMI into 3 groups: lean (18.5–24.9 kg/m²; n = 120), overweight (25–29.9 kg/m²; n = 57), and obese (≥30 kg/m²; n = 60). Within the obese group, further classification included 49 individuals with class I/low-risk obesity (BMI = 30–34.9 kg/m²), 8 with class II/moderate-risk obesity (BMI = 35–39.9 kg/m²), and 3 with class III/high-risk obesity (BMI ≥ 40 kg/m²).

Objective sleep patterns were analyzed using ActiGraph GT3X+ activity monitors, which participants were instructed to wear for 7 consecutive days, except during bathing. The device was positioned over the hip to monitor activity levels accurately. All subjects received prior training on recording their sleep, activity, and dietary intake throughout the study period. Anthropometric assessments were standardized in the PA laboratory. Participants were measured in tight-fitting clothes using consistent equipment to ensure accuracy. Height was measured to the nearest 0.1 unit, with volunteers positioned against a height scale, ensuring proper posture and head alignment. Body weight was determined using a beam balance, with BMI subsequently calculated using the following formula: weight (kg) / height (m²). Waist and hip circumferences were measured using a nonelastic tape, with waist-to-hip ratios derived accordingly. All measurements were performed by the same investigator to maintain consistency. Body composition parameters, including percent body fat, soft lean mass, and total body water, were determined using an IOI 353 Body Composition Analyzer (Jawon Medical). The anthropometric and clinical characteristics of the study participants are summarized in Table 1.

Table 1.

Demographic characteristics of the study population.

Physical characteristics of subjects	Lean group (n = 120)	Overweight group (n = 57)	Obese group (n = 60)	P value
Sex
Female	60	36	29
Male	60	24	28
Age, yr	30.4 ± 5.3	33.1 ± 5.7	32.4 ± 3.3	0.0773
Weight, kg	63.3 ± 10.1	79 ± 9.4	93.2 ± 13.9	<0.0001
Height, cm	169.9 ± 13.7	168 ± 10.6	166.8 ± 10.3	0.9332
BMI, kg/m2	21.8 ± 1.6	28 ± 1.2	33.2 ± 3.2	<0.0001
Waist circumference, inches	29.2 ± 2.1	35 ± 3.3	43.4 ± 7.1	<0.0001
Hip circumference, inches	39.8 ± 10.4	41 ± 2.8	46.1 ± 3.4	0.0492
Waist-to-hip ratio	0.76 ± 0.039	1 ± 0.0	0.94 ± 0.1	0.9784
Fat %	20.6 ± 9.4	29 ± 5.4	38.5 ± 5.9	<0.0001

Values are n or mean ± SD. P = 0.05 was considered significant.

Controlled study protocol incorporating 24-h sleep deprivation

In this rigorously designed control study, we selected a subgroup of 5 healthy, lean individuals from a pool of 120 applicants deemed eligible from the lean category. Echoing the procedure of the main cohort, all participants provided informed consent and were equipped with ActiGraph GT3X+ activity monitors. These devices were to be worn for a duration of 5 consecutive days, except during bathing, to accurately capture activity levels and sleep patterns. Anthropometric measurements and an initial blood sample were collected on day 0 (D0), marking the commencement of the study. Subsequently, participants were instructed to return to their homes and resume their usual daily activities, including their typical sleep-wake cycles. On D3, a critical modification was introduced: participants were required to undergo a continuous 24-h period of wakefulness accompanied by constant monitoring by the investigator. Following this cycle, a clinic visit was scheduled the next morning for the collection of another blood sample, to assess the physiological impacts of sleep deprivation. After this intensive sleep deprivation phase, participants were advised to take a restorative 2-d break, during which they were encouraged to revert to their normal lifestyle and sleep routines. On D5, the final day of the study, participants returned for a concluding session where a final blood sample was collected, and the ActiGraph GT3X+ monitors were submitted for data analysis. The ActiGraph data extracted from each participant were meticulously analyzed to ascertain the effects of the 24-h sleep deprivation on their physiological and metabolic parameters.

Ethical considerations and approval

Prior to their enrollment in the study, all participants were comprehensively briefed about its objectives and methodologies. To ensure adherence to ethical standards, written informed consent was obtained from each participant, aligning with the ethical principles outlined in the Declaration of Helsinki. This study received formal ethical approval from the Kuwait Ministry of Health Ethical Board (Approval ID#: 2017/542) and the Ethical Review Committee (ERC) of the Dasman Diabetes Institute, Kuwait (Approval ID#: RA HM-2019-019), underscoring our commitment to ethical research practices and participant welfare.

Nutritional intake monitoring and evaluation

To accurately track and analyze the dietary habits of all participants, they were provided with food diaries and instructed to meticulously weigh and document their consumption of both food and beverages using electronic scales (Salter Housewares). To ensure accuracy and consistency in recording, each participant received a hands-on demonstration on the proper use of the scales and how to maintain their diaries before the commencement of the study. Participants were encouraged to continue with their usual dietary practices without alterations. The food diary entries were collected and reviewed before the participants' second visit. The gathered data from the food diaries were then processed using CompEat Pro software (Nutrition Systems) Version 1.31.19, facilitating the computation of an average daily nutrient intake for each participant. This approach allowed for a detailed and comprehensive analysis of the dietary patterns among the study participants, contributing valuable insights into the relationship between nutrition and health outcomes.

PA assessment methodology

In our study, the ActiGraph GT3X+ activity monitor was employed to quantitatively assess both PA levels and sleep quality metrics. This advanced device utilizes a triaxial electronic monitoring system (wGT3X-BT; ActiGraph) to accurately record ACTi graphical activity. Participants were instructed to continue with their regular daily activities without altering their normal (habitual) PA routines throughout the duration of the study. To ensure accurate measurements, the accelerometers were affixed to elastic belts and positioned on the right side of the hip, to be worn continuously for 7 d. Exceptions were made for bathing or any activities involving water, as previously detailed in the literature, to maintain the integrity of the device.

The actigraphy methodology employed in this study has been validated as a reliable and objective tool for assessing daily PA. The accelerometer's output includes comprehensive PA metrics such as activity counts, vector magnitude, energy expenditure, step counts, activity intensity levels, and metabolic equivalents. For precise data collection, a 1-min epoch setting was utilized, with activity counts evaluated at 1-min intervals. This granularity ensured the inclusion of data from at least 4 d in which the device was worn for a minimum of 60% of the day, as per established criteria. Any period of ≥60 min with zero activity count was classified as nonwear time. To categorize PA intensity, Freedson’s cutoff points were applied, defining light intensity as 100 to 1,951 counts/min, moderate intensity as 1,952 to 5,724 counts/min, and vigorous intensity as ≥5,725 counts/min. Counts ≤99/min were classified as sedentary.⁹ The aggregated data were then analyzed to calculate the mean intensity for each activity level throughout the monitoring period, expressed as total accelerometer counts per total monitoring time.

Evaluation of sleep quality and duration

For the assessment of sleep quality metrics, the ActiGraph GT3X+ device has been previously validated as an effective tool for sleep analysis, exhibiting congruence with both polysomnography and Actiwatch methodologies.¹⁰ In our study, sleep quality and duration were objectively determined using this electronic triaxial monitor (wGT3X-BT) over a span of 7 consecutive days. To accurately calculate sleep quality metrics, a 1-min epoch setting in zero-crossing mode was utilized. Aligning with established research, which outlines the minimum requisite data for dependable sleep-wake pattern identification via actigraphy, participants contributing <4 nights’ worth of data were omitted from the analysis.¹¹ Sleep and wake periods were delineated using bedtime and wake-time entries from the participants' sleep diaries, supplemented by a visual examination of the actigraphy tracings for verification. The analysis leveraged the Cole-Kripke algorithm to differentiate between sleep and wake phases, enabling the computation of key sleep metrics, including sleep efficiency, wake after sleep onset (WASO), and total sleep time (TST). These outcomes were averaged across all recorded nights to portray a comprehensive view of each participant's typical sleep pattern. Sleep efficiency is quantified as the ratio of total measured sleep time to the total time spent in bed, with augmentation in values indicative of enhanced sleep continuity. This parameter was treated as a continuous variable for analysis purposes. WASO, defined as the cumulative minutes recorded as awake after sleep initiation (as detected by movement), was also analyzed as a continuous variable, in which higher figures signify prolonged wakefulness periods post–sleep onset. Total sleep duration encapsulates the aggregate time spent asleep during the bed occupancy period, as determined through actigraphy.

Sleep apnea evaluation protocol

In the initial phase of our study, each participant was subjected to a comprehensive health screening aimed at cataloging their medical history, with a specific focus on the identification of any pre-existing sleep apnea (SA). Those who self-reported the presence of SA were subsequently excluded from further participation to maintain the integrity of the study's focus group. Subsequently, participants were presented with a meticulously designed sleep questionnaire, constituting our primary tool for evaluating SA risk. This questionnaire was crafted to probe 2 critical aspects of potential SA indicators over the past month: (i) instances of experiencing difficulty in breathing comfortably and (ii) occurrences of coughing or loud snoring. Participants affirming either condition with a frequency surpassing 3 times per week were presumed to exhibit SA symptoms and were, therefore, considered ineligible for the study.

Quantification of plasma cytokines and chemokines

Participants were asked to return for a second laboratory visit after an overnight fast of at least 10 h. All blood draws were scheduled for the early morning. Blood samples were collected in 10 mL EDTA tubes (BD Vacutainer System). Plasma was separated, aliquoted, and stored at −80°C for subsequent analysis. A comprehensive analysis of plasma cytokines and chemokines, encompassing a total of 41 distinct biomarkers was conducted via MILLIPLEX MAP Human Cytokine/Chemokine Magnetic Bead Panel – Premixed 41 Plex – Immunology Multiplex Assay (Milliplex map kit, HCYTMAG-60 K-PX41; Millipore). Adhering strictly to the guidelines provided by the manufacturer, we ensured the accuracy and reliability of our measurements. The assay's reactions were meticulously recorded using a Luminex-based MILLIPLEX analyzer. This advanced system was complemented by a digital processor tasked with managing the resulting data efficiently. The subsequent analysis was carried out using the MILLIPLEX Analyst software, which facilitated the precise determination of the mean fluorescence intensity and the concentration of each analyte, reported in pmol/L.

Flow cytometry analysis of monocyte subclasses

To assess the expression profiles of various monocyte subclasses within whole blood specimens, we implemented a multi-color fluorescence-activated cell sorting (FACS) analysis using freshly drawn whole blood samples. The procedure commenced with the addition of 1 mL of lysing buffer to a 0.1 mL blood sample, a step crucial for the selective removal of erythrocytes and isolation of the peripheral blood leukocyte populations. Subsequent to erythrocyte elimination, cells underwent 2 washing cycles with 1 mL of phosphate-buffered saline each, followed by incubation with fluorochrome-conjugated mouse anti-human monoclonal antibodies. Specifically, we utilized BD Pharmingen PE Mouse Anti-Human CD14 (Cat # 562691) and BD Pharmingen APC Mouse Anti-Human CD16 (Cat # 561248), alongside isotype-matched control antibodies (PE Mouse IgG2b, κ Isotype Control [Cat # 559529] and APC Mouse IgG1, κ Isotype Control [Cat # 340442]; BD Pharmingen) for this purpose. Our gating strategy, detailed in Fig. S2, facilitated the precise identification and quantification of the 3 distinct monocyte subclasses that included CMs (CD14⁺⁺CD16⁻), IMs (CD14⁺⁺CD16⁺), and NCMs (CD14^lowCD16⁺⁺). Among the leukocyte populations. Data acquisition was performed using a BD FACSCanto II flow cytometer, with subsequent analysis conducted via DIVA software (version V6.1.3; BD Pharmingen).

Statistical methodology

Our data underwent rigorous analysis employing SPSS version 25 (IBM) and GraphPad Prism 7.01 (version 6.05; GraphPad Software), with results presented as mean ± SD. Initial testing for data normality was conducted using the Shapiro-Wilk normality test to ensure the appropriateness of subsequent statistical tests. For the comparative analysis between 3 group comparisons, 1-way analysis of variance and exact Kruskal-Wallis tests were respectively used for parametric and nonparametric data to discern mean differences. Following the 1-way analysis of variance, post hoc comparisons between groups were performed using the Tukey honest significant difference test to adjust for multiple comparisons while controlling the familywise error rate. Further, we conducted multiple linear regression analyses to explore the relationships between the monocyte subclasses and measured efficiency of sleep.

To investigate whether immune parameters, specifically NCMs, mediate the relationship between obesity and sleep efficiency, we conducted a series of regression analyses (Fig. S3). The mediation analysis followed these steps:

1. Direct Effect of BMI on TST (path c): To assess the direct effect of BMI on TST using linear regression.

2. Effect of BMI on NCMs (path a): To evaluate the effect of BMI on NCM levels.

3. Effect of NCMs on TST Controlling for BMI (path b): To examine the effect of NCM level on TST while controlling for BMI.

4. Direct Effect of BMI on TST Controlling for NCMs (path c’): To assess the direct effect of BMI on TST while controlling for NCMs.

For all analysis, P values of ≤0.05 were delineated as the threshold for statistical significance, guiding the interpretation of our findings and ensuring rigor in our conclusions.

Results

Demographic characteristics of the study population

To explore the intricate interactions between sleep, body weight, and inflammation, we conducted a meticulously controlled study, enlisting 237 participants who were carefully matched in terms of demographic and health characteristics. Ensuring a robust and reliable dataset, all participants were in optimal health at the time of the study, with no recorded current or past medical conditions that could confound the results. As illustrated in Fig. 1, a total of 112 male and 125 female participated in this study, and participants were stratified into 3 distinct groups based on their BMI: lean (≤25 kg/m²), overweight (26–29 kg/m²), and obese (≥30 kg/m²), as detailed in Table 1. Descriptive characteristics of the study population stratified by gender are provided in Table S1. In our rigorous examination of potential confounding variables, we observed no significant differences across the 3 groups in terms of dietary intake (as presented in Table S2) and PA levels (Table S3). This careful control of variables ensured the integrity of our findings, allowing us to focus on the primary research question concerning the relationship between sleep and systemic inflammation in varying BMI categories.

Figure 1.

A schematic illustration of the research protocol showing the chronological sequence of study procedures and associated participant activities.

Objectively measured sleep analysis

Objectively measured sleep analysis reveals intricate dynamics between obesity and sleep quality. In this investigation, actigraphic assessments provided an estimate of total sleep efficiency calculated as the proportion of time spent asleep compared with the total time in bed. In a general overview, the average sleep efficiency score measured by the actigraphy assessment was 91.4 ± 3.9%, with a total sleep duration of 466.7 ± 142.7 min per night, which is equivalent to 7.78 ± 2.38 h per night, meeting the minimum recommended duration by the National Sleep Foundation.¹² In cross BMI examination, there was a visible significant decrease in total sleep efficiency among obese participants when compared with their lean counterparts (Fig. 2A). This observation aligns with the hypothesized interplay between increased body weight and disrupted sleep architecture. The trend toward reduced sleep efficiency in the overweight cohort, though not reaching statistical significance, hints at a possible gradational impact of weight on sleep quality. Further examination of the contributing sleep quality metrics revealed that both sleep duration and WASO were significantly altered in the obese group relative to the lean group (Fig. 2B–D). Together these observations are indicative of a multifaceted impairment in sleep quality. Interestingly, the heightened WASO could be representative of a fragmented sleep pattern, potentially attributable to underlying physiological disruptions commonly associated with obesity, such as increased sympathetic nervous system activity or chronic low-grade inflammation¹³

Figure 2.

Objectively measured sleep analysis and its association with obesity. Actigraphic assessments aimed at interpreting the relationship between BMI level and sleep quality. (A) Total sleep efficiency was estimated based on the proportion of time spent asleep compared with the total time in bed. (B) Total time in bed indicates the overall time spent in bed, from attempting to sleep to the final waking. (C) TST is the cumulative amount of actual sleep obtained, excluding all periods of wakefulness. (D) WASO highlights periods of wakefulness after sleep initiation, important for understanding sleep continuity. (E) Number of awakenings and (F) average awakening period provide insights into the frequency and extent of sleep disruptions. All data are expressed as mean ± SD and compared between groups using 1-way analysis of variance with Tukey’s multiple comparisons test. P ≤ 0.05 was considered statistically significant.*P < 0.05, ****P < 0.0001.

Associations of higher BMI with subclinical inflammation

Peripheral blood chemokines and cytokines protein secretion were measured using multiplex assays. Out of the 38-inflammatory biomarkers tested, 13 markers were found to be significantly altered in overweight and obese compared with lean participants (Table 2).

Table 2.

Study population circulatory cytokine expression.

	Lean group (n = 120)	Overweight group (n = 57)	Obese group (n = 60)	P value
Sex
Female	60	36	29
Male	60	24	28
IL-15, pg/mL	1.5 ± 1.6	0.9 ± 1.3	1.7 ± 1.3	0.3590
IL-9, pg/mL	3.3 ± 2.9	5.6 ± 6.8	7.1 ± 10.8	0.8089
IL-12P40, pg/mL	5.9 ± 6.9	30.3 ± 46.8	5.3 ± 7.9	0.0285
IL-12P70, pg/mL	59.7 ± 35.1	79.1 ± 46.4	103.2 ± 60	0.0583
IL-13, pg/mL	4.9 ± 1.2	5.7 ± 1.9	6.9 ± 2.6	<0.0001
IL-1α, pg/mL	55.5 ± 109.6	96.1 ± 135.3	106.6 ± 187.8	0.3946
MCP-1, pg/mL	196.7 ± 31.8	427.2 ± 174.7	297.3 ± 67.6	<0.0001
IL-1RA, pg/mL	115.2 ± 173.6	98.5 ± 43.5	150.6 ± 298	0.2167
VEGF, pg/mL	284.4 ± 322.4	469.9 ± 260.5	573.8 ± 243.4	<0.0001
IL-1β, pg/mL	3.4 ± 2.6	3.5 ± 2.4	4.6 ± 4.8	0.0589
IL-8, pg/mL	19 ± 22.3	14.6 ± 12.5	20.4 ± 23.9	0.3192
IL-6, pg/mL	6.7 ± 6.4	5.7 ± 5.4	9.5 ± 3.3	0.0007
Eotaxin, pg/mL	100.2 ± 82.3	172.4 ± 76.9	190.8 ± 86.3	0.0111
IP-10, pg/mL	151.5 ± 23.3	175.4 ± 37.7	226.1 ± 40.8	<0.0001
IL-7, pg/mL	11.2 ± 3.7	20.7 ± 24.7	3.6 ± 2.9	0.1211
IL-17A, pg/mL	30.9 ± 22.6	17.4 ± 18.1	33.5 ± 26.0	0.0002
MCP-3, pg/mL	34.0 ± 91.1	14.3 ± 33.0	91.1 ± 134.0	0.1710
sCSD40L, pg/mL	2450.1 ± 3163.5	5103.6 ± 4051.1	4577.8 ± 2170.5	0.0643
G-CSF, pg/mL	209.4 ± 480.1	228.8 ± 330.0	257.1 ± 310.6	0.9622
TNF-β, pg/mL	12.4 ± 14.9	8.9 ± 9.8	15.6 ± 16.9	0.7170
GM-CSF, pg/mL	18.4 ± 35.9	34.7 ± 71.0	37.3 ± 42.0	0.3572
Fractalkine, pg/mL	184.9 ± 477.3	240.5 ± 327.2	232.8 ± 279.9	0.9230
TNF-α, pg/mL	2.4 ± 1.1	9.6 ± 6.6	13.5 ± 6.7	<0.0001
TGF-α, pg/mL	33.2 ± 59.6	8.2 ± 7.4	6.1 ± 5.0	0.2555
Fit-3L, pg/mL	10.1 ± 25.0	39.2 ± 50.40	7.5 ± 14.0	0.0229
FGF-2, pg/mL	465.7 ± 397.1	143.1 ± 127.3	168.6 ± 102.9	0.0040
IFNγ, pg/mL	10.0 ± 3.1	10.4 ± 2.7	12.0 ± 2.9	0.0003
IL-10, pg/mL	12.8 ± 21.5	9.5 ± 24.3	7.4 ± 4.8	0.4478
MDC, pg/mL	452.5 ± 432.8	837.8 ± 162.6	812.3 ± 185.8	0.0097
GRO, pg/mL	1052.4 ± 1994.2	2643.5 ± 3039.9	2538.3 ± 3347.5	0.0696
MIP-1β, pg/mL	38.2 ± 44.1	57.1 ± 67.2	79.4 ± 52.9	0.0642
IFN-α2, pg/mL	139.8 ± 159.8	127.8 ± 223.6	108.50 ± 199.5	0.9028
MIP-1α, pg/mL	4.9 ± 0.9	5.8 ± 1.0	7.0 ± 1.2	<0.0001
IL-3, pg/mL	2.2 ± 2.9	1.3 ± 0.8	1.0 ± 0.9	0.4252
EGF, pg/mL	150.2 ± 175.1	195.3 ± 141.6	278.6 ± 230.8	0.2178
IL-5, pg/mL	10.1 ± 19.0	5.2 ± 10.3	16.0 ± 24.7	0.5701
IL-2, pg/mL	6.9 ± 13.7	8.3 ± 16.1	11.6 ± 24.0	0.8005
IL-4, pg/mL	87.5 ± 190.7	150.7 ± 222.6	97.2 ± 149.1	0.9703

Values are n or mean ± SD. P = 0.05 was considered significant.

EGF, epidermal growth factor; FGF-2, fibroblast growth factor-2; G-CSF, granulocyte colony-stimulating factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; GRO, Growth-Regulated Oncogene; IP-10, Interferon-gamma Induced Protein 10; MDC, Macrophage-Derived Chemokine; IP, Interferon-gamma Induced Protein-10; MDC, Macrophage-Derived Chemokine; TGF-α, transforming growth factor α; VEGF, vascular endothelial growth factor.

Lean vs overweight.

Overweight vs obese.

Lean vs obese.

To test the distribution and prevalence of circulating monocyte subsets in each group, we conducted a flow cytometric analysis to identify each subclass according to their surface presentation of CD14 and CD16 as follows: CMs with surface presentation of CD14⁺CD16⁻, IMs presented as CD14⁺CD16⁺, and NCMs with CD14^lowCD16⁺ presentation. Within our cohort, we identified a significant reduction in CM subsets in the obese group compared with the lean group while no significant change was seen in the overweight group (Fig. 3A). Both IMs and NCMs showed significant upregulation in overweight and obese groups compared with lean group (Fig. 3B–D). To understand the influence of these cells, we conducted a Pearson r correlation analysis to identify the association between these monocyte subclasses and sleep efficiency. Interestingly, a positive significant association can be seen between elevated CMs and higher sleep efficiency, while a significant negative association is presented with the elevation of NCMs. No significance was detected with IMs and Sleep efficiency (Fig. 3E–G).

Figure 3.

Monocyte subclass distribution and correlation with sleep efficiency. Whole blood samples were taken from each participant. Samples were red blood cell lysed and stained for both anti-CD14-PE and anti-CD16-APC antibodies to identify monocyte subclasses and their distribution. (A) Expression level of CMs identified as CD14⁺CD16⁻. (B) Expression level of IMs identified as CD14⁺CD16⁺. (C) Expression level of NCMs identified as CD14^lowCD16⁺. (D) Representative dot plot for CD14 vs CD16 subsets cross BMI levels. (E) Pearson’s correlation coefficient analysis was conducted to investigate the relationship between CMs and efficiency of sleep. (F) IMs and efficiency of sleep and (G) NCMs and efficiency of sleep. All data are expressed as mean ± SD and compared between groups using 1-way analysis of variance with Tukey’s multiple comparisons test. Pearson r correlation was conducted for parametric correlation analysis. P ≤ 0.05 was considered statistically significant. ****P < 0.0001.

Furthermore, Pearson r correlation analysis further identified a significant positive association between the elevation of peripheral NCMs in circulation and 9 out of those 12 identified inflammatory biomarkers found to be elevated in obesity, which were IL-13, MCP-1, vascular endothelial growth factor, eotaxin, TNF-α, interferon γ (IFNγ), Macrophage-Derived Chemokine (MCD), and MIP-1α while a negative association was seen between NCMs and fibroblast growth factor-2. CMs was found to be negatively associated with MCP-1, TNF-α, IFNγ, and MIP-1α, and IM elevation was only found to be positively associated with MCP-1, IL-17A, TNF-α, IFNγ, and MIP-1α (Table 3).

Table 3.

Association between circulatory monocyte subclasses and cytokine expression.

Circulatory cytokine expression	CM CD14+CD16−		IM CD14+CD16+ (%)		NCM CD14lowCD16+
	95% CI (%)	P value	95% CI (%)	P value	95% CI (%)	P value
IL-12P40	−0.28 to 0.17	0.60	−0.27 to 0.18	0.70	−0.0078 to 0.43	0.058
IL-13	−0.16 to 0.10	0.64	−0.068 to 0.19	0.33	0.047 to 0.30	0.008
MCP-1	−0.36 to −0.11	<0.0001	0.046 to 0.30	0.008	0.35 to 0.56	<0.0001
VEGF	−0.23 to 0.038	0.15	−0.023 to 0.24	0.10	0.035 to 0.29	0.01
IL-6	−0.15 to 0.10	0.72	−0.095 to 0.17	0.57	−0.0062 to 0.25	0.06
Eotaxin	−0.29 to 0.16	0.55	−0.13 to 0.31	0.39	0.037 to 0.46	0.02
IP-10	−0.20 to 0.089	0.44	−0.011 to 0.27	0.07	0.16 to 0.42	<0.0001
IL-17A	−0.079 to 0.18	0.29	−0.13 to 0.12	0.01	−0.13 to 0.13	0.16
TNF-α	−0.48 to −0.25	<0.0001	0.23 to 0.46	<0.0001	0.45 to 0.64	<0.0001
Fit-3L	−0.26 to 0.18	0.75	−0.16 to 0.28	0.57	−0.0099 to 0.42	0.06
FGF-2	−0.025 to 0.41	0.08	−0.38 to 0.053	0.13	−0.45 to −0.031	0.02
IFNγ	−0.38 to −0.14	<0.0001	0.1428 to 0.390	<0.0001	0.28 to 0.50	<0.0001
MDC	−0.31 to 0.14	0.44	−0.10 to 0.34	0.29	0.011 to 0.44	0.04
MIP-1α	−0.36 to −0.11	<0.0001	0.18 to 0.42	<0.0001	0.31 to 0.53	<0.0001

P = 0.05 was considered significant.

FGF-2, fibroblast growth factor-2; IP, Interferon-gamma Induced Protein-10; MDC, Macrophage-Derived Chemokine; VEGF, vascular endothelial growth factor.

This suggests that the polarization of monocytes towards the NCM phenotype plays a critical role in driving systemic inflammation. These findings underscore the importance of monocyte dynamics in modulating inflammatory responses, particularly in the context of obesity and sleep disturbances, where altered immune profiles contribute to a heightened inflammatory state. This association highlights potential mechanistic pathways linking immune cell behavior to chronic inflammation and related health outcomes.

To further define the influence of sleep on the upregulation of NCMs in circulation, we investigated the independent associations of 5 sleep quality metrics with the circulation level of NCMs. However, given the potential influence of BMI on the relationship between sleep and NCMs, we included BMI as a covariate in our multivariable regression model. The regression model was adjusted for age, PA, dietary intake, and BMI to provide a more accurate assessment of these associations (Table 4).

Table 4.

Multiple regression analysis to define the impact of all variables on the upregulation of NCMs in the lean group.

Predictor	Coefficient (β)	SE	P value
NCM CD14lowCD16+ subsets
Intercept	8.561	7.532	0.2572
Dietary intake	−5.853 × 10−5	0.0003588	0.8706
Age	−0.07918	0.04063	0.0529
BMI	0.3202	0.04286	<0.0001
Percentage of sedentary time	−0.02485	0.03213	0.4404
Total efficiency of sleep	−0.07081	0.06438	0.2729
Total time in bed	−0.01513	0.006835	0.0281
TST	0.01391	0.006979	0.0477
WASO	0.02562	0.01633	0.1185
Average number of awakenings	0.04973	0.06385	0.4371
Average awakening duration	0.09667	0.07906	0.2230

As expected, our analysis indicated that BMI was a significant covariate in the relationship between sleep and NCMs (β = 0.3202, P < 0.0001). Subsequently, we conducted a mediation analysis as shown in Fig. S3. Our analysis revealed a significant direct effect of BMI on TST, with a coefficient (β) of −0.01067 (P < 0.0001) (path c), and a significant effect of BMI on NCM levels (β = 0.3767, P < 0.0001) (path a). Furthermore, NCM levels were found to significantly affect sleep, even when controlling BMI (β = −0.006449, P < 0.0001) (path b). Notably, when NCMs were included in the model, the direct effect of BMI on sleep efficiency was no longer significant (β = −0.0004399, P = 0.8373) (path c’). These findings suggest that the relationship between BMI and sleep is mediated by NCM levels, reinforcing the critical role of NCMs in the interplay between obesity, sleep disturbances, and systemic inflammation. This mediation analysis, combined with the observed associations between sleep quality and NCM levels, highlights the importance of addressing both the quality of sleep and immune modulation in managing obesity-related inflammatory states. Notably, when we stratified the data by BMI, the association between TST and NCMs remained consistent across all weight groups, including lean participants, indicating that sleep itself plays a critical role in modulating NCM levels, independent of BMI (Table S4). Interestingly, while TST was independently associated with NCMs in the lean group, in both the overweight and obese groups, NCMs were independently associated with WASO. These findings further emphasize that restful sleep characteristics, particularly TST, may significantly contribute to the elevation of NCMs in circulation, potentially driving the subclinical inflammation observed in obesity.

Validation study in healthy normal participants

To further confirm our observation, we conducted a control investigation to identify the direct influence of sleep deprivation on circulatory monocyte subclass alteration. In our controlled experimental framework designed to probe the interplay between sleep deprivation and immune cell dynamics, we incorporated a well-characterized cohort of 5 age-matched, normal-weight individuals, with BMI ≤25 kg/m². This cohort consisted of 2 males and 3 females, detailed collectively in Table S5. Initial evaluations confirmed that all participants maintained optimal sleep efficiency at baseline, without significant gender discrepancies in this metric. At the outset in D0, the distribution of monocyte subclasses within the circulation was found to be consistent across genders, with an average expression for CMs of 77.1 ± 5.0%, for IMs of 5.5 ± 2.4%, and for NCMs of 5.12 ± 2.2%, as quantified through flow cytometry and summarized in Table 5.

Table 5.

Control study population monocyte subclass distribution and expression.

Circulatory monocyte surface expression	Males (n = 2)	Females (n = 3)	P value
Classical  CD14+CD16−, %	75 ± 1.2	78.5 ± 6.5	0.5267
Intermediate CD14+CD16+, %	4.4 ± 2.6	7.1 ± 0.9	0.2732
Nonclassical CD14lowCD16+, %	5.3 ± 4.1	5 ± 1.3	0.9075

Values are mean ± SD. P = 0.05 was considered significant.

Upon the induction of a 24-h sleep deprivation protocol on D3 (Fig. 4B), a significant upregulation of the NCM subset was observed, whereas the proportions of CMs and IMs were not affected significantly. This specific alteration in NCM levels appeared to be acutely responsive to sleep patterns, as evidenced by a reversion to baseline values upon the restoration of normal sleep by D5 (Fig. 4C–F). This compelling dataset underscores a measurable influence of sleep deprivation on circulatory monocyte subclass distribution beyond the effect of obesity, emphasizing its role in the initiation of subclinical inflammatory pathways observed with increased levels of NCMs in circulation.

Figure 4.

Influence of sleep deprivation on circulatory monocyte subclass distribution. (A) Flow chart of the controlled study. (B) Representative, day-by-day actigraphy-based evaluation of sleep quality and PA throughout the 5-d period showing successful 24 h of sleep deprivation at D3. (C) Flow cytometric analysis of CM expression. (D) Flow cytometric analysis of IM expression. (E) Flow cytometric analysis of NCM expression. (F) Representative dot plot for CD14 vs CD16 subsets at D0, D3, and D5. All data are expressed as mean ± SD and compared between groups using 1-way analysis of variance with Tukey’s multiple comparisons test. P ≤ 0.05 was considered statistically significant. *P < 0.05.

Discussion

The intricate relationship between sleep deprivation, obesity, and systemic inflammation presents a complex interplay that significantly impacts general health. Our study aimed to explore the association among these factors, with a particular focus on the role of monocyte subclasses in obesity-related inflammation under circumstances of compromised sleep quality. The impact of sleep deprivation on health is well documented, with numerous studies linking poor sleep habits to a range of adverse health outcomes, including obesity and its associated metabolic syndromes.^2,14 However, the mechanisms underlying these associations, especially in the context of subclinical inflammation characteristic of obesity, remain less understood. The data collected from our cohort indicates a significant decrease in total sleep efficiency among obese participants compared with their lean counterparts, underscoring the detrimental impact of excess body weight on sleep quality. This observation aligns well with findings from prior studies, which have established a clear link between obesity and compromised sleep quality.^2,15 Several mechanisms have been proposed to explain this association, with increased activity of the sympathetic nervous system and persistent low-grade inflammation being among the most significant.^16,17 In individuals with obesity, the sympathetic nervous system is often believed to be in a state of heightened activity, which can disrupt normal sleep patterns by increasing heart rate, blood pressure, and levels of stress hormones, thereby making it difficult to achieve and maintain deep, restorative sleep. Concurrently, obesity is characterized by chronic low-grade inflammation, marked by elevated levels of proinflammatory cytokines and acute-phase reactants. This inflammatory state not only contributes to the development of obesity-related comorbidities, but also is thought to directly impact sleep architecture.¹⁸ Indeed, in our cohort, we identified 13 markers that were significantly altered in overweight and obese participants compared with their lean counterparts. Notably, the overweight group exhibited a distinctive cytokine expression profile, with cytokines such as IL-12p40, MCP-1, sCD40L, transforming growth factor α, and Flt-3L upregulated more significantly than in both the lean and obese groups. This pattern may be indicative of a compensatory biological response occurring during the preobesity stage. Existing research suggests that individuals who are overweight may undergo a transient inflammatory response as the body adjusts to initial changes in adiposity and metabolic stress.¹⁹ Specifically, inflammatory markers such as IL-12p40 and MCP-1, which play crucial roles in modulating immune responses and adipose tissue inflammation, may act to counterbalance the initial increases in fat accumulation.²⁰ Moreover, the elevated levels of sCD40L and transforming growth factor α could represent an early endothelial response to increased adiposity, distinct from the chronic inflammatory state typically observed in obesity.²¹ Thus, the unique cytokine profile in the overweight group might reflect an active, albeit insufficient, physiological adaptation to emerging metabolic challenges before the development of overt obesity.

It is crucial to consider the role of circulatory monocyte subclasses in mediating obesity-related subclinical inflammation and its impact on sleep. Monocytes, a key component of the innate immune system, are differentiated into various subclasses, each with distinct roles in immune response and inflammation. In the context of obesity, the balance among these subclasses—namely CMs, IMs, and NCMs—is often disrupted, leading to an altered immune profile that exacerbates the state of low-grade inflammation.^22,23 Several studies have shown that the elevation of NCMs, in particular, has been associated with the progression of inflammatory diseases and is indicative of a heightened inflammatory response.^24,25 These cells are known for their roles in tissue repair, patrolling endothelial integrity, and producing proinflammatory cytokines.²⁶ To understand the association of monocyte subclasses in the paradigm of obesity and sleep dysfunction, we quantified the relative percentages of each of these 3 distinct subsets in our cohort. This approach highlights changes in the distribution among subclasses that may influence the overall monocyte-mediated inflammatory response. Our results demonstrate a significant alteration in the distribution of monocyte subclasses in obese individuals, with a notable reduction in CMs and an increase in both IMs and NCMs. It is important to note that the observed shifts in monocyte subclass percentages could potentially be influenced by relative changes in other subclasses. For instance, an increase in the percentage of NCMs might reflect not only an absolute increase in NCMs, but also a decrease in the percentage of CMs.

Regardless, the observed association of NCMs with reduced sleep efficiency, highlighting a potential mechanistic link between sleep disruption and immune modulation.

Indeed, when we matched the association of this observed upregulation of NCMs with elevated levels of proinflammatory cytokines and chemokines seen in obese individuals, we detected a significant matched association to 9 inflammatory markers out of 12 found to be associated with obesity. These included both the upregulation of TNF-α and MCP-1, which have been previously reported to play a major role in the regulation of sleep.^27,28 When we conducted multiple regression analysis to determine the independent association between the upregulation of NCMs in circulation and sleep patterns with BMI considered first as a potential confounder and then as an effect modifier, we found a strong independent relationship between sleep quality and NCMs in circulation regardless of BMI level.

In a study presented by Dimitrov et al.,²⁹ the authors provided additional insights into the role of monocyte subsets in sleep regulation. They found that during continuous 24-h wakefulness, total monocyte counts showed a slight increase in the early night, which was diminished when the subject slept. Specifically, sleep distinctly decreased the number of circulating NCMs, with minimum counts reached during the second half of the night, while IM counts were not influenced by sleep. This selective drop of NCMs during nocturnal sleep led to an imbalance in the proportion of these cells in the total pool of circulating monocytes. This evidence underscores the significant influence of sleep on monocyte dynamics, particularly the reduction of proinflammatory NCMs during sleep. Our findings suggest that sleep deprivation may exacerbate the inflammatory state associated with obesity by altering monocyte subclass distribution, particularly increasing NCMs, which is known to play a role in inflammation.

To investigate this observation, we conducted a controlled experiment aimed at assessing the influence of deliberate sleep deprivation on the expression of monocyte subclasses. Within this cohort, we attempted to mitigate confounding factors such as age and BMI by including participants matched in these aspects. Remarkably, our flow cytometry analysis revealed a significant elevation in NCMs following the induction of sleep deprivation on D3 compared with its baseline levels that were observed on D0. Furthermore, upon restoration of sleep on D5, we observed a reversal of this effect, underscoring a direct correlation between sleep and peripheral immune alterations. Previous research has postulated a plausible explanation for the association between sleep deprivation and inflammation. During normal sleep, there is a decrease in blood pressure and relaxation of blood vessels. However, under conditions of sleep restriction, this physiological process is compromised, leading to an immune response mediated by cells within the blood vessel walls, which subsequently impacts circulating peripheral immune cells such as our observed impact on NCMs.^30,31

Another explanation has been proposed for the impact of sleep deprivation on the body stress response system. Studies have shown that chronic sleep deprivation profoundly impacts the hypothalamic-pituitary-adrenal axis, leading to dysregulated secretion of stress hormones like cortisol.^4,32 Although these studies are primarily conducted on animal models, they highlight how hormonal dysregulation can act as a modulator that may contribute to heightened systemic inflammation, thereby exacerbating the adverse effects of sleep deprivation on immune regulation.³¹

Future directions

Future studies should aim to explore the mechanistic pathways linking sleep deprivation, monocyte subclass distribution, and inflammation in obesity. While this study acknowledges the possibility of reverse causation and the bidirectional nature of these relationships, longitudinal studies could provide valuable insights into the directionality and potential mechanisms underlying these relationships. Additionally, interventions designed to improve sleep quality in obese individuals could be evaluated for their efficacy in modulating immune responses and reducing inflammation, offering new avenues for the treatment and prevention of obesity-related complications.

Study limitations

While our study provides valuable insights into the relationship among sleep deprivation, obesity, and inflammation, it is not without limitations. First, the cross-sectional design of the main cohort limits our ability to infer causality between sleep patterns, obesity, and immune responses. Longitudinal studies would be necessary to establish temporal relationships and causative links. Second, the sample size, especially in the controlled sleep deprivation experiment, was relatively small, potentially limiting the generalizability of our findings. Larger-scale studies are necessary to validate our observations across diverse populations. Third, although actigraphy served as a validated and practical method for sleep assessment, it does not fully capture the complexity of sleep architecture as compared with polysomnography. This could lead to an underestimation of sleep disturbances among participants. Additionally, dietary intake and PA were self-reported, which could potentially introduce bias and impact the accuracy of these measurements.

Moreover, although our results indicate a relationship between sleep deprivation and alterations in monocyte subclass distribution, potentially linked to subclinical inflammation, we must acknowledge that our study did not comprehensively account for possible confounding variables like PA, exposure to light, and dietary intake during the sleep deprivation period. Therefore, the effects that we observed cannot be definitively ascribed to variations in sleep quality alone. Finally, while we have identified associations between monocyte subclasses and sleep efficiency, the underlying biological mechanisms remain speculative. Further research is needed to elucidate the pathways through which sleep influences immune cell dynamics and inflammation. Acknowledging these limitations is crucial for the interpretation of our findings and underscores the need for further research to build on the preliminary insights provided by this study. Future investigations should aim to address these limitations through longitudinal sturdy design, larger sample size, more comprehensive sleep assessments, and mechanistic studies to fully understand the complex interplay between sleep, obesity, and the immune system.

In conclusion, our study sheds light on the previously unexplored role of monocyte subclasses in the context of sleep efficiency and obesity-related inflammation. While our findings suggest an association between sleep deprivation and the distribution of monocyte subclasses and its potential link to subclinical inflammation, it is important to note that our study did not fully control for potential confounding factors, such as ambulation, light exposure, and food intake during the sleep deprivation experiment. As such, we cannot conclusively attribute the observed effects solely to changes in sleep quality. Nonetheless, our findings highlight the importance of sleep quality in the regulation of immune responses and inflammation across varying BMI categories in obesity. These insights pave the way for further research into the mechanisms underlying these associations and the development of targeted interventions to improve sleep quality and reduce inflammation, particularly in individuals with obesity.

Acknowledgments

During the preparation of this, work the authors used QuillBot AI to improve readability and language. After using this QuillBot AI, they reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

Author contributions

F.A.-R. conceived the idea, acquired funds, designed the study, interpreted the results, verified data, and wrote the manuscript. A.A.S. conducted experiments, curated data, analyzed data, interpreted the results, and participated in writing the original draft of the manuscript. N.A. and H.A.J. participated in conducting experiments, data collection, data analysis, and data interpretation. F.A.-M. participated in study design, data interpretation, and review and editing of the manuscript. R.A. conceived the idea, participated in study design, interpreted the results, acquired funds, and reviewed and editing manuscript. All authors contributed to reviewing the paper and all authors have read and approved the final version for submission.

Supplementary material

Supplementary material is available at The Journal of Immunology online.

Funding

This work was supported by grants (RH HM-2019-019) from Kuwait Foundation for the Advancement of Sciences.

Conflicts of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Data availability

The data supporting the findings of this study are available from the F.A.-R. upon reasonable request. Due to the sensitive nature of the research involving human subjects and to ensure the privacy and confidentiality of participant information, the data will be shared in accordance with ethical guidelines and institutional policies. Interested researchers can contact F.A.-R. for access to the data, subject to approval and compliance with the necessary data sharing agreements. The consent, exercise logs, food logs, sleep logs, and the health screen forms can be found deposited at the following location https://data.mendeley.com/datasets/wns2gbpwkk/1.