Mediating effect of subclinical inflammation on the process of morning hypertension leading to atrial fibrillation in community-based older adults

ABSTRACT

Background

The impacts and mechanisms of morning hypertension (MHT) on the risk of new-onset atrial fibrillation (AF) in the elderly have not been clarified. We aimed to investigate an association between MHT and new-onset AF and explore a mediating effect of subclinical inflammation on this association.

Methods

From 2008 to 2010, 1789 older adults aged ≥60 years were recruited in Shandong area, China. Morning blood pressure (BP) was assessed using 24-hour ambulatory BP monitoring. MHT was defined as BP ≥ 135/85 mm Hg during the period from wake time to 0900 a.m. Subclinical inflammation was assessed by hypersensitive C-reactive protein (hsCRP), tumor necrosis factor-alpha (TNF-α), systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and galectin-3. New-onset AF was rated during the follow-up period.

Results

Over an average 129.0 [standard deviation (SD): 21.58] months of follow-up, the hazard ratio of new-onset AF in MHT patients was 1.39 (95% confidence interval: 1.01 to 1.91) compared with non-MHT participants (P_adjusted = 0.027). The risk of new-onset AF was 1.17-fold with one-SD increment of morning systolic BP. Subclinical inflammation was significantly associated with new-onset AF. The hazard ratios of new-onset AF were 2.29, 2.04, 2.08, 2.08, 2.03, and 3.25 for one-SD increment in hsCRP, TNF-α, SII, NLR, PLR, and galectin-3, respectively (P_adjusted < 0.001). The analysis showed that hsCRP, TNF-α, SII, NLR, PLR, and galectin-3 separately mediated the process of MHT inducing new-onset AF (P_adjusted < 0.05).

Conclusions

MHT is associated with an increased risk of new-onset AF. The subclinical inflammation might play a mediating role in this association.

KEYWORDS:

• Hypertension
• circadian rhythm
• atrial fibrillation
• inflammation
• mediating effect

Introduction

Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in the elderly and leads to increased risks of stroke, systemic embolism, and heart failure (1–5). There were approximately 33 million people with AF worldwide (1–4). With the rapid expansion of the elderly population, AF has been a growing global epidemic and posed enormous burdens on the economy and the health-care utilization (1,2).

Hypertension is regarded as one of the major contributors as well as one of the modifiable factors to AF (1,6,7). Blood pressure (BP) lowering medications have been recommended by guidelines for the treatment of hypertension to prevent the new onset of AF (6,8). However, the control rate of hypertension is not satisfactory and still low, 17% for female and 14% for male, in China (9,10), although medical staff made a great effort. Essentially, BP is constantly fluctuating throughout the day. Kario and coworkers found that there are over half patients with non-valvular AF had uncontrolled morning home-measured systolic (S)BP in those with well-controlled clinical SBP (11). It indicates that morning BP might be more associated with AF than BP measured at the other time periods.

Morning hypertension (MHT) is defined as the average of BP ≥ 135/85 mm Hg during 0700–0900 a.m. measured by home or ambulatory blood pressure monitoring (HBPM or ABPM), regardless of office BP and the levels of BP measured at the other time periods (12,13). There are two types of MHT, one characterized by an exaggerated morning BP surge is defined as “morning surge (MS)” and the other characterized by continuous hypertension from nighttime to morning is defined as “sustained nocturnal and morning hypertension (SNMH)” (12–14). One study demonstrated that MS is associated with AF (15). However, the effect of SNMH on new-onset AF and the difference in the effects of MS and SNMH on new-onset AF are still unclear. In addition, the mechanism of MHT in the development of new-onset AF has not been fully elucidated, although current dogma holds that MHT triggers sympathetic over-activity leading to the development of arrhythmias including AF (6,16,17).

Subclinical and low-grade inflammation, mediated by the infiltration of immune cells and cytokines, has been implicated in the development of chronic diseases such as AF and hypertension (18–20). Hypersensitive C-reactive protein (hsCRP), tumor necrosis factor-alpha (TNF-α), and parameters calculated from neutrophil, lymphocyte, and platelet counts including systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) are classical inflammatory markers and closely associated with chronic diseases (19–24). Galectin-3 (Gal-3), a recently identified novel inflammatory biomarker as well as AF biomarker, is involved in inflammation, fibrogenesis, and atrial electrical and structural remodeling (25,26). Although chronic inflammatory response was demonstrated to be associated with AF and hypertension (18–20), the role of inflammatory markers in the association between AF and MHT is rarely reported.

Thus, we conducted a prospective, observational, and longitudinal study to investigate the association between MHT and new-onset AF and the mediating effect of subclinical and low-grade inflammation on this association in a community-dwelling elderly population.

Materials and methods

Study population

The data of the participants were from a prospective and population-based cohort study as elsewhere (27). This cohort study has been retrospectively registered with ChiCTR.org.cn, number ChiCTR-EOC-17013598. In this study, a total of 2040 individuals ≥60 years of age were randomly enrolled from community dwellings between 2008 and 2010 in the Shandong area of China. The response rate was 97.4%. Among them, 1864 participants were eligible and included in the study. Exclusion criteria were as follows: had a plan to leave the study area within 5 years, failure to complete ABPM assessment, AF or atrial flutter, other arrhythmia such as supraventricular tachycardia, pacemaker, congestive heart failure, myocardial infarction, acute and chronic infectious diseases, using immunosuppressant, liver or renal dysfunction, malignancy, drug and alcohol abuse, and unwillingness to provide informed consent. During follow-up, 75 participants were excluded because 59 were lost and 16 were unexplained sudden death. Finally, 1789 participants were included in this study and used for further analyses (Figure 1). This study was conducted in compliance with the Declaration of Helsinki and was approved by the Research Ethics Committee of the Institute of Basic Medicine, Shandong Academy of Medical Sciences, Jinan, China. Written informed consent was obtained from each participant.

Figure 1. The protocol flowchart.

Morning blood pressure assessment

In this study, morning blood pressure was assessed using noninvasive 24-hour ABPM with validated oscillometric device (TM-2421 and TM-2430; A&D Co. Inc., Tokyo, Japan) as described elsewhere because of well-distinguish the differences between “morning surge” and “sustained nocturnal and morning hypertension” (27). With the help of family physicians, the cuff of the device was positioned on the nondominant upper limb by trained and experienced nurses who blinded to the clinical data of participants in the clinics or the participant’s home. After 24 hours monitoring, the device was detached by a nurse with the help of a family physician to download the data. The BP was recorded at 15 min intervals during the daytime (0600–2200) and 20 min intervals during the nighttime (2200–0600). Participants were encouraged to engage in normal daily activities during daytime and instructed to sleep during nighttime. Daily activities and the times of sleep and wake were asked to record in a diary. The acceptable measurement should be those BP recordings with more than 70% daytime or nighttime measurements and more than 20 valid readings with two or over valid readings per hour while awake (27,28). Morning BP was assessed using the records of BP during the period from wake time recorded of the diary to 0900 a.m. in this study because BP rises after waking up in the morning (29,30). If the participants were sleeping until 0900 a.m., the 24-hour ABP will be retested within the next 7 days, otherwise the participants will be excluded from the study. According to the results of 24-hour ABPM, participants were classified into non-MHT and MHT groups. In addition, the MHT group was further categorized into MS and SNMH subgroups (12,14,31). MS was defined as that the difference between the average SBP in the duration from wake time to 0900 a.m. and the average SBP of the three readings adjacent to the lowest SBP reading during the sleep period was 35 mm Hg and over (27,32).

Subclinical inflammation assessment

Blood samples were obtained from antecubital vein using tubes comprising standardized EDTA-K2 after overnight fasting and tested within 30 min after collection. Complete blood count including absolute counts of white blood cell, neutrophil, lymphocyte, and platelet was detected using an automated analyzer (Sysmex XN-1000; Sysmex Corporation, Kobe, Japan) in accordance with the instructions. The plasma levels of hsCRP, TNF-α, and Gal-3 were determined using ELISA kits (Beyotime, Shanghai, China) following the manufacturer's instructions. The minimum detectable concentrations were 1.4 ng/mL for hsCRP, 4.0 pg/mL for TNF-α, and 62 pg/mL for Gal-3. Intra- and inter-assay coefficients of variation for ELISAs were <5% and <10%, respectively, for hsCRP and TNF-α, and both <10% for Gal-3. SII, NLR, and PLR were calculated using the following formulas (19,20): SII = (absolute neutrophil count × absolute platelet)/absolute lymphocyte count, NLR = absolute neutrophil count/absolute lymphocyte count, and PLR = absolute platelet/absolute lymphocyte count.

Covariates

The possible covariates included age, sex, smoking, alcohol consumption, exercise, body mass index, hypertension medication, history of diabetes and medication, history of dyslipidemia and medication, fasting plasma glucose (FPG), total cholesterol (TCHO), triglycerides, low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C). Common carotid artery intima-media thickness (CCA-IMT) was also included as covariates in this study. CCA-IMT was assessed using a high-resolution ultrasound (Vivid i, GE Medical Systems Ultrasound Israel Ltd.).

Outcome

The primary study outcome was the time until the first occurrence of AF during the follow-up period. The new-onset AF was determined by the electrocardiogram test at annual follow-up visit and by ICD-10 code I48 (I48.0–I48.9) at hospitalization or specialized outpatient visit.

Statistical analyses

Continuous data are expressed as mean standard deviation (SD) or median with interquartile range (IQR; 25th and 75th percentiles) depending on the normality and categorical data as frequency with percentages. The normality of the continuous data was determined using the Kolmogorov–Smirnov test. The differences of continuous data were compared by independent samples t-test or one-way analysis of variance (ANOVA) with post hoc Bonferroni tests for normally distributed data, whereas by Mann-Whitney U-test or Kruskal–Wallis test with post hoc Bonferroni tests for nonnormally distributed data between two groups or among more than two groups. The differences of categorical data were determined by chi square test. Correlations between morning BP and inflammatory files were determined using Pearson’s or Spearman’s correlation coefficients when appropriate. Then, the independent association between morning BP and inflammatory files was identified by a multiple linear backward stepwise regression analysis. The Kaplan-Meier method with log-rank test was conducted to estimate cumulative incidence of AF and compare times to AF. The Cox proportional hazards model was used to assess the hazard ratio (HR) and 95% confidence interval (CI). For the association between AF and inflammation, participants were classified into low, moderate, and high groups separately according to the tertile of hsCRP, TNF-α, SII, and Gal-3. For the association between AF and morning BP, participants were classified into non-MHT and MHT groups. Participants with MHT were further subgrouped into MS and SNMH to assess the effects of different types of MHT on new-onset AF and inflammatory level. SPSS Hayes process (version 4.0) was conducted to further assess the mediating role of SII, NLR, PLR, hsCRP, TNF-α, and Gal-3 in the process of morning BP resulting in AF separately. Model 1 was adjusted for age and sex. Model 2 was adjusted for the confounders in model 1 and smoking, alcohol consumption, exercise, medical history and medications, body mass index, office blood pressure, blood lipids, fasting plasma glucose, and CCA-IMT. Model 3 was adjusted for the confounders in model 2 and hsCRP, TNF-α, SII, and Gal-3. SPSS for Windows software package (version 26.0, SPSS Inc) was used to conduct all statistical analyses, and GraphPad Prism (version 9.1.0, GraphPad Software) was used to create graphs. A two-sided P < .05 was considered statistically significant.

Results

Baseline demographic and clinical characteristics

Figure 1 summarizes the protocol flowchart and Table 1 details the demographic and clinical characteristics of the participants. Among 1789 participants, 892 (49.9%) were female, 713 (39.9%) were non-MHT, 507 (28.3%) were MS, and 569 (31.8%) were SNMH. The mean age was 67.54 (SD: 5.41) years. The files of 24-hour ABPM are presented in Figure 2.

Figure 2. Details of 24-hour ambulatory blood pressure files. The results are presented as the mean with standard deviation. SBP indicates systolic blood pressure; DBP, diastolic blood pressure.

Table 1. Baseline demographic and clinical characteristics.

	Value
Female, n (%)	892 (49.9)
Age, years	67.54 ± 5.41
Smoking, n (%)	415 (23.2)
Alcohol consumption, n (%)	623 (34.8)
Lacking exercise, n (%)	1075 (60.1)
Hypertension, n (%)	1301 (72.7)
Antihypertensive medication, n (%)	1011 (56.5)
Diuretics, n (%)	443 (24.8)
Beta-blocker, n (%)	265 (14.8)
Calcium channel blocker, n (%)	247 (13.8)
Angiotensin-converting enzyme inhibitor, n (%)	248 (13.9)
Angiotensin receptor blocker, n (%)	194 (10.8)
Diabetes Mellitus, n (%)	248 (13.9)
Lowering-glucose medication, n (%)	244 (13.6)
Dyslipidemia, n (%)	917 (51.3)
Anti-dyslipidemia medication, n (%)	306 (17.1)
Body mass index, kg/m^2	24.74 ± 2.52
Office SBP, mm Hg	145.03 ± 15.77
Office DBP, mm Hg	71.14 ± 7.93
Heart beat, pbm	70.40 ± 8.05
TCHO, mmol/L	4.85 ± 0.64
TG, mmol/L	1.58 ± 0.53
HDL-C, mmol/L	1.16 ± 0.33
LDL-C, mmol/L	2.97 ± 0.52
FPG, mmol/L	5.51 ± 1.42
hsCRP, mg/dL	0.69 ± 0.28
TNF-α, pg/mL	59.62 ± 20.05
WBC, cells × 10^3/µL	6.09 ± 0.88
Neutrophil, cells × 10^3/µL	4.02 ± 0.82
Lymphocyte, cells × 10^3/µL	2.07 ± 0.50
Platelets, cells × 10^3/µL	273.68 ± 27.49
SII	574.84 ± 231.01
NLR	2.08 ± 0.73
PLR	140.86 ± 40.56
Gal-3 (ng/mL)	12.28 (11.39 to 13.15)
CCA-IMT, mm	1.22 ± 0.28
Non-MHT, n (%)	713 (39.9)
MHT, n (%)	1076 (60.1)
MS, n (%)	507 (28.3)
SNMH, n (%)	569 (31.8)

The results are presented as the mean±SD, median with interquartile range, or frequencies and percentages. SBP indicates systolic blood pressure; DBP, diastolic blood pressure; TCHO, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; FPG, fasting plasma glucose; hsCRP, high-sensitivity C-reactive protein; TNF-α, tumor necrosis factor-alpha; WBC, white blood cell; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; Gal-3, Galectin-3; CCA-IMT, common carotid artery intima-media thickness; MHT, morning hypertension; MS, morning surge; SNMH, sustained nocturnal and morning hypertension.

Association between morning blood pressure and atrial fibrillation

Over an average 129.0 (IQR, 122.0 to 135.0) months of follow-up, 144 participants (7.49 per 1000 person-years) developed AF. Among them, 33 (4.31 per 1000 person-years) were in the non-MHT group, 111 (9.60 per 1000 person-years) in the MHT group. After adjustment for confounders including office BP, antihypertensive medication, hsCRP, TNF-α, SII, and Gal-3, the cumulative hazard ratio of new-onset AF was 1.385 (95% CI: 1.005 to 1.909, P_adjusted = 0.027) in the MHT group and 2.474 (95% CI: 1.408 to 4.348, P_adjusted = 0.002) in the SNMH subgroup compared with the non-MHT group (Figure 3a and Supplemental Table S1). However, the difference in the risk of new-onset AF was not significant between the MS subgroup and the non-MHT group after adjustment for confounders, although the HR was 1.284 (95% CI: 0.994 to 1.659, P_adjusted = 0.053). We also investigated the effect of one-SD increment in the morning systolic and diastolic BP on the new-onset AF. The results demonstrated that the HR was 1.166 (95% CI: 1.072 to 1.268, P_adjusted = 0.015) with one-SD increment of systolic BP and 1.025 (95% CI: 0.880 to 1.194, P_adjusted = 0.848) with one-SD increment of diastolic BP after adjustment for confounders (Supplemental Table S2). In addition, we found that the risk of new-onset AF was higher in the SNMH subgroup than that in the MS subgroup (P = .045, Figure 3b). The differences in the risk of new-onset AF between the SNMH and MS subgroups remained after adjustment for confounders (HR: 1.413, 95% CI: 1.006 to 1.984, P = .049, Supplemental Table S1).

Figure 3. Differences in the cumulative new-onset atrial fibrillation between the groups divided by morning blood pressure during the follow-up period. (a) Difference in the cumulative new-onset atrial fibrillation in participants grouped by morning blood pressure. (b) Difference in the cumulative new-onset atrial fibrillation between participants with non-MHT and MS. (c) Difference in the cumulative new-onset atrial fibrillation between participants with non-MHT and SNMH. (d) Difference in the cumulative new-onset atrial fibrillation in patients grouped by sub-type MHT. MHT indicates morning hypertension; MS, morning surge; SNMH, sustained nocturnal and morning hypertension.

Associations of morning blood pressure with subclinical inflammation

The levels of hsCRP, TNF-α, SII, NLR, PLR, and Gal-3 were significantly higher in the MHT group compared with the non-MHT group (all P < .001, Figure 4). The results of correlation and linear regression analyses showed that morning SBP and DBP were significantly and positively associated with hsCRP, TNF-α, SII, NLR, PLR, and Gal-3 even after adjustment for confounders including office BP, antihypertensive medication, and CCA-IMT (all P_adjusted < 0.001, Table 2 and Supplemental Table S3).

Figure 4. Differences in the level of inflammation between non-MHT and MHT groups. (a) Difference in hsCRP level. (b) Difference in TNF-α level. (c) Difference in SII. (d) Difference in NLR. (e) Difference in PLR. (f) Difference in Gal-3 level. MHT indicates morning hypertension; hsCRP, hypersensitive C-reactive protein; TNF-α, tumor necrosis factor-alpha; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; Gal-3, Galectin-3.

Table 2. Association between morning blood pressure and subclinical inflammation in all participants.

	Morning SBP (mm Hg)		Morning DBP (mm Hg)
	Beta (95% CI)	P value	Beta (95% CI)	P value
hsCRP (mg/dL)
Model 1	0.008 (0.008, 0.009)	<0.001	0.006 (0.005, 0.008)	<0.001
Model 2	0.008 (0.007, 0.009)	<0.001	0.005 (0.004, 0.007)	<0.001
TNF-α (pg/mL)
Model 1	0.490 (0.433, 0.548)	<0.001	0.340 (0.247, 0.432)	<0.001
Model 2	0.480 (0.422, 0.537)	<0.001	0.298 (0.205, 0.391)	<0.001
SII
Model 1	7.934 (7.321, 8.548)	<0.001	5.778 (4.727, 6.829)	<0.001
Model 2	7.562 (6.912, 8.212)	<0.001	4.927 (3.885, 5.970)	<0.001
NLR
Model 1	0.023 (0.021, 0.025)	<0.001	0.016 (0.013, 0.020)	<0.001
Model 2	0.023 (0.021, 0.025)	<0.001	0.014 (0.010, 0.017)	<0.001
PLR
Model 1	1.310 (1.200, 1.420)	<0.001	1.062 (0.878, 1.246)	<0.001
Model 2	1.257 (1.138, 1.376)	<0.001	0.934 (0.751, 1.116)	<0.001
Gal-3 (ng/mL)
Model 1	0.057 (0.051, 0.062)	<0.001	0.054 (0.045, 0.063)	<0.001
Model 2	0.047 (0.039, 0.055)	<0.001	0.039 (0.028, 0.050)	<0.001

Model 1 was adjusted for age and sex.

Model 2 was adjusted for confounders in model 1 and smoking, alcohol consumption, exercise, medical history and medications, body mass index, office blood pressure, blood lipids, fasting plasma glucose, and CCA-IMT.SBP indicates systolic blood pressure; DBP, diastolic blood pressure; hsCRP, hypersensitive C-reactive protein; TNF-α, tumor necrosis factor-alpha; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; Gal-3, Galectin-3; HR, hazard ratio; CI, confidence interval.

We also assessed the differences of the effect of the different MHT types on the inflammatory response. Supplemental Figure S1 showed that the levels of hsCRP, SII, NLR, PLR, and Gal-3 were significantly higher in the SNMH subgroup than those in the MS subgroup (all P < .05).

High levels of subclinical inflammation increase the risk of new-onset atrial fibrillation

To assess the associations between subclinical inflammatory response and the new-onset AF, we grouped the participants according to the tertiles of hsCRP, TNF-α, SII, NLR, PLR, and Gal-3 separately. The risks of new-onset AF were significantly increased from the low to moderate to high groups after adjustment for confounders (all P_trend < 0.001, Figure 5 and Supplemental Table S4). The HR of new-onset AF was 2.289 for one-SD increment in hsCRP, 2.037 for one-SD increment in TNF-α, 2.078 for one-SD increment in SII, 2.081 for one-SD increment in NLR, 2.026 for one-SD increment in PLR, and 3.248 for one-SD increment in Gal-3 (all P_adjusted < 0.001, Supplemental Table S5).

Figure 5. Differences in the cumulative new-onset atrial fibrillation among the groups divided by the tertiles of inflammatory response during the follow-up period. (a) Difference in the cumulative new-onset atrial fibrillation in patients grouped by the tertile of hsCRP. (b) Difference in the cumulative new-onset atrial fibrillation in patients grouped by the tertile of TNF-α. (c) Difference in the cumulative new-onset atrial fibrillation in patients grouped by the tertile of SII. (d) Difference in the cumulative new-onset atrial fibrillation in patients grouped by the tertile of NLR. (e) Difference in the cumulative new-onset atrial fibrillation in patients grouped by the tertile of PLR. (f) Difference in the cumulative new-onset atrial fibrillation in patients grouped by the tertile of Gal-3. hsCRP indicates hypersensitive C-reactive protein; TNF-α, tumor necrosis factor-alpha; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; Gal-3, Galectin-3.

Mediating effect of subclinical inflammation on the association between morning hypertension and atrial fibrillation

We conducted Hayes process analysis to further assess the effect of subclinical inflammation on the process of MHT leading to AF. The results showed that hsCRP, TNF-α, SII, NLR, PLR, and Gal-3 separately play a significant and indirect effect on the association of new-onset AF with morning systolic and diastolic BP after adjusted confounders (all P_adjusted < 0.05, Figure 6 and Supplemental Table S6).

Figure 6. Mediating effect of inflammatory response on the process of morning hypertension leading to atrial fibrillation. SBP indicates systolic blood pressure; DBP, diastolic blood pressure; hsCRP, hypersensitive C-reactive protein; TNF-α, tumor necrosis factor-alpha; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; Gal-3, Galectin-3. Circle and square dots indicate mediating effect and bars indicate 95% confidence interval.

Discussion

The main findings in this longitudinal and prospective cohort study were those: (1) MHT was associated with an increased risk of new-onset AF, and SNMH was more associated with new-onset AF than MS in older individual; (2) either MS or SNMH was associated with high level of subclinical inflammatory response; (3) subclinical inflammation increased the risk of new-onset AF; and (4) subclinical inflammation played a mediating effect in the process of MHT inducing AF.

The association between hypertension and cardiovascular events has been well established (1,6,7). A cohort study followed for 35 years demonstrated that baseline SBP of 128 mm Hg or higher was associated with a 1.5-fold and diastolic BP of 80 mm Hg or higher was associated with a 1.79-fold higher risk of incident AF (33). However, the data to investigate the association between MHT and the risk of AF are limited (11,15,34). In this study, we found that the risk of new-onset AF was 1.39 folds in patients with MHT compared with those with non-MHT. One-SD increment in morning systolic BP was associated with 1.17 folds risk of new-onset AF even after adjustment for confounders. The findings indicate that MHT was an important and independent risk factor of new-onset AF in the older individuals.

Few studies reported that MS, one of the sub-types of MHT, was associated with a higher risk of incident AF (11,15,34). Taş and colleagues (15) assessed MS using official BP measurement and found that the AF incidence was significantly higher in patients with larger MS than those with lower MS. In this study, we determined the sub-types of MHT using 24-hour ABPM. Twenty-four-hour ABPM is more accuracy to assess the circadian rhythm of BP fluctuation than office and home BP monitoring (25,35,36) and identify MHT in untreated as well as antihypertensive patients (37). Our findings are not completely in line with the findings of Taş (15). We found that the significance in the cumulative risk of new-onset AF between non-MHT and MS participants was abolished after the inflammatory markers including hsCRP, TNF-α, SII, and Gal-3 were included as confounders.

We also assessed the association of the risk of AF with SNMH, another sub-type of MHT, and found that the patients with SNMH were more prone to AF compared with those with MS even after adjustment for confounders including office BP and inflammatory markers. The cumulative risks of new-onset AF in SNMH patients were 2.47 and 1.41 folds higher than those in non-MHT participants and MS patients, respectively. In addition, we found that the association between SNMH and the risk of new-onset AF was significantly limited by subclinical inflammation. It indicates that subclinical and chronic inflammation might play an important role in the association between SNMH as well as MS and the risk of AF.

In a cross-section study including 343 newly diagnosed antihypertensive patients (19), the level of SII, NLR, and PLR was found to be higher in the patients with high-value MS than those in the patients with low-value MS. In this study, we assessed the subclinical inflammatory response using SII, NLR, PLR, hsCRP, TNF-α, and Gal-3. The results demonstrated that the levels of these inflammatory markers were higher in the MHT group including patients with either MS or SNMH compared with the non-MHT group. In addition, we also found that the levels of SII, NLR, PLR, hsCRP, and Gal-3 were higher in the SNMH subgroup than those in the MS subgroup. The findings indicate that MHT was associated with a higher level of subclinical inflammatory response. The inflammatory response in patients with SNMH was excessive than that in patients with MS.

The infiltration of immune cells and inflammatory factors in cardiovascular system is associated with AF (18). In this study, subclinical comprehensive inflammation was evaluated using the immune cells assessed by SII, NLR, and PLR as well as inflammatory factors assessed by hsCRP and TNF-α. Meanwhile, a novel inflammatory biomarker Gal-3 was used to assess the subclinical inflammatory response level in the study. Gal-3 is primarily secreted by macrophages, neutrophils, mast cells, and fibroblasts and involving in important regulatory roles in inflammation, adhesion, and fibrosis (25,26,38). Gal-3 has been demonstrated to enhance atrial electrical and structural remodeling and arrhythmogenesis (25,26) and elevated plasma Gal-3 is associated with the increased risk of incident AF (38). As we expected, SII, NLR, PLR, hsCRP, TNF-α, and Gal-3 were independently associated with new-onset AF. With one-SD increment of SII, NLR, PLR, hsCRP, TNF-α, and Gal-3, the risk of new-onset AF was increased more than onefold, respectively.

As immune cells and inflammatory factors can alter the atrial electrophysiology, modulate calcium homeostasis and connexins, and structural substrates, ultimately leading to an increased susceptibility to AF (18), our findings indicate that subclinical and low-grade inflammation might play an important mediating role in the process of MHT inducing AF. The results of process analyses confirmed this mediating effect of subclinical inflammation on the association between high morning BP and the risk of AF. In addition, the findings also explained the different effects between MS and SNMH on new-onset AF. The different levels of subclinical and low-grade inflammatory response induced by MS and SNMH might be the main causes.

The strength of this study was that we elucidated the association between new-onset AF and MHT including MS and SNMH and explored the mediating effect of subclinical and low-grade inflammation on the process of MHT leading to AF. It provides an insight into the mechanism of MHT induced AF. The prospective longitudinal cohort design with long-term follow-up was also a strength of our study. Third, 24-hour ABPM was used to assess the level of morning BP and determine the type of MHT in this study.

However, several limitations must be addressed. First, we used 35 mm Hg as the cutoff to determine MS in this study. It may induce a selection bias of the results, because there is no definite cutoff value for MS (19,27). Second, the changes in morning BP as well as subclinical inflammatory files during the follow-up period were not assessed, although the antihypertensive medication was adjusted as a confounder in the analysis models. This may lead to a bias in the results. Third, the participants in this study were mainly recruited from Han nationality in the Shandong area. Evidences have shown that there are differences in the characteristics of the MHT and AF among ethnic groups (39–41). In addition, we did not consider the effects of social and economic factors on new-onset AF (42–44).

Conclusions

In conclusion, our findings suggested that MHT is an independent and important risk factor of new-onset AF in older adults. SNMH is more associated with new-onset AF compared with MS. The subclinical inflammatory response might play an important mediation role in the association between MHT and new-onset AF. However, the findings should be validated in other populations including other ethnicities and social determinants.

Abbreviations

AF, atrial fibrillation; BP, blood pressure; SBP, systolic blood pressure; MHT, morning hypertension; HBPM, home blood pressure monitoring; ABPM, ambulatory blood pressure monitoring; MS, morning surge; SNMH, sustained nocturnal and morning hypertension; SII, systemic immune-inflammation index; NLR, neutrophil-to-lymphocyte ratio; PLR, platelet-to-lymphocyte ratio; hsCRP, hypersensitive C-reactive protein; TNF-α, tumor necrosis factor-alpha; Gal-3, Galectin-3; FPG, fasting plasma glucose; TCHO, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; CCA-IMT, common carotid artery intima-media thickness; SD, standard deviation; IQR, interquartile range; HR, hazard ratio; CI, confidence interval.

Author contributions

Study conception and design: YG, WL, ZL. Acquisition of data, analysis and interpretation of data: JL, ZH, LH, PL, RY, YD, HZ, YG, WL, ZL. Drafting the article: JL, ZH, LH. Final approval of the version of the article to be published: all authors, and all authors agree to be accountable for all aspects of the work.

Acknowledgments

We would like to thank all participants, general practitioners, and nurses who provided assistance in this study.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/10641963.2023.2253381.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China grants number 81973139; Shandong Provincial Natural Science Foundation of China grants number ZR2022MH179; College Students’ innovation and entrepreneurship training program grants number 2022104391161.