Authorship and contributorship
Spirometric reference values for healthy nonsmoking Saudi adults
Article first published online: 31 JUL 2013
© 2013 John Wiley & Sons Ltd
The Clinical Respiratory Journal
Volume 8, Issue 1, pages 72–78, January 2014
How to Cite
Al Ghobain, M. O., Alhamad, E. H., Alorainy, H. S., Al Hazmi, M., Al Moamary, M. S., Al-Hajjaj, M. S., Idress, M., Al-Jahdali, H. and Zeitouni, M. (2014), Spirometric reference values for healthy nonsmoking Saudi adults. The Clinical Respiratory Journal, 8: 72–78. doi: 10.1111/crj.12038
Mohammed O. Al Ghobain: principal investigator, study design, supervising the study, and writing the manuscript and journal submission; Esam Alhamad: study design, data collection and supervising the study at her center; Hassan S. Alorainy: training the technicians, study design, quality assurance, data analysis; Manal Al Hazmi: study design, data collection and supervising the study at his center; Mohamed S. Al Moamary: study design, data collection and writing the manuscript; Mohamed S. AL-Hajjaj: study design, data collection and writing the manuscript; Majdy Idress: study design, data collection and supervising the study at his center; Hamdan Al Jahdali: study design, data collection and supervising the study at his center; Mohammed Zaitoni: study design, data collection and supervising the study at his center.
The study was approved by the ethics committee of involved hospitals.
Conflict of interest
The authors have stated explicitly that there are no conflicts of interest in connection with this article.
Funding from King Abdulaziz City for Science and Technology.
- Issue published online: 6 JAN 2014
- Article first published online: 31 JUL 2013
- Accepted manuscript online: 25 JUN 2013 07:58AM EST
- Manuscript Accepted: 18 JUN 2013
- Manuscript Revised: 10 JUN 2013
- Manuscript Received: 14 JAN 2013
- King Abdulaziz City for Science and Technology
- adults ;
- references ;
- Saudi Arabia ;
To derive prediction equations of spirometric values of healthy Saudi adults and to compare the derived equations with equations reported in selected population.
Cross-sectional study of healthy nonsmoking men and women Saudi adults. The measured spirometric values were the forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), peak expiratory flow (PEF) and forced mid-expiratory flow (FEF 25%–75%).
A total of 621 spirometric tests were done. The prediction equations were derived using the following formula: Predicted spirometric value = constant + (b1 × age) + (b2 × height (cm)), where b1 and b2 represent the regression coefficients for age and height, respectively.
|Variable||Constant||Age (years)||Height (cm)||Variable||Constant||Age (years)||Height (cm)|
|Males (n = 292)||Females (n = 175)|
|FEV1/FVC (%)||98.41||−0.095||−0.068||FEV1/FVC (%)||100.67||−0.142||−0.072|
The means of the measured FVC and FEV1 were significantly lower than the predicted values derived by the American equations of −7.2% and −4.6% among males, respectively (P value < 0.00001), and −4.7%, and −5.26% among females, respectively (P value < 0.00001).
The reference spirometric values derived in our study were significantly lower than the predicted values derived by the American equations.
The spirometric prediction equations used in pulmonary function laboratories are primarily derived from western populations. These equations are affected by many variables, including race and ethnic background. Therefore, a reliable and accurate interpretation of a pulmonary function test (PFT) would require the utilization of prediction equations derived from a population with the same ethnic background. This will lead to achieving a reliable assessment of pulmonary impairment, disease progression and response to treatment upon follow-up. Although there are many studies reporting the appropriate prediction equations among adult populations in many regions of the world, literature from Saudi Arabia and other Arab countries is scant, and has many limitations [1-6] To our knowledge, the only recent study conducted in our region to derive spirometric prediction equations for an adult population was performed in Oman. Al-Rawas et al. derived prediction equations for spirometry values in 419 healthy, nonsmoking, Omani adults aged 18–65 years, including 256 men and 163 women . Among Omani subjects, the authors concluded that the predicted normal values of forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) were lower than those for Caucasians.
In Saudi Arabia, pulmonary function laboratories are currently using spirometric prediction equations derived from a western population. Such western-derived equations may not be accurate for the Saudi population, and they may affect the accuracy and reliability of pulmonary function results because of the differences in ethnic background. This would indicate an urgent need to establish reference values applicable to the local population. Therefore, the aim of this study was to derive the prediction equations for spirometric values using data from a probability sample of asymptomatic, healthy, nonsmoking, adult Saudi men and women, and to compare our derived prediction equations with equations reported from western population.
Materials and methods
A multicenter, cross-sectional study was conducted at the pulmonary function laboratories of five hospitals in Riyadh, Saudi Arabia: King Abdulaziz Medical City, King Fahad Medical City, King Khalid University Hospital, King Faisal Specialist Hospital and Riyadh Military Hospital.
The study included healthy, adult, Saudi volunteers of both sexes between 18 and 65 years of age. The subjects were lifetime nonsmokers for either cigarettes or water pipes (shisha), with a weight range from 40 kg to 120 kg, and a height range from 140 cm to 190 cm. The exclusion criteria included the following: a history of any respiratory complaints, such as cough, shortness of breath or wheezing; a history of upper respiratory tract infection in the previous 4 weeks; a history of respiratory diseases, such as asthma or tuberculosis; a history of cardiac or thoracic surgery; features suggestive of cardiac or lung disease, or evidence of chest deformities or a serious medical condition; and a history of working in an environment with a high concentration of dust or pollution.
The subjects who accepted the invitation signed an informed consent and underwent a medical evaluation, including a meticulous and thorough medical history, and a full physical examination that evaluated their respiratory and cardiac systems. The sex, age, standing height and body weight of all subjects who met the inclusion criteria were recorded. Weight was measured with the subjects wearing light clothing and barefoot on a SECA weighing scale (Hamburg, Germany). Standing height was measured without shoes with the subject's back to a vertical backboard. Both heels were placed together, touching the base of the vertical board. The spirometry tests were conducted using a Micro-Loop (Viasys Healthcare, Ireland). The spirometry device was calibrated with a JAEGER calibration pump (Hoechberg, Germany) using a 3.0-L syringe at three flow rates, in accordance with the manufacturer's recommendations, before each day's testing and after every few hours of testing. All subjects underwent spirometry tests using the techniques recommended by the American Thoracic Society (ATS) . The spirometry test was performed in the morning by two trained qualified technicians. The subjects underwent the spirometry test in a sitting position, wearing a nose clip. Uniformity of the spirometry test was assured by using the same device brand for all subjects. The validity of the test was verified according to the ATS recommendations . The measured spirometric values were the forced vital capacity (FVC), forced expiratory volume in 1 s (FEV1), peak expiratory flow (PEF) and forced mid-expiratory flow (FEF25%–75%). In addition to these measured parameters, the ratio of FEV1 to FVC was calculated and expressed as a percentage [FEV1/FVC (%)]. The study was approved by the institutional review board (IRB) of King Abdullah International Medical Research Center.
Statistical analysis was performed using SPSS 18 software (SPSS, Chicago, IL, USA). Because lung function data from males and females significantly differ, regression analyses were applied to each sex separately. A multivariate regression analysis was performed to calculate linear regression equations for each of the spirometric variables (FVC, FEV1, FEV1/FVC, PEF and FEF25%–75) by age and height. The coefficient of determination (R2) and the regression coefficient values for each lung function variable were calculated. For all indices, linear models on the original scale provided an acceptable fit to the data, and therefore linear models were chosen as the basic format for evaluating the relationships between the lung function variables as the dependent variables, and age and height as the independent variables. Regression coefficients among the spirometric variables for age and height for each sex were also derived separately. The validation was performed by calculating the predicted, percent predicted and lower limit of normal for our subjects for each of the lung function variables derived using the prediction equations of the present study. The predicted lower limit of normal (LLN) for each lung function variable was calculated by subtracting 1.64 multiplied by the residual standard deviation (SD) from the predicted mean reference value. Comparisons between the measured values of lung function variables and their respective predicted values by different equations were performed using the paired-sample Student's t-test. A P value less than 0.05 was considered statistically significant. The performance of the equations used in the current study was compared with the equations based on a random sample of the general population of the United States, published by Hankinson et al. . The means and SDs of both sets of predicted values were calculated for each spirometric variable by age and sex, and the results were compared using the same tests. Approximately three fourths of the subjects (467 subjects) were randomly selected for the development of prediction equations. The remaining 154 subjects were used as a control group to validate the derived equations.
A total of 800 spirometric tests were performed among eligible subjects; 179 (22.4%) were excluded because of poor performance on spirometry or incomplete data. The remaining 621 (77.6%) spirometric tests met the quality criteria proposed by the ATS, and were considered for the data analysis and constituted the study population. Table 1 shows the baseline characteristics and age distribution of the study subjects. The regression model was applied using age, height and weight as independent variables. Both gender and age but not the weight identified as important independent variables for all spirometric values. The addition of weight to the regression models did not demonstrate any significant improvement in the models, as indicated by the R2 values. Therefore, only age and height were included in the reference equations for all spirometric values. Table 2 shows the predicted spirometric reference equations (FEV1, FVC, FEV1/FVC %, PEF and FEF25%–75%) for male and female healthy adults in Saudi Arabia. Each predicted spirometric value can be derived using the following equation:
|Gender, N (%)||395 (63.6%)||226 (36.4%)||621 (100%)|
|Age (year), M (± SD) age distribution (year) N (%)||33 (10.4)||31 (8.9)||33 (9.9)|
|18–25||109 (27.5)||(34.9)||188 (30.3)|
|26–35||145 (36.7)||84 (37.1)||229 (36.9)|
|36–45||85 (21.5)||47 (20.7)||132 (21.4)|
|46–55||44 (11.1)||13 (5.7)||57 (9.3)|
|>55||12 (3.0)||3 (1.3)||15 (2.7)|
|Height (cm), M (± SD)||171 (6.7)||159 (5.6)||166 (8.6)|
|Weight (kg), M (± SD)||82 (14.7)||71 (15.6)||78 (15.9)|
|Body mass index, M (± SD)||28 (4.9)||28 (6.1)||28 (5.3)|
|Variable||Constant||Age (years)||Height (cm)||Residual||1.645 × residual||R2|
|Males (n = 292)|
|Females (n = 175)|
where b1 and b2 represent the regression coefficients for age in years and height in centimeters, respectively. The standard deviation of the residuals (residuals = observed – predicted) and the value of 1.64 multiplied by the residual corresponds to the prediction value at the lower fifth percentile. The lower limit of normal for each variable can be calculated by subtracting 1.64 multiplied by the residual from the mean predicted value. For example, a 40-year-old man measuring 171 cm has, according to the equation, a predicted FEV1 of (−1.886) + (−0.019) × 40 + 0.036 × 171 = 3.53 L, and his LLN would be 3.53 – 0.60302 = 2.62 L.
There were no significant differences between the mean measured and mean predicted spirometric values (FEV1, FVC, PEF and FEF25%–75%) in either males or females (Table 3).
|Variable||Measured Mean (SD)||Predicted Mean (SD)||Percent of predicted||P value|
|Male (n = 103)|
|FVC||4.5 (0.62)||4.4 (0.37)||101.8||0.108|
|FEV1||3.7 (0.51)||3.7 (0.32)||100.7||0.487|
|FEV1/FVC (%)||82.7 (3.60)||83.6 (0.96)||98.9||0.008|
|PEF||570.0 (84.77)||570.4 (27.44)||99.9||0.961|
|FEF25%–75%||3.8 (0.82)||3.9 (0.31)||98.26||0.365|
|Female (n = 51)|
|FVC||3.3 (0.49)||3.2 (0.30)||100.3||0.829|
|FEV1||2.8 (0.35)||2.70 (0.27)||100.6||0.636|
|FEV1/FVC (%)||84.1 (2.75)||84.8 (1.45)||99.1||0.059|
|PEF||399.4 (43.12)||405.2 (22.83)||98.5||0.377|
|FEF25%–75%||2.9 (0.43)||2.9 (0.28)||98.3||0.418|
We further compared the mean predicted and percent predicted values of the spirometric parameters of our subjects (derived using our equations) with the values for the same parameters derived using other selected prediction equations. The means of the measured FVC and FEV1 were significantly lower than the predicted values derived by the American equations (P value < 0.00001) (Table 4). Among males, there were significant differences in FVC and FEV1 of −7.2% and −4.6%, respectively. However, the differences in PEF and FEF25%–75% were not statistically significant at −0.52% and 0.77%, respectively. In female subjects, FVC, FEV1, PEF and FEF25%–75% were significantly different between our equations and the American equations at −4.7%, −5.26%, 4.97% and −8.51%, respectively (P value < 0.00001).
|Variable||Measured||Saudi population||NHANES III (American population)||Difference (%)||t-value||P value|
|Mean (SD)||Mean (SD)||(%)||Mean (SD)||(%)|
|FVC||4.5 (0.618)||4.4 (0.370)||101.8||4.74 (0.36)||83.5||−7.2%||8.6||0.00001|
|FEV1||3.7 (0.511)||3.7 (0.317)||100.7||3.88 (0.04)||85.8||−4.6%||11.9||0.00001|
|FEV1/FVC (%)||82.7 (3.600)||83.6 (0.960)||98.92||81.3 (2.2)||103||2.8%||10.4||0.00001|
|PEF||9.50 (1.41)||9.50 (0.45)||99.93||9.55 (0.55)||90.8||−0.52%||0.9||0.3889|
|FEF25%–75%||3.8 (0.819)||3.9 (0.305)||98.25||3.87 (0.58)||103.1||0.77%||0.5||0.6104|
|FVC||3.3 (0.487)||3.2 (0.300)||100.3||3.36 (0.33)||86.1||−4.7%||4.5||0.00001|
|FEV1||2.8 (0.350)||2.70 (0.273)||100.6||2.85 (0.33)||87.4||−5.26%||4.3||0.00001|
|FEV1/FVC (%)||84.1 (2.754)||84.8 (1.449)||99.1||83.8 (2.6)||102||1.80%||3.8||0.00001|
|PEF||6.65 (0.71)||6.75 (0.38)||98.5||6.43 (0.43)||98.0||4.97%||6.9||0.00001|
|FEF25%–75%||2.9 (0.433)||2.9 (0.282)||98.3||3.17 (0.45)||106.9||−8.51%||5.8||0.00001|
When compared with the equations derived from the Omani population, the means of the measured FVC, FEV1 and PEF values were significantly higher, although FEF25%–75% was not (Table 5). Among males, the differences in the FVC, FEV1 and PEF values were 11.39%, 11.11% and 9.57%, respectively, while the differences among females were 9.96%, 8.43% and 5.79%, respectively. As shown in Tables 4 and 5, the spirometric values derived by our equation were between the values predicted by the American and Omani equations.
|Variable||Measured||Saudi population||Omani population||Difference (%)||t-value||P value|
|Mean (SD)||Mean (SD)||(%)||Mean (SD)||(%)|
|FVC||4.5 (0.618)||4.4 (0.370)||101.8||3.95 (0.4)||100.4||11.39%||9.8||0.00001|
|FEV1||3.7 (0.511)||3.7 (0.317)||100.7||3.33 (0.33)||99.9||11.11%||9.7||0.00001|
|FEV1/FVC (%)||82.7 (3.600)||83.6 (0.960)||98.92||84.6 (0.8)||99.5||−1.18%||10.1||0.00001|
|PEF||9.50 (1.41)||9.50 (0.45)||99.93||8.67 (0.56)||100.0||9.57%||13.4||0.00001|
|FEF25%–75%||3.8 (0.819)||3.9 (0.305)||98.25||4.09 (0.37)||97.6||−4.64%||4.6||0.00001|
|FVC||3.3 (0.487)||3.2 (0.300)||100.3||2.91 (0.34)||99.8||9.96%||7.6||0.00001|
|FEV1||2.8 (0.350)||2.70 (0.273)||100.6||2.49 (0.31)||100.1||8.43%||6.0||0.00001|
|FEV1/FVC (%)||84.1 (2.754)||84.8 (1.449)||99.1||85.9 (0.9)||99.8||−1.28%||8.7||0.00001|
|PEF||6.65 (0.71)||6.75 (0.38)||98.5||6.38 (0.49)||98.9||5.79%||6.9||0.00001|
|FEF25%–75%||2.9 (0.433)||2.9 (0.282)||98.3||3.43 (0.33)||99.0||−15.45%||14.3||0.00001|
Our study identified the reference equations for the predicted spirometric values of asymptomatic, healthy, nonsmoking, adult Saudi men and women. To the best of our knowledge, this is the first recent study to report the predicted spirometric values from Saudi Arabia. Because of the lack of local reference values, practitioners in Saudi Arabia have utilized reference equations derived from western countries and their normal limits. These differences resulted in classifying a large proportion of our normal subjects as below the predicted lower limit of normal. Therefore, the predicted normal values for western populations may not be suitable for use in Saudi populations and may lead to incorrect clinical diagnoses. Such ethnic variability is likely related to genetic variations between different ethnic groups affecting body segments, the chest wall anatomy, lung mechanics and the speed of lung development [9, 10]. Although such ethnic differences have been widely suggested, there is no firm evidence to provide accurate estimates of its impact on reference values . To overcome this overestimation in local laboratories, pulmonary function machines were adjusted to the reference values for Americans because they have lower predicted values than white Caucasians [12-14] Our study had shown that this was an acceptable approach, and revealed that the FVC and FEV1 were less than the Caucasian values by −7.2% and −4.6% among male subjects, and −4.7% and −5.26% among female subjects, respectively. This is further supported by a study from the United Kingdom that revealed that the ventilatory function among western Pakistanis living in the United Kingdom was similar to that of African Americans . Until reference values for Arab patients are widely available, the utilization of the reference values for African Americans appears to be a reasonable alternative.
The only available study that supports our finding was published recently in Oman by Al-Rawas and colleagues . By comparing lung function values derived from Omani prediction equations with the values derived from Caucasians, the predicted Omani values for FVC were lower by 11%–17% in men and 7%–14% in women. The FVC and FEV1 values obtained from our Saudi population were significantly higher than the Omani population by 11.39 % and 11.11% among male individuals, and 9.96% and 8.43% among female individuals, respectively. There are similarities between the two countries, including cultural habits and factors related to exercise, nutrition, overall health status and regional pollution. Therefore, the spirometric differences may be attributed to different genetic and ethnic backgrounds between the two populations . Similar differences have been observed within the populations of high-population countries [17, 18]. Compared with white Caucasians, lower spirometric values have been observed in other ethnic groups in Asia. Another study among Asian Indians in the United States showed that their FVC was lower than white subjects by 20%–24% among males and 20%–26% among females .
The lack of reference values in Arab countries demonstrates the need for researchers to conduct large-scale studies in this region. In addition to having reference spirometric values, they may contribute to the understanding of the relative roles of genetic constitution and exogenous influence on lung function development. In the global village, every physician in any part of the world may be faced with subjects of different ethnic groups and needs to have information about possible physiological differences, including pulmonary function.
Although this study is the first recent report from Saudi Arabia, there are a few limitations that are worth mentioning. The study subjects were volunteers not randomly selected. However, the method of selection has been reported to not influence either the mean values or their ranges for lung function measurements . Further, all centers involved in this study were located in one city. However, Riyadh is the capital of Saudi Arabia and has the highest Saudi population, which makes it a reasonable representation of the whole country. Nevertheless, repeating this study in other regions in Saudi Arabia is an area for further research. Finally, the sample sizes for certain age groups were relatively small and may affect the accuracy of the predicted equations.
In conclusion, despite the aforementioned limitations of our study, it established the need to launch a national project to develop reference spirometric values. Currently, spirometric values are misinterpreted by utilizing reference values derived from other ethnic groups. These studies should not be limited to spirometry alone and should be extended to other pulmonary function indices.
The authors would like to acknowledge the respiratory care technicians who work in the pulmonary function laboratories at the included hospitals. Additionally, the authors would like to acknowledge the editing services at King Abdullah International Medical Research Centre for editing the manuscript.
- 2Normal spirometric values investigated in King Abul Aziz Teaching Hospital, Jeddah. Saudi Med J. 1982;3: 159–170., , .
- 6Normal values of ventilatory lung function tests among male Egyptians industrial workers. Respir News Bull. 1984;26: 8–10., , .
- 8Standardization of spirometry, 1994 update. American Thoracic Society. Ame J Respir Crit Care Med. 1995;152(3): 1107–1136.
- 18Pulmonary function laboratory in the tropics: needs, problems and solutions. In: Sharma O , editor. Lung Disease in the Tropics. New York, Marcel Dekker, 1991: 523–420., .