Evaluation of the Global Lung Function Initiative 2012 reference values for spirometry in an Iranian population

Between 16 July and 27 August 2019, 900 participants completed spirometry measurements. After exclusions, 622 participants (204 males and 418 females) aged 4–82 years met the selection criteria for the reference sample.

The mean (range) age of men and women was 38.34 (4–82) and 44.55 (4–80) years, respectively. In this sample, the mean height was 1.72 (0.08) m in males and 1.58 (0.08) m in females aged over 21 years. Thirty-nine (19.2%) men and 131 (31.4%) women had a BMI ≥ 30 kg/m². The demographic and clinical characteristics of the participants are described in Table 1.

The Caucasian GLI-2012 was applied to our population. Overall, the mean Z-scores of FEV₁, FVC, and the FEV₁/FVC ratio for males and females in the various age groups were higher than the Caucasian predicted values (range: 0.01 to 1.05) except for the FEV₁/FVC in the age group under 21 years (range: − 1.11 to − 0.09).

The normal distribution curves of FEV₁, FVC, and FEV₁/FVC based on observed GLI Z-score mean and standard deviation values in men and women are presented in Fig. 1a–f.

(a–f) Normal distribution curves of FEV1, FVC and FEV1/FVC based on observed mean GLI Z-score values in men and women.

The distribution of the Z-scores of FEV₁, FVC, FEV₁/FVC, and FEF_25–75% stratified by sex and age in the reference sample of healthy individuals is shown in Table 2.

The FEV₁ Z-score was smaller than 0.5 in men and women aged 10–21, 22–29, and 30–39 years; however, its standard deviation was often above 1. Moreover, this value was below 0.5 in girls under 10 years old and men 40–49 years old and over 70. The FEV₁ Z-score was not different from zero (by one-sample t-test analysis) in the age groups 10–29 and over 70 years in both genders (P > 0.05) (Table 2).

The FVC Z-score exceeded the predicted values (0.5) across age groups < 10, 30–39, 40–49, and 50–59 years in both genders, but it was below 0.5 in individual aged 10–21 (males), 22–29 (both gender), 60–69 (males), and > 70 (both gender) years old (Table 2).

The Z-score of FEV₁/FVC was below 0.5 in all age groups except for the age group under 10 (both genders) and 60–69 (males) years (Table 2).

The mean Z-score of FEF_25–75% was between 0 and − 0.5 in males in all age groups; the same finding was found in women aged 10–21, 40–49, 50–59, and 70–84 years (Table 2).

In the age group over 21 years, the Z-score of FEV₁/FVC was below the LLN in seven men (3.43%) and eight women (2.01%). However, these values were significantly higher in six boys (46.2%) and eight girls (40.0%) under 21 years old.

Z-score of FEV₁/FVC less than LLN was zero in women over the age of 60 and in men aged 22–39 and 60–69 years. It was also less than 5% in women aged 50–59 and men aged 30–59 years (1–3.4%).

Totally, the FEV₁/FVC Z-score was above the upper limit of normal in 20 (9.8%) men and 69 (16.5%) women (ULN > 1.64).

The mean Z-score of FEV₁/FVC above the upper limit of normal (ULN, > 1.64) ranged between 0% (age group 22–29 years) and 25% (over 70 years) in men and between 0% (age group 10–21 and over 70 years) and 28.3% (age group 50–59 years) in women.

The FEV₁ Z-score was within the 90% limits of normality (− 1.64 to + 1.64) in 81.3% of the observations (83.7% in males and 80.4% in females). The corresponding figure was 78.9% for FVC (84.5% in males and 76.1% in females) and 93.3% (90.2% in males and 94.8% in females) for the FEV₁/FVC ratio in the age group over 21 years.

The Z-scores of FEV₁, FVC, FEV₁/FVC, and FEF_25–75% were analyzed according to age, height, weight, and gender using a linear regression model.

Age, weight, and height (but not gender) had an impact on the FEV₁/FVC Z-score in univariate regression (P < 0.05). In multiple linear regression (in the presence of height, weight and age as variables with P < 0.2 in univariate linear regression) height and age remained associated (B-coefficient = 0.012 and 0.007; P = 0.008, P = 0.001 respectively) with the FEV₁/FVC Z-score (Fig. 2a). There was a significant association between age and FEV₁ Z-score (B-coefficient = 0.007; P = 0.011). In the multiple linear regression (in the presence of age, gender and height), age remained statistically significant (B-coefficient = 0.008; P = 0.006) (Fig. 2b).

(a–d) Association of GLI2012 mean Z-score values of FEV1, FVC, FEV1/FVC, and FEF_25–75% with gender, age, and anthropometric variables according to multiple linear regression.

There was a significant association between gender and FVC Z-score (B-coefficient = 0.389; P < 0.001) in univariate linear regression; this association was maintained in the multiple regression model too (B-coefficients = 0.346; P = 0.004) (Fig. 2c).

Height and gender had an effect on the FEF_25–75% Z-score (B-coefficients =  − 0.371 and 0.006; P < 0.001 and P = 0.021, respectively) in univariate linear regression. In the multiple linear regression model, gender had a significant relationship with the FEF_25–75% Z-score (B-coefficient =  − 0.358; P = 0.001) (Fig. 2d).

The prevalence of COPD defined by spirometry based on the fixed ratio (FR) criterion increased with age from 0.8% in the age group 30–39 years to 16.7% in the age group > 70 years (P_{for linear by linear association} < 0.001). The prevalence of COPD according to the LLN criterion did not follow a specific trend (P_{for linear by linear association} = 0.749). There was a 98.13% agreement between FR and LLN method (Fleiss’ kappa coefficient = 0.58, P < 0.001). The prevalence of COPD based on FR and LLN according to age and sex is presented in Table 3.