Abstract

Dietary antioxidants, waist circumference, and pulmonary function were measured in the Fourth Scottish MONICA cross-sectional survey of 865 men and 971 women aged 25–64 years. Waist circumference was inversely related to forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC), even after adjustment for age, height, weight, working status, energy intake, and smoking variables in a multiple linear regression model (men: β = −0.017 for FEV1 l/cm, p < 0.01 and β = −0.008 for FVC, p = 0.04; women: β = −0.009 for FEV1, p < 0.01 and β = −0.007 for FVC, p = 0.01). After additional adjustment for waist circumference, estimated vitamin C and β-carotene intakes were positively associated with lung function in men (vitamin C: β = 0.102 for FEV1 l/mg/day, p = 0.03; β-carotene: β = 0.073 for FVC l/μg/day, p = 0.02). Retinol and vitamin E were not significantly related to lung function for either sex. A case-control study of airway obstruction showed that waist circumference was significantly associated, while vitamin C could be protective. The study suggests that adequate intake of antioxidants and avoidance of increasing girth could help to preserve lung function.

Lung function is important in medical practice, as it is predictive of both morbidity (1, 2) and mortality (3, 4). It is known that age and anthropometric characteristics affect lung function and that environmental toxic free radicals/oxidants, for example, occupational exposure and smoking, reduce lung function. Conversely, dietary vitamin C, an important antioxidant, could potentially protect lung function because it can directly neutralize free radicals and suppress macrophage secretion of superoxide anions, and it acts at the lipid membrane (5). Several recent epidemiologic studies have reported a relation between dietary vitamin C and lung function (68). However, the association of other dietary antioxidant vitamins with lung function has been investigated less often in community studies. While large waist circumference predicts an excess burden of ill health (2, 9, 10), little is known about the effect of waist circumference on lung function and airway obstruction. To investigate the relation of dietary antioxidants and waist circumference with pulmonary function and airway obstruction, we carried out analyses within the Fourth Scottish MONICA population survey.

MATERIALS AND METHODS

Adults aged 25–64 years were recruited randomly from general practitioner lists in north Glasgow, United Kingdom, for the Fourth Scottish MONICA survey in 1995, which followed a procedure similar to the Scottish Heart Health Study (11). Patients were stratified by sex and 10-year age bands and were sampled to obtain similar numbers of participants of each age. They were asked to complete a self-administered personal health record, including details on smoking and a food frequency questionnaire (12). They brought the health record to the survey clinic, where the survey team of nurses checked it for completeness.

Height, weight, and waist circumference were measured by using standard techniques (13), as follows: height to within 0.5 cm, without shoes, by using a portable stadiometer (Holtain, Crymych, United Kingdom); weight to within 100 g, without heavy clothing, by using digital scales (Seca, Hamburg, Germany) that were calibrated weekly; and waist circumference to within 0.1 cm by using plastic measuring tapes, with the waist defined as midway between the lowest rib and the iliac crest as the participant breathed out gently. Forced expiratory volume in the first second (FEV1) and forced vital capacity (FVC) were measured by using a Microlab 3300 electronic spirometer (Micro Medical Limited, Kent, England). Standardized field-measurement procedures were used, and subjects were seated and did not use nose clips. The study nurse demonstrated the technique, emphasizing the tight fit between lips and tube and encouraging the subject to breathe out as long and forcefully as possible. After one trial run, the best of three technically satisfactory maneuvers was recorded by the machine. A nonfasting blood sample was also taken for cotinine measurement (a stable metabolite of nicotine) and other analyses.

On the basis of the personal health record information, smoking status was defined as current, former, or never. Asked the question, “Do you smoke cigarettes now?” current cigarette smokers responded positively with, “Yes, regularly.” Former smokers responded negatively to this question but positively to the question, “Did you ever smoke cigarettes?” Never smokers gave negative answers to both of these questions and smoked neither a cigar nor a pipe. The questions regarding amount of smoking were the following: “On average, about how many cigarettes do you now smoke a day?” for current smokers, and “What is the maximum number of cigarettes you ever smoked a day for as long as a year?” for both current and former smokers. Smoking pack-years were estimated for both current and former smokers by multiplying the maximum number of cigarettes smoked for as long as 1 year by the number of years of smoking. Dietary antioxidant intakes were calculated by using a computer program that multiplied the reported frequency of intake by a standard portion size for each food and linked that information with the average nutrient composition of the foods; the United Kingdom references tables were used (14).

The SPSS statistical package (Windows version 8.0; SPSS Inc., Chicago, Illinois) was used to analyze the data. Smoking was considered as an important confounding effect in this study. We used serum cotinine concentration (measuring recent exposure to tobacco smoke (11) and, for the current smokers, relating that information to the number of cigarettes smoked) and smoking pack-years for adjustment. A square-root transformation was made for cotinine concentration and smoking pack-years. Natural logarithmic transformation was applied for all of the antioxidants, and linear regression analysis was used to examine the relation between antioxidants, waist circumference, and lung function. Lung function data for men and women were analyzed separately. Age, height, weight, working status, total energy intake, and smoking variables (serum cotinine and smoking pack-years) were considered as adjustment factors in multiple linear regression analyses (the basic model). The quadratic term on age squared was added to all models because it improved the visual impression of linearity in plots of residuals by age; weight was squared in the analysis to reduce the collinearity between weight and waist circumference in the model. In addition, the variables age, age squared, height, weight squared, and waist circumference were centered in the model to minimize any possible collinearity. In these multiple linear regression models, a subanalysis for the current, former, and never smokers was performed to examine effects of antioxidants on lung function.

We used the definition of airway obstruction as the ratio of FEV1 to FVC less than or equal to 0.75 (15) to carry out a case-control study of airway obstruction to further examine the effects of antioxidants and waist circumference on lung function. Data for both sexes were combined to estimate the association of antioxidants and waist circumference with airway obstruction by using the logistic regression model to derive odds ratios. Data on men and women were divided separately into fourths according to the quartile values for waist circumference and antioxidants in the model, while age, height, weight, total energy intake, cotinine levels, and smoking years were kept in continuous form for adjustment.

RESULTS

A total of 1,836 subjects (865 men, 971 women) participated in the Fourth Scottish MONICA survey (65.8 percent response rate); 846 men and 958 women had their lung function measured. Significant differences in lung function, waist circumference, and antioxidant vitamin intakes were found between the two sexes. Men had a higher FEV1 and FVC, a larger waist circumference, and higher retinol and vitamin E (α-tocopherol equivalents) intakes but lower vitamin C and β-carotene intakes. A difference was also found in height, weight, cotinine levels, smoking pack-years, and energy intake but not in age (table 1). There were similar proportions of men (44.4 percent) and women (44.3 percent) not working and of current smokers (43.0 and 41.1 percent, respectively), but more men than women worked full-time (53.0 vs. 35.7 percent, respectively; p < 0.001) and fewer were never smokers (31.5 vs. 39.9 percent, respectively; p < 0.001).

TABLE 1.

Lung function and relevant factors* for men and women participants in the Fourth Scottish MONICA Survey, 1995

Factor Men
 
Women
 
Mean SD Mean SD 
FEV1 (liter) 3.28 0.81 2.42 0.64 
FVC (liter) 3.87 0.89 2.80 0.71 
Waist circumference (cm) 91.9 11.9 80.3 12.8 
Vitamin C (mg/day) 72.2 34.1 76.0 38.8 
β-Carotene (μg/day) 3,202 2,142 3,761 2,398 
Retinol (μg/day) 570 395 484 362 
Vitamin E (mg/day) 8.1 4.2 7.2 3.52 
Age (years) 46.35 11.34 45.50 11.70 
Height (m) 1.723 0.070 1.597 0.064 
Weight (kg) 78.95 13.94 67.31 14.39 
Cotinine (ng/ml) 131.4 179.6 122.1 171.4 
Smoking pack-years 35.31 27.15 27.41 20.52 
Total energy intake (kJ/day) 2,578.6 695.0 2,019.4 551.0 
Factor Men
 
Women
 
Mean SD Mean SD 
FEV1 (liter) 3.28 0.81 2.42 0.64 
FVC (liter) 3.87 0.89 2.80 0.71 
Waist circumference (cm) 91.9 11.9 80.3 12.8 
Vitamin C (mg/day) 72.2 34.1 76.0 38.8 
β-Carotene (μg/day) 3,202 2,142 3,761 2,398 
Retinol (μg/day) 570 395 484 362 
Vitamin E (mg/day) 8.1 4.2 7.2 3.52 
Age (years) 46.35 11.34 45.50 11.70 
Height (m) 1.723 0.070 1.597 0.064 
Weight (kg) 78.95 13.94 67.31 14.39 
Cotinine (ng/ml) 131.4 179.6 122.1 171.4 
Smoking pack-years 35.31 27.15 27.41 20.52 
Total energy intake (kJ/day) 2,578.6 695.0 2,019.4 551.0 
*

All means between men and women, except for age, were significantly different at p < 0.05.

SD, standard deviation; FEV1, forced expiratory volume in the first second; FVC, forced vital capacity.

For current and former smokers.

A simple linear regression model (table 2) showed a significant inverse relation between waist circumference and lung function for men and women. Vitamin C intake was positively associated with lung function for both sexes, but retinol intake was inversely related. The associations of β-carotene and vitamin E intake with lung function were not found to be significant. In multiple linear regression, a significant relation was found between waist circumference and lung function for both sexes; that is, waist circumference was significantly inversely associated with FEV1 and FVC (table 3). In a separate analysis of fifths of weight for men and women, the negative effect of waist circumference on lung function was slightly greater in the heaviest versus the lightest weight group (data not shown). In the models, lung function was also significantly related to age, height, smoking variables, and other factors, which varied between men and women.

TABLE 2.

Simple linear regression between lung function and relevant factors for men and women participants in the Fourth Scottish MONICA Survey, 1995

Variable Men
 
Women
 
FEV1
 
FVC
 
FEV1
 
FVC
 
β SE β SE β SE β SE 
Waist circumference (cm) −0.013 0.002* −0.013 0.003* −0.010 0.002* −0.010 0.002* 
Vitamin C (mg/day) 0.261 0.060* 0.210 0.066* 0.111 0.040* 0.128 0.044* 
β-Carotene (μg/day) −0.002 0.037 0.040 0.040 0.021 0.028 0.035 0.031 
Retinol (μg/day) −0.208 0.043* −0.200 0.047* −0.179 0.030* −0.194 0.033* 
Vitamin E (mg/day) 0.061 0.063 0.073 0.069 −0.059 0.048 −0.071 0.052 
Variable Men
 
Women
 
FEV1
 
FVC
 
FEV1
 
FVC
 
β SE β SE β SE β SE 
Waist circumference (cm) −0.013 0.002* −0.013 0.003* −0.010 0.002* −0.010 0.002* 
Vitamin C (mg/day) 0.261 0.060* 0.210 0.066* 0.111 0.040* 0.128 0.044* 
β-Carotene (μg/day) −0.002 0.037 0.040 0.040 0.021 0.028 0.035 0.031 
Retinol (μg/day) −0.208 0.043* −0.200 0.047* −0.179 0.030* −0.194 0.033* 
Vitamin E (mg/day) 0.061 0.063 0.073 0.069 −0.059 0.048 −0.071 0.052 
*

p < 0.01.

Natural logarithmic transformation for all vitamin variables.

FEV1, forced expiratory volume in the first second; FVC, forced vital capacity; SE, standard error.

TABLE 3.

Basic multiple linear regression between lung function and waist circumference for men and women participants in the Fourth Scottish MONICA Survey, 1995

Variable FEV1
 
FVC
 
Men
 
Women
 
Men
 
Women
 
β SE β SE β SE β SE 
Waist circumference (cm) −0.017 0.004** −0.009 0.002** −0.008 0.004* −0.007 0.003* 
Age (years) −0.029 0.002** −0.029 0.001** −0.028 0.003** −0.029 0.002** 
Age2 −0.0003 <0.001 0.0004 <0.001** −0.0003 <0.001 −0.0005 <0.001** 
Height (m) 3.86 0.371** 3.21 0.238** 5.39 0.437** 4.166 0.274** 
Weight (kg)2 0.00006 <0.001** 0.00004 <0.001** 0.000002 <0.001 0.00002 <0.001 
Working status −0.091 0.024** −0.0030 0.017 −0.1170 0.028** −0.0371 0.020 
Total energy intake (kJ/day) 0.00004 <0.001 −0.00008 <0.001** 0.00002 <0.001 −0.00009 <0.001** 
Cotinine (ng/ml)§ −0.004 0.003 −0.009 0.002** −0.005 0.004 −0.012 0.003** 
Smoking pack-years§ −0.0306 0.0090** −0.0107 0.0070 −0.01146 0.0100 −0.0017 0.0080 
Intercept 3.64 0.054** 2.649 0.042** 4.220 0.063** 3.042 0.048** 
Variable FEV1
 
FVC
 
Men
 
Women
 
Men
 
Women
 
β SE β SE β SE β SE 
Waist circumference (cm) −0.017 0.004** −0.009 0.002** −0.008 0.004* −0.007 0.003* 
Age (years) −0.029 0.002** −0.029 0.001** −0.028 0.003** −0.029 0.002** 
Age2 −0.0003 <0.001 0.0004 <0.001** −0.0003 <0.001 −0.0005 <0.001** 
Height (m) 3.86 0.371** 3.21 0.238** 5.39 0.437** 4.166 0.274** 
Weight (kg)2 0.00006 <0.001** 0.00004 <0.001** 0.000002 <0.001 0.00002 <0.001 
Working status −0.091 0.024** −0.0030 0.017 −0.1170 0.028** −0.0371 0.020 
Total energy intake (kJ/day) 0.00004 <0.001 −0.00008 <0.001** 0.00002 <0.001 −0.00009 <0.001** 
Cotinine (ng/ml)§ −0.004 0.003 −0.009 0.002** −0.005 0.004 −0.012 0.003** 
Smoking pack-years§ −0.0306 0.0090** −0.0107 0.0070 −0.01146 0.0100 −0.0017 0.0080 
Intercept 3.64 0.054** 2.649 0.042** 4.220 0.063** 3.042 0.048** 
*

p < 0.05

**

p < 0.01.

FEV1, forced expiratory volume in the first second; FVC, forced vital capacity; SE, standard error.

Working status: 1, full-time; 2, part-time; 3, not working.

§

Square-root transformation.

By using the basic multiple linear regression model, including waist circumference, we examined the dietary antioxidant intakes individually and found that vitamin C and β-carotene were significantly positively associated with lung function in men and that β-carotene also seemed to be positively associated with FVC in women, while retinol and vitamin E were not significantly related to lung function (table 4). In these analyses, waist circumference consistently showed the same significant inverse relation with lung function as reported in table 3. When we considered all of the antioxidants together in the model, we observed that vitamin C was still associated with FEV1 in men (β = 0.109, p < 0.05), while β-carotene was associated with both FEV1 (β = 0.042, p = 0.05) and FVC (β = 0.049, p < 0.05) in women; the findings regarding the other antioxidants were not significant. Furthermore, when analyses were performed separately for smoking status, we found that the significance of the association between retinol and lung function in women former smokers increased (β = 0.09 for FEV1, p = 0.07; β = 0.08 for FVC, p = 0.13). All other results were similar to those shown in table 4, with a reduced level of significance, probably due to the smaller numbers.

TABLE 4.

Multiple linear regression* between lung function and antioxidant vitamins for men and women participants in the Fourth Scottish MONICA Survey, 1995

Variable Men
 
Women
 
FEV1
 
FVC
 
FEV1
 
FVC
 
β SE p value β SE p value β SE p value β SE p value 
Vitamin C (mg/day) 0.102 0.048 0.03 0.074 0.056 0.19 −0.019 0.030 0.52 −0.004 0.034 0.90 
β-Carotene (μg/day) 0.038 0.027 0.16 0.073 0.031 0.02 0.030 0.019 0.12 0.041 0.022 0.07 
Retinol (μg/day) −0.051 0.035 0.15 −0.059 0.041 0.15 0.021 0.023 0.37 0.009 0.027 0.73 
Vitamin E (mg/day) −0.055 0.048 0.25 −0.019 0.057 0.73 0.039 0.035 0.26 0.033 0.040 0.40 
Variable Men
 
Women
 
FEV1
 
FVC
 
FEV1
 
FVC
 
β SE p value β SE p value β SE p value β SE p value 
Vitamin C (mg/day) 0.102 0.048 0.03 0.074 0.056 0.19 −0.019 0.030 0.52 −0.004 0.034 0.90 
β-Carotene (μg/day) 0.038 0.027 0.16 0.073 0.031 0.02 0.030 0.019 0.12 0.041 0.022 0.07 
Retinol (μg/day) −0.051 0.035 0.15 −0.059 0.041 0.15 0.021 0.023 0.37 0.009 0.027 0.73 
Vitamin E (mg/day) −0.055 0.048 0.25 −0.019 0.057 0.73 0.039 0.035 0.26 0.033 0.040 0.40 
*

All adjusted variables were taken from the basic model shown in table 3.

Natural logarithmic transformation.

FEV1, forced expiratory volume in the first second; FVC, forced vital capacity; SE, standard error.

In the case-control study, 112 men and 77 women were identified as cases, while all 252 men and 297 women who had FEV1/FVC ratios of more than 0.90 were taken as controls. Both univariate and multivariate analyses showed that the odds ratio of airway obstruction increased significantly with waist circumference but decreased with vitamin C intake (table 5). The risk of airway obstruction was significantly lower in the second-lowest fourth of β-carotene intake compared with the lowest fourth, while retinol and vitamin E intakes were not significantly associated with airway obstruction. In a separate analysis of smoking status, all results were similar to those shown in table 5 or were not significant because of the smaller numbers. In the multivariate analysis, after an additional adjustment for smoking status, the results did not change.

TABLE 5.

Risk (odds ratios) of airway obstruction, by waist circumference and quartile of dietary antioxidant vitamin level, Fourth Scottish MONICA Survey, 1995

Variable and quartile* 
Crude Adjusted 
OR 95% CI OR 95% CI 
Waist circumference (cm)  
 1 1.00  1.00  
 2 1.75 1.05, 2.92 2.30 1.13, 4.68 
 3 2.21 1.34, 3.62 3.23 1.41, 7.39 
 4 2.15 1.30, 3.53 4.90 1.60, 15.02 
Vitamin C (mg/day)  
 1 1.00  1.00  
 2 0.58 0.37, 0.92 0.60 0.32, 1.12 
 3 0.35 0.21, 0.57 0.45 0.22, 0.90 
 4 0.52 0.33, 0.83 0.60 0.27, 1.15 
β-carotene (μg/day)  
 1 1.00  1.00  
 2 0.55 0.34, 0.96 0.40 0.21, 0.78 
 3 1.08 0.68, 1.72 0.93 0.49, 1.76 
 4 1.11 0.70, 1.77 1.11 0.56, 2.22 
Retinol (μg/day)  
 1 1.00  1.00  
 2 1.23 0.73, 2.06 1.28 0.64, 2.56 
 3 1.89 1.15, 3.12 1.33 0.69, 2.75 
 4 1.72 1.05, 2.83 0.96 0.46, 1.99 
Vitamin E (mg/day)  
 1 1.00  1.00  
 2 0.73 0.44, 1.21 0.89 0.45, 1.75 
 3 1.12 0.70, 1.79 1.75 0.90, 3.39 
 4 0.77 0.47, 1.25 0.77 0.38, 1.54 
Variable and quartile* 
Crude Adjusted 
OR 95% CI OR 95% CI 
Waist circumference (cm)  
 1 1.00  1.00  
 2 1.75 1.05, 2.92 2.30 1.13, 4.68 
 3 2.21 1.34, 3.62 3.23 1.41, 7.39 
 4 2.15 1.30, 3.53 4.90 1.60, 15.02 
Vitamin C (mg/day)  
 1 1.00  1.00  
 2 0.58 0.37, 0.92 0.60 0.32, 1.12 
 3 0.35 0.21, 0.57 0.45 0.22, 0.90 
 4 0.52 0.33, 0.83 0.60 0.27, 1.15 
β-carotene (μg/day)  
 1 1.00  1.00  
 2 0.55 0.34, 0.96 0.40 0.21, 0.78 
 3 1.08 0.68, 1.72 0.93 0.49, 1.76 
 4 1.11 0.70, 1.77 1.11 0.56, 2.22 
Retinol (μg/day)  
 1 1.00  1.00  
 2 1.23 0.73, 2.06 1.28 0.64, 2.56 
 3 1.89 1.15, 3.12 1.33 0.69, 2.75 
 4 1.72 1.05, 2.83 0.96 0.46, 1.99 
Vitamin E (mg/day)  
 1 1.00  1.00  
 2 0.73 0.44, 1.21 0.89 0.45, 1.75 
 3 1.12 0.70, 1.79 1.75 0.90, 3.39 
 4 0.77 0.47, 1.25 0.77 0.38, 1.54 
*

Quartile 1, lowest; quartile 4, highest.

Waist circumference and the antioxidants were adjusted for each other and for age, sex, height, weight squared, working status, total energy intake, cotinine levels, and smoking pack-years (square root).

OR, odds ratio; CI, confidence interval.

DISCUSSION

The strength of this study was the highly standardized data on a random population sample, which included information on waist circumference and lung function (both FEV1 and FVC) and measurements of four dietary antioxidants and serum cotinine. The weakness was the lack of information on the average daily number of cigarettes smoked by former smokers. However, this weakness was overcome by estimating pack-years from the maximum number of cigarettes smoked a day for as long as a year; the robustness of the results was consistent in the separate analysis of smoking status. We used this estimate because it was available for both former and current smokers and was significantly related to the amount of smoking by current smokers; however, adjusting for either the average daily amount or the maximum number of cigarettes smoked a day for as long as a year did not change the results. Our findings from this population-based study suggest that waist circumference and intake of some antioxidant vitamins are significantly and independently associated with lung function.

Waist circumference measurement has been increasingly related to ill health (10). It reflects total and abdominal fat accumulation and is not greatly influenced by height (16). In a recent Australian study of the effects of body composition and fat distribution on ventilatory function, Lazarus et al. (17) investigated a single measurement, FVC only, and observed that it was significantly negatively associated with waist circumference in men but not in women. Analyzing Dutch population data, Lean et al. found that waist circumference was significantly related to “shortness of breath when walking uphill or upstairs” (10, p. 855). In the present study, we used more objective indices—both FEV1 and FVC—to identify the relation of waist circumference to lung function with, importantly, adjustment for height and weight. Our findings indicated that lung function (both FEV1 and FVC) was negatively associated with waist circumference for both sexes, even in multiple linear regression analyses that included important factors such as weight, smoking, and antioxidant vitamin intake, and that a larger waist circumference was associated with airway obstruction in a monotonic dose-response relation. The reason for this significant negative association is not entirely clear, but it seems likely that a large waist circumference could have mechanical effects on lung function, that is, at least partially affecting movement of the diaphragm and chest wall. However, a large waist circumference is associated with fat, a metabolically active tissue, and other biologic effects cannot be ruled out.

The relation between diet and lung diseases has created recent interest (18); it has been suggested that the balance of oxidants and antioxidants may be important in the pathophysiology of chronic obstructive pulmonary disease. Dietary intake, a major source of nonenzymatic antioxidants, including vitamin C, appears relevant to the risk of this disease. Our study showed that increased vitamin C intake was associated with a reduced risk of airway obstruction, and our lung function data for men are consistent with the findings of previous population-based studies in the United States, England, and China (68) both in the direction and magnitude of the association of vitamin C intake with lung function. The results for FEV1 were somewhat more consistent than those for FVC and were similar to previous reports (7, 8).

In our data, an increase of 50 mg/day in vitamin C intake from 50–100 mg/day was associated with an increase of 70.7 ml in FEV1 for men; in the English data (both sexes combined) (7), a 40 mg/day higher vitamin C intake led to a 25.0 ml FEV1. A US study (6) showed that a 40 ml difference in FEV1 occurred from the lowest (17 mg/day) to the highest (178 mg/day) thirds of dietary vitamin C intake, while Chinese data (8) from a rural population suggested that an increase of 100 mg/day in vitamin C intake was associated with an increase of 21.6 ml in FEV1. In our study, the effect of vitamin C on lung function was greater than that reported in other investigations. This finding could be due to the separate analysis by sex or a highly sensitive target population in our study, because Scottish people had lower estimated mean vitamin C intakes (19) than these other populations did—for example, 88 mg/day in the US study (6), 99 mg/day in the English data (7), and 151 mg/day in the Chinese rural population (8) versus 72 mg/day for men and 76 mg/day for women in the present study.

In the multivariate linear regression models, the lack of a significant association of vitamin C with lung function for women could have been due to other factors. For example, total energy intake was a significant predictor in women (not in men) for both measures of lung function. The difference between men and women regarding the relation between total energy intake and waist circumference may be due to a greater propensity for men to gain weight around the waist, as distinct from the hips, given the same energy intake or to differences in activity level. Cotinine concentration was a significant predictor in women (not in men), suggesting possible effects of passive smoking or indoor air pollution (2022) on women's lung function. In addition, women had a significantly higher intake of vitamin C than men did (table 1); although its effect on lung function was significant in univariate analysis, the effect size was smaller than that in men (about half) (table 2). These findings could at least partly explain the lack of a significant effect of vitamin C on lung function in women.

This lack of a significant association of vitamin C with lung function in women has recently been reported by others as well. Ness et al. (23) investigated 835 men and 1,025 women aged 45–75 years registered with general practitioner practices in Norfolk, England, and found that although plasma vitamin C was positively correlated, after adjustment for age and height, with both FEV1 and FVC in men, the association in women was weaker and was not statistically significant. The difference between men and women could be attributed in part to a greater error in the lung function measures in women, since they were noted to have poorer compliance with the technique than men did (23). The present investigation adds to current knowledge by demonstrating a link between vitamin C intake and pulmonary function, including airway obstruction from Scottish data, after adjustment for waist circumference.

This population-based study also provides evidence that dietary β-carotene intake (derived primarily from carrots and green vegetables plus fruits) protects lung function. Studies that measured serum β-carotene have shown similar results. For example, van Antwerpen et al. (24) investigated 30 asymptomatic male cigarette smokers and 34 age-matched nonsmoking controls and found that for the smokers, but not the nonsmokers, plasma β-carotene levels correlated significantly and positively with FEV1. Chuwers et al. (25) found that among asbestos-exposed men with a high rate of current and former cigarette smoking, serum β-carotene concentration was significantly related to ventilatory function. Concerning retinol, our data show only a possible positive relation for the women former smokers, while Chuwers et al. reported, in the same study of asbestos-exposed men, that serum retinol levels had a protective effect against loss of lung function. Morabia et al. (15) investigated 83 White male subjects and found a significant effect of serum retinol concentration on airway obstruction, a finding that was the same as in their previous study of dietary vitamin A (milk intake) and airway obstruction (26).

The US and Chinese studies (6, 8) did not report vitamin E or retinol data in their papers on the relation between dietary vitamin C intake and lung function. Others have examined the relation between vitamin E and lung function. In a study of 178 men and women aged 70–96 years selected on the basis of reported respiratory symptoms, Dow et al. (27) found that vitamin E intake was significantly associated with FEV1 and FVC. These authors suggested that dietary intake of vitamin E may influence lung function in the elderly but that food frequency questionnaires were not sufficiently sensitive to explore this hypothesis further. Britton et al. (7) indicated that in data from Nottingham, England, a significant association was found between vitamin E intake and lung function in initial analyses; however, vitamin C and vitamin E intakes were correlated (r = 0.29, p < 0.001), and, after adjusting for the effects of vitamin C, no additional independent effect of vitamin E on either FEV1 or FVC was found. As mentioned by Dow et al., food composition data are often inadequate for quantifying vitamin E intake; as such, estimates of dietary intake may correlate poorly with serum vitamin E concentrations. The current food frequency questionnaire had previously been shown to produce good to moderate correlation between diet and plasma vitamin C and β-carotene but not between retinol and vitamin E (12).

Our findings from this general population survey suggest that waist circumference and vitamin C and β-carotene intakes are significantly associated with lung function, independently of other relevant factors. The apparent absence of an association with antioxidants whose estimated intakes are less reliable does not preclude their having a role. A longitudinal study is needed to investigate the effect of dietary antioxidants and waist circumference on age-related deterioration in lung function.

Reprint requests to Dr. Ruoling Chen, Cardiovascular Epidemiology Unit, University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, Scotland, United Kingdom (e-mail: rchen@cve.dundee.ac.uk).

This study was supported by the Scottish Office and British Heart Foundation.

REFERENCES

1.
Kannel WB, Hubert H, Lew EA. Vital capacity as a predictor of cardiovascular disease: the Framingham study.
Am Heart J
 
1983
;
105
:
311
–15.
2.
Welin L, Svardsudd K, Wilhelmsen L, et al. Analysis of risk factors for stroke in a cohort of men born in 1913.
N Engl J Med
 
1987
;
317
:
521
–6.
3.
Rodriguez BL, Masaki K, Burchfiel C, et al. Pulmonary function decline and 17-year total mortality: the Honolulu Heart Program.
Am J Epidemiol
 
1994
;
140
:
398
–408.
4.
Hole DJ, Watt GC, Davey-Smith G, et al. Impaired lung function and mortality risk in men and women: findings from the Renfrew and Paisley prospective population study.
BMJ
 
1996
;
313
:
711
–15.
5.
McGowan SE, Parenti CM, Hoidal JR, et al. Ascorbic acid content and accumulation by alveolar macrophages from cigarette smokers and nonsmokers.
J Lab Clin Med
 
1984
;
104
:
127
–34.
6.
Schwarts J, Weiss ST. Relationship between dietary vitamin C intake and pulmonary function in the First National Health and Nutrition Examination Survey (NHANES I).
Am J Clin Nutr
 
1994
;
59
:
110
–14.
7.
Britton JR, Pavord ID, Richards KA, et al. Dietary antioxidant vitamin intake and lung function in the general population.
Am J Respir Crit Care Med
 
1995
;
151
:
1383
–7.
8.
Hu G, Zhang X, Chen J, et al. Dietary vitamin C intake and lung function in rural China.
Am J Epidemiol
 
1998
;
148
:
594
–9.
9.
Han TS, van Leer EM, Seidell JC, et al. Waist circumference action levels in the identification of cardiovascular risk factors: prevalence study in a random sample.
BMJ
 
1995
;
311
:
1401
–5.
10.
Lean MEJ, Han TS, Seidell JC. Impairment of health and quality of life in people with large waist circumference.
Lancet
 
1998
;
351
:
853
–6.
11.
Tunstall-Pedoe H, Woodward M, Tavendale R, et al. Comparison of the prediction by 27 different factors of coronary heart disease and death in men and women of the Scottish Heart Health Study: cohort study.
BMJ
 
1997
;
315
:
722
–9.
12.
Bolton-Smith C, Casey CE, Gey KF, et al. Antioxidant vitamin intakes assessed using a food-frequency questionnaire: correlation with biochemical status in smokers and non-smokers.
Br J Nutr
 
1991
;
65
:
337
–46.
13.
World Health Organization. Measuring obesity: classification and distribution of anthropometric data. Copenhagen, Denmark: World Health Organization, 1989. (Nutr UD, EUR/ICP/NUT 125).
14.
Holland B, Welch AA, Unwin ID, et al. McCance and Widdowson's the composition of foods. 5th ed (revised and extended). London, United Kingdom: HMSO, 1991.
15.
Morabia A, Menkes MJS, Comstock GW, et al. Serum retinol and airway obstruction.
Am J Epidemiol
 
1990
;
132
:
77
–82.
16.
Han TS, Seidell JC, Curral JEP, et al. The influence of height and age on waist circumference as an index of adiposity in adults.
Int J Obes
 
1997
;
21
:
83
–98.
17.
Lazarus R, Gore CJ, Booth M, et al. Effects of body composition and fat distribution on ventilatory function in adults.
Am J Clin Nutr
 
1998
;
68
:
35
–41.
18.
Sridhar MK. Nutrition and lung health.
BMJ
 
1995
;
310
:
75
–6.
19.
Gregory J, Foster K, Tyler H, et al. The Dietary and Nutritional Survey of British Adults. Office of Population Censuses and Surveys (Social Survey Division). London, United Kingdom: HMSO, 1990.
20.
Coggon D. Air pollution in homes. (Editorial).
BMJ
 
1996
;
312
:
1316
.
21.
Spitzer WO, Lawrence V, Dales R, et al. Links between passive smoking and diseases: a best-evidence synthesis. A report of the Working Group on Passive Smoking.
Clin Invest Med
 
1990
;
13
:
17
–42.
22.
Neas LM, Dockery DW, Ware JH, et al. Concentration of indoor particulate matter as a determinant of respiratory health in children.
Am J Epidemiol
 
1994
;
139
:
1088
–99.
23.
Ness AR, Khaw KT, Bingham S, et al. Vitamin C status and respiratory function.
Eur J Clin Nutr
 
1996
;
50
:
573
–9.
24.
van Antwerpen VL, Theron AJ, Richards GA, et al. Relationship between the plasma levels of beta-carotene and lung functions in cigarette smokers.
Int J Vitam Nutr Res
 
1995
;
65
:
231
–5.
25.
Chuwers P, Barnhart S, Blanc P, et al. The protective effect of beta-carotene and retinol on ventilatory function in an asbestos-exposed cohort.
Am J Respir Crit Care Med
 
1997
;
155
:
1066
–71.
26.
Morabia A, Sorenson A, Kumanyika SK, et al. Vitamin A, cigarette smoking, and airway obstruction.
Am Rev Respir Dis
 
1989
;
140
:
1312
–16.
27.
Dow L, Tracey M, Villar A, et al. Does intake of vitamins C and E influence lung function in older people?
Am J Respir Crit Care Med
 
1996
;
154
:
1401
–4.