Longitudinal study of the influence of obesity, C-reactive protein, and smoking on FEV1 decline in young adulthood

Abstract

Background

Chronic obstructive pulmonary disease (COPD) is a disease with a high socioeconomic burden for the global population. Identifying those individuals with a higher potential to develop the disease is essential for reducing its incidence.

Methods

This is an observational, longitudinal study that uses data from the 1978/1979 Ribeirão Preto City birth cohort (São Paulo State, Brazil). The study included 895 individuals who participated at the age of 23–25 and 37–38 years. Asthmatics were diagnosed by methacholine bronchial challenge test and were excluded from the analysis. A multiple linear regression was performed to evaluate the association of active smoking, passive smoking, body mass index (BMI), C-reactive protein (CRP) levels, and respiratory symptoms with FEV1 variation between ages.

Results

The analysis showed an association between BMI, CRP levels, and active smoking with FEV1 fall. Active smoking increased FEV1 decline by 1.95%. For each 1 kg/m² increase in BMI, there was a 0.28% loss in FEV1, while an increase in CRP level of 1 mg/dL was associated to a 0.76% additional FEV1 decline.

Conclusion

In addition to the well-known relationship between smoking and pulmonary function decline, there was also an association with BMI and CRP levels, suggesting the hypothesis that a metabolic process may contribute to the development of COPD.

Background

Chronic obstructive pulmonary disease (COPD) is a condition that involves chronic respiratory symptoms and structural lung changes, leading to airflow limitation. Population aging partly explains the increase in both its incidence and the costs associated with its management and treatment [1–3]. There is a need for methods aiming at early diagnosis or detection of susceptible individuals.

Several studies have aimed to better understand the pathophysiological process of COPD [4] or develop early detection methods for susceptible patients [5] or design appropriate prevention and treatment interventions [6].

The main risk factor for loss of respiratory capacity - and the leading cause of COPD - continues to be smoking [7]. However, not all individuals exposed to smoke will develop the disease, which highlights the role of candidates mechanisms involved in the process, such as genetic characteristics¹ (genes regulating alpha-1 antitrypsin production, and those related to lung development) [5, 8] morphological traits, and metabolic factors [1].

The causal relationship between obesity and certain respiratory diseases, such as asthma, is already well established, as it is a condition that leads to systemic inflammation [9]. For other pulmonary diseases, this relationship has not yet been widely explored [10]. We hypothesized that metabolic syndrome may also be correlated with other respiratory changes such as COPD [11] considering the association between inflammation and its genetic regulation with failure in the development and formation of the respiratory tract [12].

An individual’s lung function trajectory can be influenced by numerous potential risk factors, and identifying those susceptible to developing respiratory diseases becomes a complex but necessary task. Thus, the aim of this article is to amplify our knowledge about the risk factors for early pulmonary function decline.

Methods

Study sample

The data used in this article were obtained during the 4th and 5th assessments of the 1978/1979 Ribeirão Preto City birth cohort, São Paulo State, Brazil [13, 14]. The 4th (T1) and 5th (T2) cohort assessments took place when the participants were 23–25 years old and 37–38 years old, respectively; 895 individuals were assessed in both these time-points and were included in this study (Fig. 1).

Fig. 1.

Composition of the Birth Cohort 1978/1979

Ribeirão Preto is located in the northeastern region of the state of São Paulo, 320 km from the state capital. It is considered a wealthy and industrialized city, with a population of 698,942 and a Human Development Index of 0.835.

Protocol

At the ages of 23–25 (T1) and 37–38 years (T2), individuals from the birth cohort were recruited for re-assessment. In the first assessment, questionnaires were administered, and spirometry, methacholine bronchial challenge tests, and blood sampling were performed. In the second assessment, questionnaires, blood sampling, and spirometry were performed. All procedures were carried out in a single day, in the morning. In the first evaluation (T1), spirometry preceded bronchial challenge test.

The bronchoprovocation test is capable of identifying bronchial hyperresponsiveness even in patients with mild or absent symptoms, and it can also detect individuals with asthma outside of bronchospasm episodes. Following international guidelines [15] individuals were classified as asthmatic if they met one of the following criteria: (1) a concentration causing a 20% FEV1 fall (PC20) below 1 mg/mL, with or without reported respiratory symptoms; or (2) a PC20 between 1 and 8 mg/mL combined with respiratory symptoms.

Individuals with a confirmed asthma diagnosis based on the methacholine challenge test showed wide variability, with both large decreases and large increases in FEV1 between ages. These intense fluctuations significantly interfered with the overall behavior of lung function between the two time points; therefore, asthmatic individuals were excluded from the final analyses.

Procedures

Data collection was conducted by a trained team responsible for: (1) administering structured questionnaires; (2) collecting blood for laboratory tests (C-reactive protein measurement); (3) performing spirometry and methacholine bronchial challenge test.

C-reactive protein measurement was carried out in association with other laboratory measurements that were not described in this study. Serum concentrations of C-reactive protein were quantified in an automated biochemistry analyzer (Weiner, Rosario, Argentina).

Questionnaires

Data regarding the presence of respiratory symptoms - such as wheezing in the last 12 months, chest tightness, or shortness of breath - were obtained with the use of a structured questionnaire with questions on respiratory health derived from the European Community Respiratory Health Survey.

Participants were considered active smokers if they reported smoking throughout the entire study period, meaning they answered positively to smoking at least one cigarette per day at both T1 and T2.

Methacholine bronchoprovocation tests and spirometry procedures followed international guidelines [15–17]. A Koko^® pneumotachograph (PDS Instrumentation Incorporation, Louisville, Colorado, USA) was used for these assessments.

Studied variables

The variable of interest was FEV1 variation, calculated using data from the spirometries performed. The FEV1 value (in liters) obtained at T1 was subtracted from the value obtained at T2. This result was then transformed to demonstrate its corresponding percentage relative to the initial FEV1 value. Subsequently, this percentage value was multiplied by 10 solely for the convenience of the calculations. FEV1 variation was considered decline when the result was negative and gain when the result was positive.

The independent or explanatory variables were: (1) body mass index (BMI) variation: calculated by subtracting the BMI value at the second time-point (T2) of the cohort from the BMI value at the first time-point (T1); (2) wheezing in the last 12 months; (3) allergic rhinitis; (4) active smoking; (5) passive smoking; (6) abdominal circumference; (7) C-reactive protein level.

Statistical analysis

An exploratory data analysis was performed considering measures of central tendency and dispersion. Histograms and boxplot graphs were created to evaluate the dispersion of each quantitative variable. Means and medians were calculated for central tendency. Standard deviations, minimum and maximum values were determined for dispersion. Qualitative variables were summarized in frequency tables. Comparisons were performed using Student’s t-test and Chi-square test.

A multiple linear regression model was constructed to examine which independent variables are associated with the outcome of FEV1 variation. The following assumptions were tested: (1) collinearity among independent variables using the variance inflation factor; (2) homoscedasticity of the residuals through graphical analysis.

A final model was constructed using only variables that showed a significant association with FEV1 variation.

A p-value of less than 5% (p < 0.05) was considered statistically significant. The statistical software used was R version 4.3.3 and RStudio.

Results

The sample comprised 895 individuals, 444 women (45.7%) and 451 men (54.3%). The demographic, clinical, and pulmonary function data, considering gender, are presented in Table 1 (categorical variables) and Table 2 (continuous variables).

Table 1.

Characteristics of the population studied (n:895)

	Male (n:451)		Female (n:444)		p-value
	n	%	n	%
Abdominal circumference					0.000
High	140	31	191	43
Normal	311	69	253	57
Rhinitis					0.000
Yes	106	24	191	43
No	345	76	253	57
Wheezing (last 12 months)					0.539
Yes	21	5	12	03
No	430	95	427	97
Other symptoms						0.476
Yes	17	4	21	5
No	434	96	423	95
Smoking						0.134
Yes	44	10	31	7
No	406	90	412	93
Secondhand smoking						0.594
Yes	126	28	117	26
No	325	72	327	74
FEV1 < 80% of predicted value at age 23–25						0.063
Yes	31	7	46	10
No	420	93	398	90
FEV1 < 80% of predicted value at age 37–38						0.102
Yes	38	8	25	6
No	413	92	419	94

FEV1 Forced expiratory volume in one second

Table 2.

FEV1 variation, CRP levels and BMI in each time and variation by sex and smoking status

	Male	Female	ρ	Current smoker	Never smoker	ρ
FEV1 variation in liter (SD)	−0.21 (± 0.222)	−0.15 (± 0.169)	0.000	−0.238 (± 0.20)	−0.183 (± 0.19)	0.020
FEV1% variation (SD)	−5.13 (± 5.14)	−4.88 (± 5.71)	0.499	−6.53 (± 5.6)	−4.88 (± 5.4)	0.012
CRP [ml/dL; median (IQR)]	0.14 (0.07–0.27)	0.3 (0.11–0.61)	0.000	0.2 (0.01–0.49)	0.18 (0.08–0.43)	0.528
BMI at 23–25 ys [mean (SD)]	24.80 (± 4.11)	23.59 (± 5.09)	0.000	25.08 (± 5.11)	24.12 (± 4.61)	0.120
BMI at 37–38 ys [mean (SD)]	28.95 (± 4.63)	28.45 (± 6.17)	0.167	28.01 (± 5.10)	28.77 (± 5.49)	0.220
BMI variation [mean (SD)]	4.38 (± 3.72)	4.42 (± 4.22)	0.871	4.48 (± 3.63)	4.39 (± 4.01)	0.855

Variation was 37–38 years value minus 23–25 years value

FEV1 Forced expiratory volume in one second, CRP C-reactive protein, BMI Body mass index, IQR Interquartile range, SD Standard deviation

Women showed a higher prevalence of elevated abdominal circumference (43% versus 31% in men) and rhinitis (43.0% versus 23.5% in men). These two variables showed a significant difference between genders (p < 0.001).

The frequencies of wheezing in the last 12 months, active smoking (at T1 and T2), and passive smoking were 4.6%, 9.7%, and 27.9%, respectively, among men, and no significant differences were observed when compared to women (p > 0.05).

C-reactive protein (CRP) levels were higher in women (p < 0.001). When the sample was divided into smokers and non-smokers, regardless of gender, no significant difference was found (Table 2). Conversely, the FEV1 variation relative to baseline value (percentage variation) did not have difference between genders (p > 0.05), but it was significantly higher among current smokers (p < 0.05). The mean variation in BMI did not differ significantly either when comparing genders or between smokers and non-smokers (p > 0.05).

Multiple linear regression (Table 3) showed a significant negative association between FEV1 variation and BMI variation. Each increase of 1 kg/m² in BMI was associated with a loss of 0.28% of FEV1 (β = 0.285; t = −5.923; p < 0.001).

Table 3.

Multiple linear regression analysis for association between variation in FEV1 and variables of interest

	Coefficient	Standard error	p-value	95% CI	Beta
Intercept	−3.856	2.127	0.070	−8.033 - 0.319
Sex	−0.611	0.369	0.098	−1.336 - 0.113	−0.056
BMI variation	−0.285	0.048	< 0.001	−0.379 - −0.190	0.210
Abdominal circumference	−0.266	0.405	0.511	−1.062 - 0.530	−0.023
Wheezing (last 12 months)	1.049	1.275	0.410	−1.454 - 3.552	0.039
Other symptoms	0.996	1.267	0.431	−1.491 - 3.484	0.037
Rhinitis	0.244	0.387	0.527	−0.515 - 1.005	0.021
CRP	−0.765	0.238	0.001	−1.232 - −0.298	−0.109
Smoking	−1.957	0.664	0.003	−3.260 - −0.653	−0.099
Secondhand smoking	−0.119	0.403	0.766	−0.911 - 0.671	−0.009

BMI Body mass index, CRP C-reactive protein, CI Confidence interval. The variables of interest were sex, BMI variation, abdominal circumference, wheezing in the last 12 months, other symptoms, rhinitis, CRP, smoking, and passive smoking

Similarly, a significant negative association was found between FEV1 variation and CRP levels as well as active smoking. For every 1 mg/dL increase in CRP levels, there was a loss of 0.76% in FEV1 (β = −0.765; t = −3.214; p = 0.001), while smoking resulted in a 1.95% loss in FEV1 (β = −1.957; t = −2.946; p = 0.003). When the impact of each of these variables on FEV1 variation was assessed, we observed that smoking had the strongest association, followed by CRP levels and BMI variation, in this order.

The other variables studied - gender, abdominal circumference, presence of wheezing in the last 12 months, presence of other symptoms or allergic rhinitis, and passive smoking - did not show a significant association with FEV1 fall.

The model created by using only the variables that were found to be associated (Table 4) was more appropriate for predicting FEV1 variation [F (3.886) = 22.46; p < 0.001; R² = 0.07].

Table 4.

Correlation of variables of interest with FEV1 variation

	Rho*	p-value
Gender	−0.013	0.6779
BMI variation	−0.263	< 0.0001
Abdominal Circumference	−0.144	< 0.0001
Wheezing (last 12 mo)	0.084	0.0112
Other symptoms	−0.079	0.0178
Rhinitis	0.021	0.5242
CRP	−0.142	< 0.0001
Smoking	0.084	0.0113
Secondhand smoke	0.044	0.184
PC20 (methacholine)	0.017	0.621

*Pearson correlation. BMI Body mass index, CRP C react protein, PC20 The provocative concentration causing a 20% FEV1 fall

There was no high correlation between the explanatory variables that could influence the multivariable linear analysis.

The same evaluation was performed by separating the smoking and non-smoking groups, and the associations between BMI variation and CRP levels, as well as FEV1 variation, remained consistent.

Discussion

The study was conducted with young adults between their third and fourth decades of life. During this period, a large part of the patients started the process of pulmonary function decline. Based on the FEV1 assessments, 75% of the sample showed fall. The whole sample had mean FEV1 decrease of 5%. This pattern is consistent with those described by previous authors, who highlighted the onset of a decline in pulmonary function starting in the third decade of life [18–22].

The proportion of active smokers in this study was 7.6%, while for the same age group, the average is 12% in Brazil and 21.7% worldwide [23, 24]. Our results show that active smoking is the factor most strongly associated with FEV1 decline, even in these young people without established respiratory diseases.

Exposure to passive smoking was similar between genders and did not have association with FEV1 decline, which is consistent with previous studies [25–27]. Similarly, the presence of respiratory symptoms - defined here as wheezing, shortness of breath, chest tightness, and symptoms of rhinitis - did not have association.

Regarding the potential influence of a systemic inflammatory process, we analyzed the measurement of CRP and higher CRP levels were associated with greater FEV1 decline. This is a widely available test used in population studies investigating its relationship with cardiovascular diseases [28] as well as with evolution of pulmonary function [28, 29]. We hypothesized that there may be a systemic metabolic issue that could make or identify individuals with greater susceptibility to pulmonary function loss and the development of future lung disease.

Some studies evaluated other factors that could cause systemic inflammation, as well as their relationship with pulmonary function. One of the major research topics has been obesity [7, 9, 30]. Zhu et al. (2021) demonstrated the existence of shared genes between the presence of obesity and the decline in pulmonary function, suggesting that the influence of obesity on lung function goes beyond the mechanical issue and may also involve issues related to immune response, cellular proliferation, and embryological, skeletal, and tissue development [10].

The presence of abdominal fat is strongly associated with an increased mortality from cardiovascular diseases [9]. However, it does not have a noticeable relationship with respiratory function. The data available on the subject is still unclear. In cross-sectional studies like that of Garcia-Larsen et al. (2014), no significant association was found between the two parameters [19] whereas, in a longitudinal study, in which the relationship between variation in abdominal circumference and lung function over a period was assessed, a significant and negative association was found [29]. Other extrapulmonary factors such as body composition have been studied in association with FEV1 decline. Doi et al. showed that left handgrip strength in males and bone mineral content in females exhibited a significantly positive association with FEV1. Moreover, the decline in these gender-specific extrapulmonary factors may serve as a potential screening tool for the early detection of low FEV1 in early adulthood [30].

In our data, abdominal circumference did not have a significant association with FEV1 variation. This measure was assessed only at the 23–25 years evaluation, which may have led to an incomplete analysis. On the other hand, BMI was calculated in both assessments, and the variation between the two measures was calculated. We showed that an increase in BMI was correlated with greater FEV1 loss. This finding supports the hypothesis that there is indeed a systemic inflammatory process involved in FEV1 decline (Fig. 2).

Fig. 2.

Proposed relationship between variables and FEV1 decline

This study has the following strengths: (1) longitudinal design, which is beneficial in cases of insidiously developing diseases; (2) the precise exclusion of individuals with asthma, avoiding potential bias from high pulmonary function variability, which is characteristic of asthma; and, (3) standardization of spirometry performed at the same conditions by the same staff 12 years apart.

Study limitations

Some study limitations are (1) the use of spirometric values without bronchodilator administration; (2) the impossibility of confirming the onset of COPD in those individuals with large FEV1 decline; and, (3) other variables could be tested such as the sagittal abdominal diameter - most accurate indicator of central obesity and cardiometabolic risk. The original cohort was initially designed for other primary objectives; therefore, some potentially useful and interesting variables could not be analyzed.

Conclusion

As with active smoking, increased BMI and higher CRP levels are associated with greater loss of pulmonary function in young adults. It can be inferred that, in the evolutionary process of lung function toward COPD, there may also be interference from metabolic phenomena.

Abbreviations

COPD

Chronic obstructive pulmonary disease

FEV1

Forced expiratory volume in one second

PC20

The methacholine concentration causing a 20% FEV1 fall

T1

First time-point at the age of 23-25 years

T2

Second time-point at the age of 37-38 years

BMI

Body mass index

CRP

C-Reactive protein

IQR

Interquartile range

SD

Standard deviation

CI

Confidence interval

FVC

Forced vital capacity

Authors’ contributions

A.C.C. and E.O.V. contributed substantially to the study design, data analysis, interpretation and writing of the manuscript. C.C.R. and M.F.Jr. contributed substantially to the study design, data analysis and interpretation. M.A.B. and H.B. contributed substantially to the cohort creation and description, study design, data collection, database preparation and analysis. They had full access to all the data in the study and takes responsibility for the integrityof the data and the accuracy of the data analysis. These authors are guarantors of the paper. All authors have read and approved the manuscript.

Funding

This study was supported by FAPESP (Grant 2017/21035-8), and CNPq, Brazil.

This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brazil (CAPES) - Finance Code 001.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study adhered to the principles outlined in the Declaration of Helsinki and was carried out with the approval of the Research Ethics Committee (number 2.947.100). The requirement for informed consent was waived by the Institutional Review Board due to the use of a database belonging to the 1978 birth cohort previously approved by the Ethics Committee, under approval number 1.282.710. The researchers ensured no physical and/or biological risks were present and ensured the personal confidentiality of the participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.