Abstract
BACKGROUND: Measured maximum voluntary ventilation (MVV) correlates with maximum ventilatory capacity during exercise. As a shortcut, MVV is often estimated by multiplying measured FEV1 times 35 or 40, but this index varies with altitude due to reduced air density. The objective was to describe MVV in healthy individuals residing at 2,240 m above sea level and compare it with the reference values customarily employed.
METHODS: We recruited a convenience sample of respiratory-healthy, non-obese volunteers >10 y of age who had resided for >2 y in Mexico City. All participants performed forced spirometry and MVV according to current standards. Multiple regression models were fitted, including age, height, and measured FEV1, separately for males and females to obtain reference values. The impact of lower air density on MVV at this elevation was estimated from the reported increase in peak flow in relation to altitude.
RESULTS: We studied 381 individuals (210 females [55.1%]) age 10–80 y with a mean MVV of 145.6 ± 48 L/min. Both FEV1 × 35 and FEV1 × 40 underestimated the MVV observed: in males by approximately 26% and in females by approximately 10%. MVV for our population approached FEV1 × 45 (98 ± 15.6% of real MVV). Multiple regression models including height, weight, and measured FEV1 explained 70% of residual variability once sex was taken into account.
CONCLUSIONS: At an altitude of 2,240 m, MVV is about 45 times the measured FEV1, and it can be estimated for other altitudes. The best predicting equations for MVV were calculated separately for females and males and included the following predictors: age, age2, and measured FEV1. The study found that reference values for MVV from studies conducted at sea level are inaccurate at this altitude.
Introduction
The study of lung function is key for the diagnostic evaluation and monitoring of patients with respiratory diseases and in assessing surgical risk, disability, and prognoses. Measuring maximum voluntary ventilation (MVV) is one of the tests used to evaluate respiratory mechanics. It is relatively simple and reproducible and can be performed with most spirometers. MVV correlates with maximum ventilation achievable during exercise, and residual respiratory reserve is routinely employed to interpret exercise tests. According to current diagnostic algorithms, a decrease in the value of the residual respiratory reserve indicates lung disease if FEV1 is low and indicates neuromuscular disease or poor effort if it is normal.1–4
Expected MVV values, usually based on sea-level reference values,2,3 may fail at higher altitudes because the reduced air density increases forced respiratory flows. In addition, spirometric values from residents of Mexico City differ from those reported in developed countries, including Mexican-American populations.5–7 Because inadequate reference values lead to errors in interpretation, our aim was to establish proper reference values for MVV in a population residing at 2,240 m above sea level and then compare them with published reference values to determine whether adjusting sea-level values for air density explains the differences in MVV measured in healthy people.
QUICK LOOK
Current knowledge
Instead of measuring maximum ventilatory capacity (MVV) before performing a cardiopulmonary exercise test, it is often recommended to estimate MVV by multiplying FEV1 times 35-40; however, this recommendation does not take into account the changes in air density that occur at altitudes and may increase MVV.
What this paper contributes to our knowledge
Our study reveals that the recommended prediction strategies underestimated the MVV of a population residing at moderate altitude. At the elevation of Mexico City (2,240 m above sea level), the MVV was very close to FEV1 × 45 and thus could be estimated for other altitudes.
Methods
The study protocol was approved by the institutional ethics committee, and the adults or parents of the children who participated signed a letter of informed consent. Testing was conducted from May 2013 to August 2014 at the Pulmonary Function Testing Laboratory of the National Institute of Respiratory Diseases (INER) in Mexico City, a reference center for respiratory diseases that attends primarily patients who have no social security or other health insurance coverage. We included Mexican male and female individuals residing in the Metropolitan Area of Mexico City (2,240 m above sea level), >10 y of age, never-smokers (<400 cigarettes smoked in lifetime), without previous respiratory diseases (COPD, chronic bronchitis, asthma, chest surgery, regular use of respiratory medications) or respiratory symptoms (wheezing, dyspnea, coughing, phlegm), and non-obese (body mass index < 30 kg/m2 in individuals > 18 y old and in the < 95th percentile of Centers for Disease Control and Prevention values in younger individuals).6
Participants were selected by invitation or through posters, mostly from among INER employees, their relatives, and students (of medicine, respiratory care, and nursing) enrolled in courses at our institution. No participants were active in competitive sports. Ethnic origin is usually not included in surveys applied in Mexico City, and all participants were considered to be Mexican mestizos.
Spirometry tests and MVV were performed with pneumotachograph-based equipment (Ergospirometry 5.22.1.149, CareFusion, San Diego, California) according to American Thoracic Society/European Respiratory Society standards.1 All tests were performed by experienced trained personnel certified by the National Institute of Occupational Safety and Health after completing spirometry courses at the INER training site. We utilized reference values for spirometry obtained in Mexican-American individuals.5 The volume accuracy of the instrument was verified daily using a 3.00-L calibration syringe.
Briefly, the technician explained and demonstrated the maneuver to all participants. Subsequently, the maneuvers were performed with subjects seated and wearing a nose clip. Each maneuver consisted of at least 3 tidal breaths followed by breathing as fast and as deeply as possible for 12 s. The 12-s ventilation observed was multiplied by 5 to obtain MVV at 1 min. Additional maneuvers were carried out until the highest 2 MVV values matched each other within 10 L/min. The highest of the 3 MVV results was reported.1
Statistical Analyses
Descriptive statistics (means ± SD) were utilized to characterize the study population. Linear regression models were employed to predict MVV utilizing the following independent variables: age (crude, but also age2 and age3 because we included individuals during growth with increasing MVV and adults with decreasing MVV due to aging), height, weight, sex, and measured FEV1. Variables significantly associated with MVV in the univariate models were then incorporated into multivariate models to test fit with and without measured FEV1, to generate predicting equations for the case in which FEV1 was not measured. For each model, we observed the determination coefficient, r2, and the SD of the residuals (root mean square error) as general indicators of the model's fit. We compared the MVV measured in of the participants with common reference values employing a graphic smoothing technique, locally weighted scatterplot smoothing, to observe relationships among variables and their tendencies. Also evaluated was the appropriateness of estimating MVV as FEV1 times 35 or 40, as is commonly recommended,7 before comparing our MVV measurements with several reference values.8–13 Statistical analyses were performed with Stata 13 statistical software (StataCorp, College Station, Texas).
Results
We studied a total of 381 healthy residents of Mexico City age 9–81 y, including 210 females (55.1%). Mean age was 36.3 y, and body mass index was 24 ± 4.2 kg/m2 (Fig. 1 and Table 1). Potential participants who responded to an invitation to participate in the study were predominantly young adults with only a small proportion of elderly men and women. Subjects had FEV1, FVC, and FEV1/FVC expressed as a percentage of prediction within normal range and similar in females and males. Table 1 depicts MVV in L/min and as a percentage of predicted from several reference equations, with percentages considerably above the expected 100% (for properly adjusted reference equations), but also significant between-sex differences (all P < .001). Although FEV1 × 45 or FEV1 × 40 × 1.11 (altitude adjusting factor; see the supplementary materials at http://www.rcjournal.com) in participants is near the average MVV, significant between-sex differences continue to exist, although they disappear upon applying the reference multiple regression equation (bottom line of Table 1).
Subject distribution by age and sex.
Baseline Characteristics of the Population
Figure 2A depicts MVV as a function of age in males and females. It demonstrates increasing MVV (ascending curve on the left) and then aging (descending curve on the right), with higher values for males than females. When MVV was adjusted by measured FEV1 (MVV/FEV1; Fig. 2B), age variations diminished significantly, as MVV and FEV1 increased and decreased more-or-less proportionally until showing an early horizontal distribution of points. As Table 1 shows, measured MMV was approximately 45 times the measured FEV1 (the exact value obtained by linear regression for both sexes together was 46.6 × FEV1; bottom line of Table 2).
Maximum voluntary ventilation (MVV) (A) and MVV/FEV1 (B) as a function of age. Measured FEV1 considerably reduces age variation, which approaches 45 for nearly all ages. Results from males (circles) and females (filled circles) justify the practice of estimating MVV by multiplying measured FEV1 by a constant.
Regression Coefficients From Significant Univariate Predictors of Maximum Voluntary Ventilation
Table 2 presents the regression coefficients of variables significantly associated with MVV, including the correlation (r) and determination coefficients (r2). These variables plus age2 and age3 were tested by multiple linear regression due to the curvilinearity of the relationship between age and MVV (Fig. 2A) and then fitted separately for males and females (Table 3), including equations with and without measured FEV1. The proportion of MVV variability explained by the models was ∼70% for those that included FEV1 and slightly below 60% for those that did not consider FEV1 (Table 3). Figures 3 and 4 summarize the study data (continuous line) and the MVV expected by different predicting equations9–15 (several also in Table 1) for males (Fig. 3) and females (Fig. 4). Lines represent smoothing (by locally weighted scatterplot smoothing) of the measured MVV and common reference equations.
Best Multivariate Predicting Equations for Maximum Voluntary Ventilation With and Without Measured FEV1
Comparison of the observed measurements of maximum voluntary ventilation (MVV) in men from the current study with those expected from previously published studies for other populations. Equations basically involve the age-related decline in maximal voluntary ventilation but not the growth during childhood. In addition, predictions from the majority of equations are considerably below the measured values in Mexico City. Lines were obtained by smoothing of the data points using locally weighted scatterplot smoothing.
Comparison of the observed measurements of maximal voluntary ventilation (MVV) in women from the current study with those expected from previously published studies for other populations. Equations basically involve the age-related decline in maximal voluntary ventilation but not the growth during childhood. In addition, predictions from the majority of equations are considerably below measured values in Mexico City. Lines were obtained by smoothing of the data points using locally weighted scatterplot smoothing.
Discussion
On average, MVV in Mexico City at 2,240 m above sea level was 126.3 ± 20% of FEV1 × 35 and 110 ± 17.5% of FEV1 × 40, as is commonly recommended before an exercise test to estimate MVV without performing the test,2,3which is considered potentially fatiguing. In fact, MVV approached 45 times the measured FEV1 (yielding 98 ± 15.6% of predicted), probably due to the lower air density at 2,240 m above sea level and the resulting higher flows during maneuvers that involved turbulence. The expected increase in peak expiratory flow in Mexico City compared with sea level and, probably, MVV due to lower air density was 11%,7 which fits the measured data if we estimate predicted MVV at Mexico City as 40 × FEV1 × 1.11 (measured values were 99.2 ± 15.7% of this estimate; see Table 1 of the supplementary material). A similar procedure could be used to estimate expected values for other altitudes. Because the predicting equation for changes in peak expiratory flow with altitude utilized a power function of barometric pressure (see Methods), an estimate based on altitude would be simpler, since from sea level to 3,000 m, the calculated value is very close to a straight line: MVV correction factor = 1.0 + (0.051 × altitude [km]) (see the supplementary material).7 Because adding measured FEV1 to the reference equations substantially improves their accuracy for predicting MVV, we recommend including the former if MVV is not measured directly.
Even in healthy individuals, estimates of predicted MVV may differ considerably from measured MVV (−50 to +50 L/min). The coefficient of variation for MVV is 3.8% and 5.1 L/min, slightly higher than for FVC (3.3% and 0.24 L) and FEV1 (2.6% and 0.18 L). Therefore, before cardiopulmonary exercise testing, we recommend measuring MVV instead of only estimating it from FEV1. In our experience, healthy individuals and patients can perform this maneuver properly without experiencing fatigue, dizziness, or coughing.
Figures 2 and 3 synthesize the comparison of MVV measured in our study and common reference values. Average predicted values were below measured MVV in both males and females. In addition, the age-slope of several reference values was different from that of the individuals studied, especially the females. Finally, in Table 1, the MVV expressed as a percentage of predicted produced, in general, significantly higher values for males than females.
Several limitations of our study should be mentioned. Although the distribution of participants by age adheres to that observed in the Mexican population, the number of elderly individuals included was low, which could reduce the accuracy of our equations for older age groups. Also, participants formed a convenience sample, although they were healthy residents of Mexico City with a range of socioeconomic levels characteristic of that population. Gathering a population-based sample representative of Mexico was not feasible, but also probably unnecessary to demonstrate the change in MVV with altitude compared with previously reported sea-level values. Finally, once adjustment for FEV1was performed, variations of MVV with age were small.
Conclusions
Published equations and prediction strategies for MVV do not apply to individuals residing at moderate altitudes. Results in Mexico City were approximated by FEV1 × 45, but this is expected to change at other altitudes. We therefore recommend measuring MVV before cardiopulmonary exercise testing.
Footnotes
- Correspondence: Silvia Cid-Juarez MD MSc, National Institute of Respiratory Diseases, Calzada de Tlalpan 4502, Col. Sección XVI, Deleg. Tlalpan, Mexico City 14080, Mexico. E-mail: scidj{at}hotmail.com.
Dr Cid-Juárez presented a version of this paper at the European Respiratory Society Conference, held September 26–30, 2015, in Amsterdam, Netherlands.
The authors have disclosed no conflicts of interest.
Supplementary material related to this paper is available at http://www.rcjournal.com.
- Copyright © 2017 by Daedalus Enterprises
