Racial or Ethnic Variation in Spirometric Lung Function Norms

Racial or Ethnic Variation in Spirometric Lung Function Norms· Recommendations Based on Study of Ethiopian Jews

Haim Bibi, M.D.;]ohn R. Goldsmith, M.D.; and Hillel Vardi, M.S.

It is widely known that different ethnic (or racial) groups do not appear to have lung function test results which 6t the same prediction formulae. Israel, with ethnic minorities from many countries, is £aced with a dilemma as to what to use as a basis for normative estimates. We studied 146 Ethiopian Jewish immigrants (68 males and 78 females), all DODSmoken with no evidence of any chronic disease. The FEV1 and FVC were analyzed separately by sex and age (those leu than 25 and those 25 years or more of age). Standard prediction formulae based on height overpre-

dieted the values found by from 15 to 29 percent. Logarithmic formulae based on sitting height provided a good 8t for values for children «25 years). Compared to other populations the sitting height-standing height ratio was low for this population, (0.48 vs 0.52) reftecting their relatively long limbs and shorter thoracic height. Our 6ndings suggest that as a guideline, ethnic groups with deviant standing height-sitting height ratios either use ethnic-group speci8c prediction formulae or derive prediction formulae using sitting height.

lung function tests have been studied Spirometric among many different population groups.':" Nu-

Israels population, like that of the United States in previous periods, is made up largely of immigrants, and a high proportion are dark-skinned. Initial studies of immigrants from Kurdistan and Yemen indicated that otherwise healthy immigrants had lung function that differed from that of United States population norms, based as they are on predominantly white individuals. 11 Goren and Bruderman'" have shown that immigrants from Iran or Iraq or North African countries other than Morocco have lower lung function than Israeli-born or other immigrants. There is, therefore, a substantial question as to which prediction formulae should be applied to dark-skinned ethnic groups in Israel. The same questions are relevant in Europe and the United States. Tohelp answer this question, we report the observed lung function test results for a group of Ethiopian Jews who by history x-ray and physical examination were found to have no pulmonary disease. Using both sitting and standing height, sex, and age, we analyzed the appropriateness of various prediction formulae.

merous prediction formulae and nomograms have been published for predicting what values should be for persons with no evidence of respiratory or other disease, based usually on standing height (less frequentlyon sitting height), age and sex of the subjects. 7.8 It is widely agreed that blacks in the United States have about 15 percent lower lung function than whites of similar ages and heights, and this is said to apply to children and adults of both sexes.9.10 It is not biologically plausible that skin color can affect ventilatory function, but the basis for these observations is not otherwise analyzed. In this report we intend to identify as "ethnic" differences those associated with skin color or country of origin. By "racial" differences are understood differences including not only skin color but inheritable biologic variation of many sorts, including facial characteristics and stature. It follows that ethnic group differences may include those due to race as well as to other variables. Hsi et aI, 7 in studies on children in Texas, showed that using sitting height instead of standing height in logarithmic prediction formulae overcame the apparent differences in the predicted/observed ratio for white, black and Mexican children. Otherwise available prediction formulae for "normal" blacks, including Ethiopian and East African subjects, are based on standing height, age, and sex.4-6 *From the Pediatrics Department, Barzilai Medical Center, Ashkelon, Israel, and the Epidemiology and Health Services Evaluation Unit, Ben-Gurion University of the Negev Beer Sheva, Israel. Manuscript received May J8; revision accepted October 2. Reprint requeat,: Dr. Goldsmith, PO Box 473, Orner 85-965 Israel

1021

MATERIAL AND METHODS

A total of 146 Ethiopian Jews living in the same neighborhood in Ashkelon were studied. They ranged in age from 11 to 70 years; there were 68 male and 78 female subjects. All were nonsmokers, and each had been examined by a pulmonologist, using chest x-ray film and PPD skin test to rule out tuberculosis or other chronic lung disease. Physical examination was done as well to exclude persons with chronic or intercurrent illness. They had been in Israel for about two years, and therefore, would no longer have manifested the acute impairments associated with their arduous travel. Spirometric lung function tests (SLm weight, standing height, and sitting height were measured. The SLFT included the FVC and FEVit measured with a portable spirometer, and calibrated Racial or Ethnic VariatIon in Spirometric Lung FunctionNorms (Bibl. Goldsmith. Vardl)

Table I-Mean Sitting Height, Standing Height, Age, FVe and FEV1!or Ethiopian]ew. in Age and Sa Grou".

ratio by age is shown in Figure 1, along with comparable data for a population ofJapanese-Americans from Hawaii studied by Massey and Fournier-Massey" When similar equipment and procedures were used to study Israeli school children, about nine years of age, in the Ashkelon region according to country of birth of the father, the ratios of sitting height to standing height varied in the narrow range of 0.5214 to 0.5262 for female subjects and 0.5170 to 0.5197 for male subjects. This provides the basis, along with the results of Hsi et al7 for the line at 0.52 in Figure 1.

(±SD)

Groups

Age

Sitting Standing Height, Height, SitJ FVC, Stand M L M

0.774 <25 Males 16 (3.6) (0.062) N=21 47.0 0.820 25+ Males (13.0) (0.035) N=47 0.739 <25 Females 15.9 (4.6) (0.()58) N=33 0.773 25+ Females 43.3 (13.0) (0.051) N=45

FEV h L

3.09 (0.869) 0.482 3.19 (0.665) 0.485 2.09 (0.502) 0.490 2.25 (0.505)

1.597 (0.121) 1.700 (0.079) 1.525 (0.095) 1.578 (0.086)

0.486

2.66 (0.748) 2.75 (0.579) 1.85 (0.441) 1.97 (0.450)

Lung Function Prediction Equations Table 2 gives the lung function equations based on standing height and on sitting height for the population studied for FVC and FEV1. As can be seen, the multiple correlation coefficient, R, is slightly greater for equations based on standing height than on sitting height in seven out of the eight sets of equations. The R values are especially low for the equations for adult women, a group in which variation in height contributes very little to the prediction. As anticipated, the coefficient for age is positive in most of the equations for younger persons. In Table 3A is shown the extent to which various prediction formulae (shown in Appendices A and B) seem to describe the relationships of height and age for these populations. An estimation of the degree of

each day at an ambient temperature of 24 to 2frC. The SLFfwere measured in the sitting position after the procedures badbeen both demonstrated and explained by a trained translator in the Amharit language. We selected the best of three measurements from subjects who cooperated to give consistent results. Data are reported in terms of ambient temperature and pressure, Ashkelon being at sea level. RESULTS

Table 1 shows the results according to sex and age classes. We chose, arbitrarily to divide the subjects by age at age 25, since there is some evidence that subjects may increase their lung function up to that age. 13 The distribution ofthe sitting height-standing height

0..58 - - - - - - - - - - - - - - - - ' - - - - - - - - - - - - - - - -......

_----1-:.

------;--------

....&:.~

--a----._____-------

-------

0..54 0..53

~~--__e-H_----~~~~ ...... .....-_::11

--

'....... ..... ,. _..... ...

"".•••":1".....-.-.,.

.

---------

APPROXIMATE RATIOS FOR ISRAELI AND US (WHITE) CHILDREN

-- -- -- -

---- -- -

---

0.51

0..5

~---==-~....--. ........----~/.......--------"?'iI!r------..

/.

0.49 0...48

0.47

--._-._-._-._-.

~

//

."-

-~~..-;....-----

...-

...----

./.".,.

......'"

----

.,.-....

/

....

...

/"

~

-------1

0..46 .......- - - - - - . . . . . - - - - -.....- - - - - - . - - - - - - -....... 20-29 J. FE ....ALE

[]

.6.GE GIiDUP I)

E.1I.4ALE

70+

E. FElr.4ALE

1. Sitting-standing height ratios, Japanese and Ethiopians, by "age and sex. Source of data for Japanese-Americans: from Massey and FournieJl-Mass~ Environ Res 1986; 39:418-33, by permission. FICURE

CHEST I 83 I 5 I MAY, 1888

1027

Table I-Lung Funceion Pndkeion EqutJIionafor EdaiopitmJew Baaedon Age, Sltmding Height (H) Gnd Sitting Height (S) Males Boys (to age 25) N == 21 Mean FEV1=S.0973H-+0.0253age-S.07 2.66 =S.755 8+0.0395 age-2.421 FVC=S.I94 H-0.0232 age-S.562 3.09 =6.033 8 + 0.0866 age - 2.963 Adults (age 26 + ) N = 47 FEV1 =2.260 H -0.0221 age-O.0593 =4.722 8-0.0221 age-O.859 2.7S FVC=3.994 H-O.0234 age-2.507 =4.329 8 -0.0226 age+O. 701 3.19 Females Girls (to age 25) N =33 FEV1==2.829 H+O.OO404 age-2.S28 =2.046 8+0.0364 age-2.414 1.85 FVC = 3.642 H + 0.0059 age - 3.552 =3.124 8 +0.0444 age-O.922 2.09 Adults (age 26 + ) N = 45 FEV1t= -0.00846 age +2.336 =0.289 5-0.0088 age +2. 128 1.97 FVC = 0.505 H - 0.0144 age + 2.084 = 0.S11 8 - 0.0145 age + 2.490 2.24

H we assume that the reason for the poor fit is the relatively low sitting height-standing height ratio for the Ethiopian Jews, then it is reasonable to try to make some adjustment to account for the relatively long limb length of the populations. We can do this by using the observed sitting height and assuming a ratio of 0.52, which seems to be characteristic of other populations. We have been unable to locate data on sitting height-standing height ratios for populations of adults. (Gilson and Hugh-Jones in their report, "Lung function in Coal Workers pneumoconiosis," MRC Spec Rep Ser 290, London: HMSO, 1955, give such data [range 0.5204-05402].) No doubt they are found in the literature of physical anthropology Using this calculated standing height-sitting height/0.52 in the formulae which use that variable, we get the results shown in Table 3B, representing a much better fit.

R=0.0.745 R=0.0.616 R=O.O.799 R=0.0.714 R=O.O.566 R=O.O.553 R=O.O.636 R=0.0.481 R=O.O.640 R=0.0.561 R=0.0.729 R=0.0.662

DISCUSSION

R=O.O.245 R=0.0.247 R=O.O.369 R=O.O.362

The sitting height-standing height ratio for the Ethiopian population is lower than for the black children in the Hsi et al? study It is possible that the Ethiopian immigrants from Northeast Africa are even more slender and relatively longer limbed than the West African blacks who comprise most of the United States black population. Another possibility is that the passage of time since their immigration to the United States has moderated their stature either due to improved nutrition or to genetic drift. Massey and Foumier- Massey 8 who found that Japanese-Americans tend to have a higher ratio than other United States populations, also felt that this tendency was diminishing in groups that had relatively more

- Heights in meters; lung function in liters.

tThe contribution of standing height to the regression for FEV1 is trivial.

fit can be found in the ratio of observed to predicted, which is underlined in the tables. Except for the Hsi et al,14 formulae which is a logarithmic function of sitting height alone and used only fur the groups under 25 years of age, the usual prediction formulae overestimate the lung function by 15 to 29 percent, that is, given prediction ratios respectively of 0.85 to 0.71. Table

3-""'"from

Applying Sewral Pndkeion Forrraalae to Predicting Lung Function Among EthiopitJn Jew. 3A

Observed Mean

Ratio

Hsi et al7 Prediction (EEC Prediction Modified)t

Ratio

OIP Ratio

EEC-HSI et al1 Prediction

Ratio

Knudson et al 13 Prediction (Modified)t

3.187 3.730

0.836 0.828

2.408 2.702

1.081 1.144

2.607 3.066

1.022 1.008

2.408* 2.702-

1.081* 1.144-

3.248 4.166

0.845 0.765

3.4:56 4.208

0.794 0.757

2.438 3.122

1.126 1.020

2.924 3.497

0.939 0.911

2.610 2.945

0.709 0.711

1.9942.131-

0.928 0.983

2.259 2.512

0.819 0.834

1.9942.131-

0.9280.983-

2.601 3.085

0.757 0.731

2.551 2.661

0.772 0.847

2.316 2.692

0.850 0.837

2.187 2.253

0.900 1.000

Knudson et al13 Prediction

Males Boys (to 25 years~ N = 21 FEVI 2.664 FVC 3.090 Men (26 + years~ N = 47 2.745 FEV1 FVC 3.186 Females Girls (to 25 years~ N =33 FEV1 1.850 FVC 2.094 \\bmen (26+years~ N=4S FEV1 1.969 FVC 2.254

3B

-Since the Hsi et aJ1 fOrmula Cor children is alreadybased on the sitting height, these values are unchanged from Table 3, A. tu: ~ we use the observed sitting height and compute the standing height on the basis of an assumed ("normal") sitting height to standing height ntio of 0.52, then we get the following results.

1028

RacIal or Ethnic vartaIion InSpIrometrIc Lung Function Norma (81b1. GoIdamlth. \WdI)

years of living under conditions usual in the United States. It is likely that the better fit of regression equations based on standing height has led to ignoring of sitting height, which is the more valid indicator of racial differences in lung volumes. Mengesha and Mekkonen" also studied lung function of Ethiopian populations in Ethiopia. Their prediction equations based on age and standing height had similar coefficients of multiple correlation to ours for men and lower coefficients for women. They based their comparisons of 6ts to prediction among different ethnic groups on values which would be found for a "standard man or woman," namely: one aged 40 and 170 cm tall for men and 160 cm for women. Inserting these values into the equations, we derived, as shown in Table 2 for adults, somewhat lower predictions than they found. For example, they found for a 40-year-old, 170 cm man, FVC of 4. 35L, and our data would predict 3.35. Their volumes were corrected to body temperature and pressure saturated with water vapor (BTPS) which leads to about a 10 percent increase in the observed values recorded at 20°C. The differences, however, are greater than 10 percent. If there were no technical differences, this discrepancy may indicate that Ethiopian Jews differ physically from other Ethiopians in having lower respiratory volumes. Differences observed in these indices could be due to genetic factors, physical make-up inherent in the different ethnic groups, altitude or other environmental differences, or habitual physical activities. In neither of the studies were smoking subjects included.

Relevance to Ethnic or Racial Variation of Normal Lung Function Our findings lead us to propose a practical guideline for estimation of norms of lung function among different ethnic or racial groups. If such groups show a deviant ratio of sitting height to standing height, namely: an average ratio deviating appreciably from 0.52, then it is likely that "standard" prediction formulae are not appropriate. In such a situation, the preferable route is to derive ethnic-specific prediction equations. An alternate possibility is to compute an expected standing height from the observed sitting height divided by 0.52. Although the fit for children using a logarithmic formula based on sitting height alone is good and shows no effect of ethnicity it seems premature to recommend such formulae for adults. We cannot justify the continuation of arbitrarily reduced lung function norms for ethnic minorities, as . this is likely to lead to misclassification of a large fraction of those studied. ApPENDIX A PREDICTION FORMULAE TESTED WITH

DATA

FOR

ETHIOPIAN IMMIGRANTS· Males Age 6-<12 FEV! =0.0348 H -2.8142 FVC =0.0409 H -3.3756 Age 12-<25 FEV1 = 0.0519 H + 0.0636 age - 6.1181 FVC =0.059 H+0.0739 age-6.8865 Age 25+ FEVl =0.0665 H -0.0292 age-6.5147 FVC = 0.0844 H - 0.0298 age - 8.7818 Females Age 6-<11 FEV1 = 0.0336 H - 2.7578 FVC =0.043 H-3.7486 Age 11-<20 FEVl =0.0351 H +0.0694 age-3.7622 FVC =0.0416 H +0.0699 age-4.447 Age 20+ FEVl = 0.0309 H - 0.0201 age -1.405 FVC =0.0427 H -0.0174 age-2.9001 *From Knudson et al. l3 ApPENDIX B ALTERNATE PREDICTION FORMULAE· Males Up to age 25 Hsi FEV} =0.000818 S exp 3.42 FVC =0.000771 S exp 3.46 Age 26 and up EEC FEV. =0.04301 H-O.0209 age-2.492 FVC =0.0575 H -0.026 age-4.345 Females Up to age 25 Hsi FEV} =0.00196 S exp 3.21 FVC =0.00142 S exp 3.3 Age 26 and greater EEC FEVl = 0.03953 H - 0.025 age - 2.604 FVC = 0.04426 H - 0.024 age - 3.284 *From Hsi et al (for black children) and EEC.l4 REFERENCES 1 Cotes JE, Russitee CE, Higgins In: GilsonJC. Average normal values for the forced expiratory volumes in white caucasian males. Br Med J 1966;312:1018-19 2 Oscherowitz M, Edlavitch SA, Baker TR, Jarboe 1: Differences in pulmonary functions in various racial groups. Am J Epidem 1972;96:319-27 3 De Costa JL. Pulmonary function studies in healthy Chinese adults in Singapore. Am Rev Respir Dis 1971;104:128-31 4 Mustafa KY. Spirometric lung function tests in normal men of African ethnic origin. Am Rev Respir Dis 1977; 116:209-13 5 Cookson JB, Blake G~ Faranistic C. Normal values for CHEST I 93 I 5 I MAY, 1888

1021

6 7

8 9

ventilatory function in Rhodesian Africans. Br J Dis Chest 1976; 70:107-11 Menegesha YA, Mekonen Y. Spirometric lung function tests in normal non-smoking Ethiopian men and women. Thorax 1985; 40:465-68 Hsi B~ Hsu KHK, Jenkins DE. Ventilatory function of normal children and young adults: Mexican-American, white and blacks: III. Sitting height as predictor. J Pediatr 1983; 102:860-65 Massey DG, Fournier-Massey G. Japanese-Americans pulmonary reference values: influence of environment on anthropology and physiology Env Res 1986;39:418-33 Miller AC, Thornton JC. The interpretation of spirometric measurements in epidemiologic surveys. Env Res 1980; 23:44468

1030

10 Dockrey D~ Berkey C, Ware J8, Speizer FE, Ferris BG Jr. Distribution offorced vitalcapacity and forced expiratory volume in one second in children 6 to 11 years of age. Am Rev Respir Dis 1983; 128:405-12 11 Lourie JA. Spirometric studies of Yemenite and Kurdish Jews. Phil Thm Roy Soc London 1973;266:113-19 12 Goren A,Bruderman I. Pulmonary functions and respiratory symptoms and diseases among adult Israelis: variation by country of origin. Isr J Moo Sci 1986;22:761-65 13 Knudson RJ, Lebowitz MD, Holberg CJ.Changes in the normal maximal expiratory Bow-volume curve with growth and aging. Am Rev Respir Dis 1983; 127:725-34 14 European economic community Bull Europ Physiopath Respir 1983;45-51

Racialor Ethnic vartation In Spirometric Lung Function Norms (Blb/, Goldsmith, \Wdl)