Abstract
Background
The association between lung function and headache risk remains unclear. This study aims to explore the association between lung function and headache risk through cross-sectional and longitudinal analyses.
Methods
This study used data from the China Health and Retirement Longitudinal Study (CHARLS) spanning 2015 to 2020 to conduct cross-sectional and longitudinal analyses. The cross-sectional analysis included 10,917 middle-aged and older adults aged ≥ 45 years. For the longitudinal component, 5,194 participants from this cohort who had no history of headache diagnosis in the 2015 cross-sectional survey were followed up until 2020. The cross-sectional association between predicted PEF% and headache was examined using multivariate logistic regression and restricted cubic spline (RCS) analysis. For the longitudinal association, Kaplan–Meier curves, multivariable Cox proportional hazards regression models, and RCS analysis were applied to assess the relationship between baseline predicted PEF% levels and the risk of new-onset headache. Additionally, subgroup analyses were performed to explore the consistency of these associations across different subgroups.
Results
The results of the multivariate logistic regression analysis showed that the probability of headache decreased with increasing PEF% predicted (P < 0.05); compared with the Q1 group, the Q4 group had a lower probability of experiencing headaches (OR = 0.81, 95% CI 0.68–0.96, P = 0.016). During follow-up, a total of 2,428 people (46.7%) developed new-onset headaches. The results of the multivariable Cox regression analysis showed that an increase in PEF% predicted was a protective factor for the occurrence of headaches, and as PEF% predicted increased, the risk of headaches decreased (P < 0.05); compared with the q1 group, the q4 group had a lower risk of headache (HR = 0.94, 95% CI 0.90–0.97, P < 0.001). Subgroup analysis results showed that gender and PEF% predicted had an interaction effect on headache risk (P = 0.005).
Conclusions
In middle-aged and elderly people, especially middle-aged and elderly women, improving PEF levels may have potential value in the primary prevention of headaches.
Supplementary Information
The online version contains supplementary material available at 10.1186/s40001-025-03069-2.
Keywords: Lung function, PEF% predicted, Headache, Middle-aged and elderly people
Background
Chronic lung disease (CLD) encompasses chronic bronchitis, chronic obstructive pulmonary disease, asthma, and other conditions. Its incidence rises significantly with age, posing a serious threat to patients’ health and imposing a heavy burden on the economy and healthcare system [1]. Notably, headache is a common symptom in patients with CLD. This type of headache is known as pulmonary migraine, and its pathophysiological mechanisms are thought to be closely linked to inflammation and immune dysfunction caused by the disease [2]. In-depth exploration of the pathological mechanisms underlying CLD-related headaches necessitates assessment of lung function, a core physiological indicator [3].
Lung function is a key indicator of respiratory health. Peak expiratory flow (PEF) refers to the maximum instantaneous flow rate during forced expiration in a forced vital capacity maneuver. As a convenient and easily measurable lung function parameter, it can effectively screen for declining lung function in adults aged 40 years and older [4, 5]. According to guidelines from the American Thoracic Society and the European Respiratory Society, PEF can sensitively detect central airway obstruction caused by premature airway aging and bronchoconstriction. Compared with forced expiratory volume in the first second (FEV₁) and forced vital capacity (FVC), PEF offers advantages in the early identification of central and upper airway obstruction [6]. However, it remains unclear whether reduced PEF is associated with an elevated risk of headaches. Additionally, while previous studies have noted a higher incidence of headaches in asthma patients with poor lung function [7], few large-population studies have examined the relationship between PEF and headache risk after comprehensive adjustment for covariates.
This study uses data from CHARLS to examine the association between lung function and headache risk in middle-aged and elderly Chinese adults.
Methods
Study sample
The data for this study were obtained from CHARLS, a database that employs a random multistage stratified probability proportional sampling method. It is a nationwide longitudinal survey of community-dwelling residents aged 45 years and older across 28 provinces in China. The national baseline survey of CHARLS was conducted in 2011 (first wave). To date, four follow-up surveys have been administered every 2 years (second wave in 2013, third wave in 2015, fourth wave in 2018, and fifth wave in 2020) [8]. CHARLS was approved by the Biomedical Ethics Review Committee of Peking University (IRB 00001052-11015), and all participants provided written informed consent. This study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.
This study used CHARLS data from 2015 to 2020 for cross-sectional and longitudinal analyses, with 2015 as the baseline. A total of 21,095 participants were initially involved. Based on the following exclusion criteria, 10,917 participants were included in the cross-sectional analysis: (1) Age < 45 years or missing age information (n = 1645); (2) Missing PEF data (n = 4369); (3) Missing headache-related information (n = 96); (4) Missing or abnormal covariate data (n = 4068). For the longitudinal study, 5194 middle-aged and elderly individuals without a headache diagnosis in the 2015 cross-sectional survey were included and followed up until 2020. The sample selection process is illustrated in Fig. 1.
Fig. 1.
Flowchart of the study participants
Measurement of PEF and calculation of PEF% predicted
PEF was measured and recorded by trained technicians using a peak flow meter with a disposable mouthpiece. Participants were instructed to take a deep breath and then exhale as forcefully and rapidly as possible. Three measurements were taken, and the highest value was used for data analysis. PEF values are influenced by individual differences, with normal ranges varying by age, gender, and height. To further improve the accuracy of lung function assessment, we analyzed the ratio of actual PEF values to predicted values (PEF% predicted) [9]. Predicted PEF values were calculated using the formula recommended by Academician Zhong Nanshan [10] for Chinese adults: for males, the formula is 75.6 + 20.4 × age − 0.41 × age2 + 0.002 × age3 + 1.19 × height; for females, it is 282.0 + 1.79 × age − 0.046 × age2 + 0.68 × height.
Assessment of headache
Headaches were assessed using the health status module of the CHARLS questionnaire: First, participants were asked, “Do you often experience discomfort due to physical pain?” and required to answer “yes” or “no.” For those who answered “yes,” they were further asked, “Please indicate all areas of your body where you feel pain.” If participants reported head pain, this was defined as a headache.
Assessment of covariates
Based on previous studies, covariates were initially screened [11–13], followed by collinearity analysis with a threshold of variance inflation factor < 5. Boruta analysis was then used to determine the covariates included in this study (see S1). Potential covariates included age, gender, educational attainment, marital status, place of residence, drinking habits, smoking habits, BMI, sleep duration, CLD medication use, depression, and comorbidities (including hypertension, diabetes, heart disease, stroke, and kidney disease). Among these, age was categorized into two groups: 45–59 years and ≥ 60 years. Educational attainment was classified as illiterate, primary school, lower secondary school, upper secondary school, and tertiary education or above. Place of residence was divided into rural and urban areas. Marital status was categorized as married and single (including unmarried, divorced, or widowed). Smoking status was classified as smoker or non-smoker. Drinking status was categorized as drinker or non-drinker. BMI was divided into three groups: underweight (< 18.5 kg/m2), normal weight (18.5–24.0 kg/m2), and overweight/obese (≥ 24.0 kg/m2). Depression was assessed using the 10-item Center for Epidemiological Studies Depression Scale (CES-D10), with a total score of 30 points. Higher scores indicated more severe depressive symptoms, and a score of ≥ 10 points was used as the criterion for identifying depression [14]. The diagnosis of comorbidities was based on both self-reported physician diagnoses and current medication use.
Statistical analyses
Statistical analyses were performed using the R statistical software package (http://www.R-project.org, R Foundation) and Free Statistics software version 2.1. Quantitative data conforming to a normal distribution are expressed as mean ± standard deviation (SD), with intergroup comparisons performed using independent samples t-tests. Quantitative data not conforming to a normal distribution are expressed as M (P25, P75), with intergroup comparisons performed using the rank-sum test. Categorical variables are expressed as frequencies (percentages), with intergroup comparisons performed using the chi-square test. Multivariate logistic regression and restricted cubic spline (RCS) curve analyses were used to examine the cross-sectional association between PEF% predicted and headache, and subgroup analyses were conducted to assess the consistency of this association. Kaplan–Meier curves, multivariable Cox proportional hazards regression models, and RCS curves were used to analyze the longitudinal association between baseline PEF% predicted levels and the risk of new-onset headaches in 2015. Subgroup analyses were performed to explore the moderating effects of sociodemographic characteristics and health status on the association between PEF% predicted and the risk of new-onset headaches. A P-value < 0.05 was considered statistically significant.
Results
Baseline characteristics of the study population
This study ultimately included 10,917 participants in the cross-sectional analysis. Comparisons between the headache group (n = 1648) and non-headache group (n = 9269) revealed no statistically significant differences in age or BMI (both P > 0.05). However, significant differences were observed in gender, place of residence, educational attainment, marital status, smoking status, drinking status, comorbidities, depression, CLD medication use, sleep duration, and PEF (all P < 0.05). Details are presented in Table 1.
Table 1.
Baseline characteristics of the study population in 2015
| Variables | Total (n = 10,917) | Non-headache (n = 9269) | Headache (n = 1648) | P | Statistic |
|---|---|---|---|---|---|
| Age, n (%) | |||||
| 45–59 | 5307 (48.6) | 4542 (49) | 765 (46.4) | 0.053 | 3.735 |
| ≥ 60 | 5610 (51.4) | 4727 (51) | 883 (53.6) | ||
| Gender, n (%) | |||||
| Male | 4796 (43.9) | 4350 (46.9) | 446 (27.1) | < 0.001 | 224.223 |
| Female | 6121 (56.1) | 4919 (53.1) | 1202 (72.9) | ||
| Residence, n (%) | |||||
| Rural | 8035 (73.6) | 6694 (72.2) | 1341 (81.4) | < 0.001 | 60.32 |
| Urban | 2882 (26.4) | 2575 (27.8) | 307 (18.6) | ||
| Education, n (%) | |||||
| Illiterate | 4740 (43.4) | 3783 (40.8) | 957 (58.1) | < 0.001 | 234.107 |
| Primary school | 2544 (23.3) | 2157 (23.3) | 387 (23.5) | ||
| Lower secondary school | 2396 (21.9) | 2173 (23.4) | 223 (13.5) | ||
| Upper secondary school | 1032 (9.5) | 954 (10.3) | 78 (4.7) | ||
| Tertiary education or above | 205 (1.9) | 202 (2.2) | 3 (0.2) | ||
| Marital status, n (%) | |||||
| Married | 9593 (87.9) | 8196 (88.4) | 1397 (84.8) | < 0.001 | 17.534 |
| Single | 1324 (12.1) | 1073 (11.6) | 251 (15.2) | ||
| Smoking status, n (%) | |||||
| Smoker | 4164 (38.1) | 3710 (40) | 454 (27.5) | < 0.001 | 92.327 |
| Non-smoker | 6753 (61.9) | 5559 (60) | 1194 (72.5) | ||
| Drinking status, n (%) | |||||
| Drinker | 4816 (44.1) | 4202 (45.3) | 614 (37.3) | < 0.001 | 37.023 |
| Non-drinker | 6101 (55.9) | 5067 (54.7) | 1034 (62.7) | ||
| BMI, n (%) | |||||
| Underweight | 573 (5.2) | 468 (5) | 105 (6.4) | 0.061 | 5.584 |
| Normal weight | 5720 (52.4) | 4852 (52.3) | 868 (52.7) | ||
| Overweight/obesity | 4624 (42.4) | 3949 (42.6) | 675 (41) | ||
| Hypertension, n (%) | |||||
| No | 7705 (70.6) | 6720 (72.5) | 985 (59.8) | < 0.001 | 109.2 |
| Yes | 3212 (29.4) | 2549 (27.5) | 663 (40.2) | ||
| Diabete, n (%) | |||||
| No | 10,049 (92.0) | 8598 (92.8) | 1451 (88) | < 0.001 | 42.497 |
| Yes | 868 (8.0) | 671 (7.2) | 197 (12) | ||
| Heart attack, n (%) | |||||
| No | 9255 (84.8) | 8053 (86.9) | 1202 (72.9) | < 0.001 | 210.798 |
| Yes | 1662 (15.2) | 1216 (13.1) | 446 (27.1) | ||
| Stroke, n (%) | |||||
| No | 10,301 (94.4) | 8817 (95.1) | 1484 (90) | < 0.001 | 67.687 |
| Yes | 616 (5.6) | 452 (4.9) | 164 (10) | ||
| Kidney disease, n (%) | |||||
| No | 9995 (91.6) | 8626 (93.1) | 1369 (83.1) | < 0.001 | 180.688 |
| Yes | 922 (8.4) | 643 (6.9) | 279 (16.9) | ||
| Depression, n (%) | |||||
| No | 7198 (65.9) | 6743 (72.7) | 455 (27.6) | < 0.001 | 1269.265 |
| Yes | 3719 (34.1) | 2526 (27.3) | 1193 (72.4) | ||
| Medicine, n (%) | |||||
| No | 10,222 (93.6) | 8804 (95) | 1418 (86) | < 0.001 | 187.589 |
| Yes | 695 (6.4) | 465 (5) | 230 (14) | ||
| Sleep duration (h) | 6.4 ± 1.9 | 6.5 ± 1.8 | 5.5 ± 2.2 | < 0.001 | 363.875 |
| PEF (L/min), Med (IQR) | 358.1 (321.5, 435.2) | 360.5 (323.7, 443.9) | 343.7 (312.1, 370.5) | < 0.001 | 164.144 |
BMI is classified into three groups: underweight (< 18.5 kg/m2), normal weight (18.5–24.0 kg/m2), and overweight/obese (≥ 24.0 kg/m2). Medicine represents CLD medication use. PEF stands for peak expiratory flow
Cross-sectional data analysis of the association between PEF% predicted and headache
Multivariate logistic regression analysis of PEF% predicted and headache
A multivariate logistic regression analysis was conducted with headache occurrence as the dependent variable and PEF% predicted and its quartiles as independent variables. The quartile groups were defined as follows: Q1 (< 66.0%, n = 2720), Q2 (66.0%–84.1%, n = 2723), Q3 (84.2%–101.5%, n = 2740), and Q4 (≥ 101.6%, n = 2734). Three models were constructed: Model 1 was adjusted for age and gender; Model 2 further adjusted for educational attainment, place of residence, marital status, BMI, smoking status, drinking status, sleep duration, and CLD medication use on the basis of Model 1; Model 3 additionally adjusted for hypertension, diabetes, heart disease, stroke, kidney disease, and depression on top of Model 2. The results showed that PEF% predicted was significantly negatively correlated with headache. As PEF% predicted increased, the probability of headache decreased (P < 0.05) as shown in Table 2.
Table 2.
Multivariate logistic regression analysis of PEF% predicted and headache
| Variables | Model 1 | Model2 | Model3 | |||
|---|---|---|---|---|---|---|
| OR (95%CI) | P-value | OR (95%CI) | P-value | OR (95%CI) | P-value | |
| PEF% predicted | 0.41(0.34, 0.51) | < 0.001 | 0.64(0.51, 0.79) | < 0.001 | 0.76 (0.60, 0.95) | 0.018 |
| PEF% predicted quartile | ||||||
| Q1 (< 66.0%) | 1 (Ref) | 1 (Ref) | 1 (Ref) | |||
| Q2 (66.0%–84.1%) | 0.84 (0.73, 0.97) | 0.017 | 0.98 (0.84, 1.13) | 0.766 | 1.01 (0.86, 1.18) | 0.916 |
| Q3 (84.2%–101.5%) | 0.67(0.58, 0.78) | < 0.001 | 0.83 (0.71, 0.97) | 0.020 | 0.90 (0.77, 1.07) | 0.227 |
| Q4 (≥ 101.6%) | 0.57 (0.49, 0.66) | < 0.001 | 0.73 (0.62, 0.86) | < 0.001 | 0.81 (0.68, 0.96) | 0.016 |
| Trend.test | 0.83 (0.79, 0.87) | < 0.001 | 0.90 (0.85, 0.94) | < 0.001 | 0.93 (0.88, 0.98) | 0.009 |
Model 1 adjusted for age, gender
Model 2 adjusted for age, gender, education, residence, marital status, BMI, smoking status, drinking status, sleep duration, medicine
Model 3 adjusted for age, gender, education, residence, marital status, BMI, smoking status, drinking status, sleep duration, medicine, hypertension, diabetes heart attack, stroke, kidney disease, depression
RCS analysis of PEF% predicted and headache
After fully adjusting the model, the RCS curve was further used to explore the dose–response relationship between PEF% predicted and headache. The results showed that there was no nonlinear association between PEF% predicted and headache (P for overall = 0.034, P for nonlinear = 0.323). The higher the PEF% predicted, the lower the probability of headache, as shown in Fig. 2.
Fig. 2.
Dose–response relationship between predicted PEF% and headache
Subgroup analysis of PEF% predicted associated with headache
Subgroup analysis and interaction tests revealed that an increase in PEF% predicted was consistently associated with a decreased probability of headache. No significant interaction was observed between PEF% predicted and headache across subgroups stratified by population characteristics (all P > 0.05) (see S2), indicating that the effect of PEF% predicted on headache remains relatively stable across different population characteristics.
Longitudinal data analysis of the association between PEF% predicted and the risk of new-onset headaches
Kaplan–Meier cumulative incidence curve analysis of PEF% predicted and the risk of new-onset headache
By the end of the follow-up period, a total of 2,428 middle-aged and elderly individuals (46.7%) were newly diagnosed with headaches. Participants were grouped into quartiles based on PEF% predicted values: the q1 group (< 67.0%, n = 1296), q2 group (66.0–84.1%, n = 1296), q3 group (84.2%–101.5%, n = 1296), and q4 group (≥ 101.6%, n = 1303). The number of new headache cases in groups q1, q2, q3, and q4 was 668, 617, 599, and 544, respectively. Kaplan–Meier cumulative incidence curve analysis showed that the cumulative incidence of new-onset headaches increased with longer follow-up time in all four groups, and there were statistically significant differences in the cumulative incidence of new-onset headaches among the four groups (P < 0.05), as shown in Fig. 3.
Fig. 3.
Comparison of PEF% predicted with the cumulative incidence of new-onset headaches
Multivariate Cox regression analysis of PEF% predicted and the risk of new-onset headache
A multivariate Cox regression analysis was performed with headache occurrence as the dependent variable and PEF% predicted and its quartiles as independent variables. After adjusting for covariates, the results showed that an increase in PEF% predicted was a protective factor against the occurrence of headaches. As PEF% predicted increased, the risk of headaches decreased (P < 0.05) as shown in Table 3.
Table 3.
Multivariate Cox regression analysis of PEF% predicted and the risk of new-onset headache
| Variables | Model 1 | Model2 | Model3 | |||
|---|---|---|---|---|---|---|
| HR (95%CI) | P-value | HR(95%CI) | P-value | HR(95%CI) | P-value | |
| PEF% predicted | 0.68 (0.58, 0.79) | < 0.001 | 0.74 (0.63, 0.86) | < 0.001 | 0.75 (0.64, 0.88) | 0.001 |
| q1 (< 67.0%) | 1 (Ref) | 1 (Ref) | 1 (Ref) | |||
| q2 (67.6%–85.0%) | 0.88 (0.79, 0.98) | 0.022 | 0.90 (0.80, 1.00) | 0.055 | 0.90 (0.80, 1.00) | 0.055 |
| q3 (85.1%–101.9%) | 0.83 (0.74, 0.93) | 0.001 | 0.86 (0.77, 0.97) | 0.012 | 0.87 (0.78, 0.98) | 0.019 |
| q4 (≥ 102.0%) | 0.76 (0.68, 0.85) | < 0.001 | 0.80 (0.71, 0.90) | < 0.001 | 0.80 (0.72, 0.91) | < 0.001 |
| Trend.test | 0.91 (0.88, 0.95) | < 0.001 | 0.93 (0.90, 0.97) | < 0.001 | 0.94 (0.90, 0.97) | < 0.001 |
Model 1 adjusted for age, gender
Model 2 adjusted for age, gender, education, residence, marital status, BMI, smoking status, drinking status, sleep duration, medicine
Model 3 adjusted for age, gender, education, residence, marital status, BMI, smoking status, drinking status, sleep duration, medicine, hypertension, diabetes, heart attack, stroke, kidney disease, depression
RCS analysis of PEF% predicted and the risk of new-onset headache
After fully adjusting the model, the RCS curve was further used to explore the dose–response relationship between PEF% predicted and the risk of new-onset headaches. The results showed that there was no nonlinear association between PEF% predicted and the risk of new-onset headache (P overall = 0.014, P nonlinear = 0.647), as shown in Fig. 4.
Fig. 4.
Dose–response relationship between PEF% predicted and the risk of new-onset headaches
Subgroup analysis of the association between PEF% predicted and the risk of new-onset headache
Subgroup analysis results showed that in individuals aged ≥ 60 years, 45–59 years, females, those living in rural areas, those with hypertension, non-diabetics, non-heart disease patients, non-stroke patients, non-kidney disease patients, and non-depression patients, an increase in PEF% predicted was a protective factor against the risk of new-onset headache (all P < 0.05). In men, people with diabetes, stroke, kidney disease, and depression, no significant association was observed between PEF% predicted and the risk of new-onset headache (P > 0.05). Interaction analysis revealed no significant interaction between PEF% predicted and gender with respect to new-onset headache (P > 0.05), as shown in Fig. 5.
Fig. 5.
Subgroup analysis of the association between PEF% predicted and the risk of new-onset headache
Discussion
This large-scale cross-sectional and longitudinal study involving middle-aged and elderly Chinese individuals found that higher PEF% predicted was associated with a lower prevalence and risk of headaches. RCS analysis further confirmed the negative correlation between PEF% predicted and headache risk. Subgroup analyses revealed that this correlation showed a consistent trend across the majority of subgroups. Notably, gender interacted with PEF% predicted in relation to headache risk: a higher PEF% predicted was significantly associated with a reduced headache risk only in women, but not in men.
Previous studies have indicated that low atmospheric pressure, work pressure, and mental stress are common triggers for chronic migraines [15]. In a controlled trial involving 77 volunteers, Broessner et al. [16] observed that even healthy volunteers exposed to normobaric hypoxic environments experienced migraine-like headaches. A domestic study involving 1,143 community residents revealed an association between migraine and chronic lung disease [17]. These studies suggest that hypoxia-related factors may be linked to migraine onset. However, research on the association between lung function and headache based on PEF remains scarce. Using data from the CHARLS, a large national cohort, this study conducted cross-sectional and longitudinal analyses to explore the association between lung function and headache among middle-aged and elderly Chinese individuals, with the results further validating the aforementioned conclusions.
The pathogenesis of headache is complex and not yet fully elucidated. The association between PEF% predicted and headache can be explained through the following potential mechanisms: (1) activation of the thalamus, hypothalamus, and brainstem; (2) cortical spreading depression; (3) activation of the trigeminal vascular system. First, the mechanism underlying hypothalamic activation remains unclear. It is hypothesized that the hypothalamus detects a reduction in brain-tissue oxygen partial pressure via central chemoreceptors or intrinsic oxygen-sensitive neurons, thereby initiating its own activation. The ensuing increase in hypothalamic blood flow may serve as a key signal that adjusts the balance between cerebral oxygen supply and consumption, subsequently precipitating migraine-related pathophysiological responses [18, 19]. Second, hypoxia can induce the occurrence of cortical spreading depression (CSD) in the body, which is closely related to the function of astrocytes in brain tissue that can rapidly detect the oxygenation status of brain tissue. At the same time, hypoxia-induced CSD can inhibit the respiratory function of astrocyte mitochondria, leading to mitochondrial depolarization, free radical production, and other effects, thereby inhibiting energy production and causing the onset of headaches [20]. Furthermore, hypoxia not only induces CSD in the body [21] but also reduces ATP levels [22]. Since the restoration of ion homeostasis after CSD requires a high energy input, hypoxia can sustain CSD, leading to the prolonged activation of the trigeminal vascular system and triggering migraine symptoms [23].
Subgroup analysis and interaction analysis showed that gender and PEF% predicted had an interaction effect on headache risk. An increase in PEF% predicted was significantly associated with a reduced risk of headache in women, but no significant association was found in men. This phenomenon may be attributed to the long-term exposure of women in rural China to indoor air pollution. As the primary caregivers responsible for cooking and heating in rural Chinese households, women are likely exposed to high concentrations of indoor pollutants at a rate far exceeding that of men [24]. Indoor air pollution mainly consists of carbon-based particulate matter and irritating gases produced by inefficient combustion. This complex mixture can cause airway damage, which can affect lung function over time [25]. Secondly, middle-aged women often bear dual responsibilities in the family and society, such as career development, raising children, and supporting elderly parents. These factors bring about high levels of social and psychological stress [26].
The strengths of this study are as follows: First, based on data from the CHARLS, a large-scale national cohort study, this research explored the association between PEF% predicted and headache risk among middle-aged and elderly Chinese individuals. The cohort employed multistage stratified probability sampling, ensuring the representativeness and reliability of the data. Second, this study focused on the relationship between PEF% predicted (calculated from PEF) and headache risk. Other lung function indicators, such as FVC and FEV1, involve complex measurement procedures requiring specialized equipment and technical personnel. These devices are primarily restricted to medical institutions, necessitating multiple hospital visits for testing and making real-time monitoring challenging [12, 27]. In contrast, PEF measured via a peak flow meter offers advantages including ease of acquisition, simplicity of use, patient-friendly operation, suitability for home use, and adaptability to long-term monitoring [5, 13]. Additionally, PEF% predicted is calculated based on PEF, gender, and height, and has the characteristics of simple operation, convenient calculation, and data standardization [6]. To enhance the generalizability of the results, this study performed stratified analyses across multiple subgroups and adjusted for variables related to exposure and outcomes, thereby obtaining a more precise correlation between PEF% predicted and headache risk.
This study has several limitations. First, headache diagnosis relied solely on self-reported history, which may introduce recall bias and lacks details on specific subtypes. Second, the observational design precludes causal inferences about the relationship between PEF and headache. Third, although we adjusted for many potential confounders, unmeasured factors may still influence the results. Finally, lung function was assessed only with PEF; while easy to obtain, this measure may not fully capture the diversity of pulmonary abnormalities.
Future studies should include multicenter clinical cohort studies, expand the age and racial range, refine headache classification and lung function indicators, to explore the mechanism between the two.
Conclusion
In summary, among middle-aged and elderly populations in China, particularly among middle-aged and elderly women, increasing PEF levels may hold significant implications for the primary prevention of headache. This finding provides a theoretical basis for stratifying the risk of headache onset and formulating intervention strategies in middle-aged and elderly populations in China.
Supplementary Information
Acknowledgements
We extend our heartfelt gratitude to the staff involved in establishing the CHARLS database, as well as to the respondents. Additionally, we thank Guangxi Critical Care Medicine Clinical Research Center for their support. We also thank all the authors who contributed to this article.
Abbreviations
- CHARLS
China Health and Retirement Longitudinal Study
- PEF% predicted
Peak expiratory flow predicted
- PEF
Peak expiratory flow
- FEV1
Forced expiratory volume in the first second
- FVC
Forced vital capacity
- BMI
Body mass index
- CES-D10 Scale
10-Item Center for Epidemiologic Studies Depression Scale
- DLCO
Diffusion Capacity of Carbon Monoxide in the Lung
- CSD
Cortical spreading depression
- OR
Odds ratio
- CI
Confidence interval
- RCS
Restricted cubic spline
- IQR
Interquartile range
Author contributions
Conceptualization, D.-B.H, L.-Y.H.; data curation and analysis, L.-Y.H., M.-J.X. and L.L.; validation, L.-Y.H. and Y.-Y.Q.; writing—original draft preparation, B.-H.F., Q.-J.C.; writing—review and editing, L.-Y.H., M.-J.X., and L.L.; supervision, D.-B.H. All authors have read and agreed to the published version of the manuscript.
Funding
This study was supported by the First Affiliated Hospital of Guangxi Medical University’s 2024 Hospital Self-Set Scientific Research Cultivation Project—Clinical Nursing Research Climbing Plan (Project Name: Clinical Study on Early Resistance Training to Prevent Venous Thromboembolism in ICU Patients on Mechanical Ventilation; Project Number: YYZS2023018).
Data availability
The CHARLS datasets analyzed in this study are publicly accessible via the National School of Development, Peking University (http://charls.pku.edu.cn/). Researchers can obtain these datasets by submitting a data use agreement to the CHARLS team.
Declarations
Ethics approval and consent to participate
Ethics approval and consent to participate: CHARLS study is an open dataset. The CHARLS was approved by the Ethics Review Committee of Peking University (IRB 00001052-11015) and all participants signed an informed consent. The research protocol was designed under the Declaration of Helsinki.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Liuyun Huang and Mingjie Xie are co-first authors.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Hu Z, Tian Y, Song X, Zeng F, Yang A. Associations between sarcopenia with asthmatic prevalence, lung function and comorbidity. BMC Geriatr. 2022;22:703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wang L, Deng Z, Zu M, Zhang J, Wang Y. The comorbid relationship between migraine and asthma: a systematic review and meta-analysis of population-based studies. Front Med (Lausanne). 2021;7: 609528 (Přejít k původnímu zdroji. Přejít na PubMed). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bouwens JD, Bischoff EW, in’t Veen JC, Schermer TR. Diagnostic differentiation between asthma and COPD in primary care using lung function testing. NPJ Prim Care Respir Med. 2022;32: 32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ikeda H. Effectiveness of adding a peak flow meter for the identification of patients with chronic obstructive pulmonary disease in real-world clinical practice. BMC Pulm Med. 2025;25:1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhang Q-J, Huang J-C, Chen J, Hu W, Xu L-Q, Guo Q-F. Peak expiratory flow is a reliably household pulmonary function parameter correlates with disease severity and survival of patients with amyotrophic lateral sclerosis. BMC Neurol. 2022;22:105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hansen MRH, Schmid JM. Screening for impaired pulmonary function using peak expiratory flow: Performance of different interpretation strategies. Respir Med Res. 2023;83: 101015. [DOI] [PubMed] [Google Scholar]
- 7.Alrasheedi S, Alhumaidi KA, Alharbi AM, Aldhowyan NA, Alobaid NM, Alturaif NA, et al. Relationship between asthma and headache attacks in the Qassim region, Saudi Arabia. Cureus. 2023;15: e39784. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chen X, Smith J, Strauss J, Wang Y, Zhao Y. China health and retirement longitudinal study (CHARLS). In: Encyclopedia of geropsychology. Springer; 2015. p. 1–8. [Google Scholar]
- 9.Dong B, Yue Y, Wang Z, Sun M, Wang Y. Association between physical activity, peak expiratory flow, and cognitive function in aging: a cross-sectional analysis. BMC Geriatr. 2024;24:460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jian W, Gao Y, Hao C, Wang N, Ai T, Liu C, et al. Reference values for spirometry in Chinese aged 4–80 years. J Thorac Dis. 2017;9:4538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Wang M-H, Pan L-J, Zhang Y-H, Zhu H-Q, Zhu X-B, Wang X-Q. Prevalence and risk factors of headache in Chinese with stroke: a cross-sectional study based on CHARLS. J Headache Pain. 2024;25: 217. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Xu H, Gong X, Cui K, Li X, Chen L, Lu Y, et al. Association of peak expiratory flow with motoric cognitive risk syndrome among older adults. Front Aging Neurosci. 2024;16:1412542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Trevisan C, Rizzuto D, Maggi S, Sergi G, Welmer A-K, Vetrano DL. Cross-sectional and longitudinal associations between peak expiratory flow and frailty in older adults. JCM. 2019;8:1901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li M, Yang Y, Pang L, Wu M, Wang Z, Fu Y, et al. Gender-specific associations between activities of daily living disability and depressive symptoms among older adults in China: evidence from the China health and retirement longitudinal study. Arch Psychiatr Nurs. 2019;33:160–6. [DOI] [PubMed] [Google Scholar]
- 15.Mungoven TJ, Henderson LA, Meylakh N. Chronic migraine pathophysiology and treatment: a review of current perspectives. Front Pain Res. 2021;2: 705276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Broessner G, Rohregger J, Wille M, Lackner P, Ndayisaba J-P, Burtscher M. Hypoxia triggers high-altitude headache with migraine features: a prospective trial. Cephalalgia. 2016;36:765–71. [DOI] [PubMed] [Google Scholar]
- 17.Wang X, Xing Y, Sun J, Zhou H, Yu H, Zhao Y, et al. Prevalence, associated factors, and impact on quality of life of migraine in a community in northeast China. J Oral Facial Pain Headache. 2016;30:139–49. [DOI] [PubMed] [Google Scholar]
- 18.Roy TK, Secomb TW. Functional implications of microvascular heterogeneity for oxygen uptake and utilization. Physiol Rep. 2022;10: e15303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Angelova PR, Kasymov V, Christie I, Sheikhbahaei S, Turovsky E, Marina N, et al. Functional oxygen sensitivity of astrocytes. J Neurosci. 2015;35:10460–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Schain AJ, Melo-Carrillo A, Stratton J, Strassman AM, Burstein R. CSD-induced arterial dilatation and plasma protein extravasation are unaffected by fremanezumab: implications for CGRP’s role in migraine with aura. J Neurosci. 2019;39:6001–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Chen S-P, Qin T, Seidel JL, Zheng Y, Eikermann M, Ferrari MD, et al. Inhibition of the P2X7–PANX1 complex suppresses spreading depolarization and neuroinflammation. Brain. 2017;140:1643–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Andersen JV, Jakobsen E, Waagepetersen HS, Aldana BI. Distinct differences in rates of oxygen consumption and ATP synthesis of regionally isolated non-synaptic mouse brain mitochondria. J Neurosci Res. 2019;97:961–74. [DOI] [PubMed] [Google Scholar]
- 23.Yao G, Man Y-H, Li A-R, Guo Y, Dai Y, Wang P, et al. NO up-regulates migraine-related CGRP via activation of an akt/GSK-3β/NF-κB signaling cascade in trigeminal ganglion neurons. Aging (Albany NY). 2020;12: 6370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Chen N, Ma L-L, Zhang Y, Yan Y-X. Association of household solid fuel use and long-term exposure to ambient air pollution with estimated 10-year high cardiovascular disease risk among postmenopausal women. Environ Pollut. 2024;342: 123091. [DOI] [PubMed] [Google Scholar]
- 25.Xu T, Ye X, Lu X, Lan G, Xie M, Huang Z, et al. Association between solid cooking fuel and cognitive decline: three nationwide cohort studies in middle-aged and older population. Environ Int. 2023;173: 107803. [DOI] [PubMed] [Google Scholar]
- 26.Santosa A, Rosengren A, Ramasundarahettige C, Rangarajan S, Gulec S, Chifamba J, et al. Psychosocial risk factors and cardiovascular disease and death in a population-based cohort from 21 low-, middle-, and high-income countries. JAMA Netw Open. 2021;4:e2138920–e2138920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Nathan SD, Wanger J, Zibrak JD, Wencel ML, Burg C, Stauffer JL. Using forced vital capacity (FVC) in the clinic to monitor patients with idiopathic pulmonary fibrosis (IPF): pros and cons. Expert Rev Respir Med. 2021;15:175–81. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The CHARLS datasets analyzed in this study are publicly accessible via the National School of Development, Peking University (http://charls.pku.edu.cn/). Researchers can obtain these datasets by submitting a data use agreement to the CHARLS team.