The population in Thailand is aging and life expectancy has been increasing. This creates an increased use of health care resources. Better understanding of the physiology and pathology of the elderly is needed to decrease unnecessary invasive investigations and avoid misdiagnosis of important diseases.
Chronic respiratory diseases are a significant health problem in the elderly and tend to relate to increasing age [1]. Aging and accumulating exposure to environmental toxins can lead to numerous anatomical, physiological, and immunological modifications of the respiratory system [2-8]. Previous studies investigated the aging respiratory system. A study of rats demonstrated collagen accumulation and progressive fibrosis of the aging lung [9]. Microscopic examination of aging human lungs shows increasing airspace size without evidence of destruction of the alveolar wall [10-12]. Plain radiographs reveal hyperinflation of the lungs and bullous spaces in the lungs of both normal and emphysematous patients [13]. Several computed tomography (CT) studies examining lung morphology in elderly revealed air trapping [6, 14, 15], a subpleural basal reticular pattern, lung cysts [16] and increases in the bronchoarterial ratio, which reflects bronchiectasis [16, 17].
Although CT studies of morphological changes in aging lungs are available, most are limited to a subgroup of urban dwellers. The results are not truly representative of the general Thai elderly population in respect of ethnicity, environment, and prevalent diseases.
The purpose of the present study was to describe morphological changes of lungs in elderly Thais without any respiratory illnesses in comparison with younger individuals.
The institutional review board of the Medical Faculty of Chulalongkorn University (Bangkok, Thailand) reviewed our study protocols and gave approval for additional limited CT examination and for submission of data for publication (IRB 296/56). Written informed consent to participate in this study and undergo the additional CT examination was obtained from each patient, or their nearest relative or legal guardian in the case of senile patients or those not capable of providing their own consent.
This cross-sectional study included prospectively selected patients was designed to evaluate the morphological changes in lungs by performing limited additional thoracic CT in Thai patients classified into three age groups; 20-40 years, 41-60 years, and >60 years.
The enrolled participants were Thai patients who had been referred for CT of the abdomen or head and neck at King Chulalongkorn Memorial Hospital from October 2013 to September 2014. Participants with symptoms or problems involving the respiratory system or related systems, which might affect morphological changes of the lungs defined by clinically constructed questionnaire were excluded from the study. Dyspnea was graded using the modified Medical Research Council breathlessness scales [18] (grade 0 = breathless with strenuous exercise, grade 1 = breathless when hurrying on level ground or walking up hill, grade 2 = breathless when walking on the level, grade 3 = breathless after walking about 100 yards, grade 4 = breathless at rest).
Exclusion criteria were active pulmonary diseases, any respiratory symptom other than grade 1 dyspnea, primary lung cancers or lung metastasis, morphological abnormalities on previous chest X-ray or CT images, any known chronic pulmonary disease, known cardiac failure with pulmonary edema or pulmonary hypertension at the time of the CT study, known connective tissue diseases affecting the respiratory tract, previous thoracic operations or radiation, previous trauma affecting lungs, and pregnancy. Any patients with confusion, inability to stay still in a supine position, or inability to hold their breath were also excluded from our study.
Sixty-seven consecutive asymptomatic patient volunteers (20 years or older) were enrolled after receiving complete comprehensible information concerning the purpose of this study. Smoking history in pack-years, dyspnea scores, and underlying disease were recorded. Seven participants failed to meet these criteria and were excluded from our study. The remaining 60 patient participants were 20 in each age group.
Limited-sliced supine inspiratory and expiratory CT imaging of the thorax was performed using Somatoms sensation 16 (Siemens, Germany) and Discovery CT750 HD (GE healthcare, United Kingdom) with scanning parameters as follows; 0.75 mm collimation, slice thickness 1 mm, 100 kVp, 134 mA, and 0.625 mm collimation, slice thickness of 1 mm, 120 kVp, 165 mA, respectively. Images were acquired with 8 sequential axial slices; 4 slices on full inspiratory phase and other 4 slices on full expiratory phase to assess the degree of air trapping. The slices were performed at 4 levels: at the top of the aortic arch, at the carina, at the right inferior pulmonary vein, and at 2 cm above the right hemidiaphragm. Before scanning, participants were provided breathing instructions and practiced the breathing. No contrast medium was used in our protocol. However, patients may have received contrast medium for their primary study after the additional study CT chest was completed. CT scans were reconstructed using an algorithm with high spatial resolution.
Standard lung window settings were routinely used for the evaluation (level –500 Hounsfield units (HU), width 1500). All images were rendered anonymously and researchers were blinded to the clinical data. Two researchers; including a thoracic radiologist (NP with 15 years’ experience of thoracic CT) and a resident training in diagnostic radiology (NL with 3 years’ experience of radiology), independently reviewed all sections. Where the two readers could not reach a consensus, they scored the sections together. The final consensus results were recorded. CT features recorded included the presence, extent, grading, and distribution of air trapping, bronchiectasis, bronchial wall thickening, and reticulation. Other significant features were recorded as remarks.
The degree and lobar involvement of air trapping were defined from consensus of visual assessment and measurement of the mean lung attenuation. In normal lungs, the mean increase in lung attenuation at expiration should be approximately at 110 HU [15]. Measurement of mean lung attenuation was performed by freehand drawing regions of interest (ROIs) at a work station (Fuji PACS, Japan) in selected areas, which were the areas of most correspondence between inspiration and expiration. The mean lung attenuation value during inspiration was subtracted from that during expiration at each level. If visual assessment was consistent with air trapping and mean lung attenuation was increasing less than 110 HU, then the area was recorded as air trapping. The ROIs excluded the chest wall and large hilar vessels. Grading of air trapping was defined as: lobular, composed of small areas of hypoattenuation that corresponded to 1 or 2 adjacent secondary pulmonary lobules in 1 or 2 regions per lung level, 3 or more areas of lobular air trapping observed alternating with areas of normally attenuating lung, usually in a multilobular distribution, extensive with contiguous areas of air trapping in more than 3 adjacent pulmonary lobules, and sub-segmentally or lobar in distribution.
Reticulation within basal segments or other segments of lung was recorded. A reticular pattern adjacent to thoracic vertebral osteophytes was not included. Bronchiectasis was recorded and defined as bronchial dilatation with internal diameter of bronchus greater than the adjacent pulmonary artery [17], lack of tapering of bronchi, and identification of bronchi within 1 cm of the pleural surface [19]. Degree of bronchiectasis was defined as cylindrical, varicosities or cystic. Lobar involvement was also noted.
For the last important CT feature collected during this study, the presence or absence of bronchial wall thickening with involvement of lobes was then determined by visual analysis. For ambiguous cases, measurements were performed using a work station (Fuji PACS, Japan) at 5× magnification and measurement using electronic calipers, with wall thickness derived from these measurements (T = (D– L)/2) (
Statistical analyses were conducted using IBM SPSS Statistics for Windows, version 20 (IBM corporation, Armonk, NY, USA) and program R version 3.1.1 (GNU General Public License, USA). The three groups were compared using a Pearson chi-square test. A ROC curve was used to determine new age groups according to zero count in some of the characteristics of the original age groups. The comparison of variables of the new age groups were compared using a Fisher exact test and binomial proportion test.
Demographic data are shown in
Demographic data
Group 1 | Group 2 | Group 3 | Pearson chi-square test | |
---|---|---|---|---|
Range (age) | 20-40 | 41-60 | ≥6 1 | |
Median | 32 | 51 | 64.5 | |
Mean (Standard deviation) | 30.5 (6.1) | 50.4 (4.9) | 66.7 (5.7) | |
Number of participants | 20 | 20 | 20 | |
Male (%) | 17 (85) | 17 (85) | 8 (40) | 0.01 |
<0.05 | <0.05 | 0.50 | ||
Smoking | 4 | 10 | 5 | 0.09 |
Smoking history (Pack-years) Data are medians, with ranges in parentheses | 0 (0-2.5) | 0.05 (0-40) | 0 (0-60) |
Pearson chi-square statistics revealed an association between age group and sex (
Correlation between each outcome and age group
CT scan outcome | Age group 1 | Age group 2 | Age group 3 | Pearson chi-square test |
---|---|---|---|---|
Air trapping present | 5 | 7 | 18 | <0.01 |
Air trapping absent | 15 | 13 | 2 | |
Reticulation present | 0 | 1 | 0 | 0.36 |
Reticulation absent | 20 | 19 | 20 | |
Bronchiectasis present | 0 | 4 | 6 | 0.03 |
Bronchiectasis absent | 20 | 16 | 14 | |
Bronchial wall thickening present | 0 | 3 | 6 | 0.03 |
Bronchial wall thickening absent | 20 | 17 | 14 |
Because there were no bronchiectasis or bronchial wall thickening found in the youngest age group, a ROC curve was used to re-categorize participants into two age categories (<56 and >56 years old) to reduce the error from the Pearson chi-square test. Sensitivity and specificity of the coordinate of the ROC curve for bronchiectasis (
Sensitivity and specificity of the coordinate of the ROC curve at age 56
Bronchiectasis | Bronchial wall thickening | |
---|---|---|
Sensitivity | 70% | 78% |
Specificity | 68% | 69% |
Fisher’s exact test was used to determine the differences in bronchiectasis and bronchial wall thickening between the two groups (at ≤56 years and age >56 years) because the count in first group was less than 5 (
Correlation between each outcome and age group
Parameters | Age ≤56 | Age >56 | |
---|---|---|---|
Number | 37 | 23 | 0.09 |
Bronchiectasis | 3 | 7 | 0.04 |
Bronchial wall thickening | 2 | 7 | 0.02 |
A Pearson chi-square statistic showed no relationship between smoking status and the presence of air trapping, reticulation, bronchiectasis and bronchial wall thickening in CT images of any participant (
Correlation between each outcome and smoking status in all participants
CT imaging outcome | Smoking | No smoking | |
---|---|---|---|
Air trapping present | 10 | 20 | >0.10 |
Air trapping absent | 9 | 21 | |
Reticulation present | 1 | 0 | >0.10 |
Reticulation absent | 18 | 41 | |
Bronchiectasis present | 3 | 7 | >0.10 |
Bronchiectasis absent | 16 | 34 | |
Bronchial wall thickening present | 3 | 6 | >0.10 |
Bronchial wall thickening absent | 16 | 35 |
a Binomial proportion test,
b Fisher’s exact test
Distribution of air trapping, bronchiectasis and bronchial wall thickening in each group
Parameter | Lobe | Age group 1 | Age group 2 | Age group 3 | Total number in each lobe (%) |
---|---|---|---|---|---|
Number of air trapping | RUL | 0 | 1 | 5 | 6 (8) |
RML | 0 | 1 | 1 | 2 (3) | |
RLL | 5 | 6 | 17 | 28 (39) | |
LUL | 0 | 1 | 5 | 6 (8) | |
Lingular | 0 | 1 | 1 | 2 (3) | |
LLL | 4 | 7 | 17 | 28 (39) | |
Bronchiectasis | RUL | 0 | 1 | 2 | 3 (13) |
RML | 0 | 4 | 1 | 5 (22) | |
RLL | 0 | 2 | 3 | 5 (22) | |
LUL | 0 | 1 | 0 | 1 (4) | |
Lingular | 0 | 1 | 0 | 1 (4) | |
LLL | 0 | 2 | 4 | 6 (26) | |
Bronchial wall thickening | RUL | 0 | 2 | 2 | 4 (15) |
RML | 0 | 2 | 2 | 4 (15) | |
RLL | 0 | 2 | 5 | 7 (26) | |
LUL | 0 | 2 | 2 | 4 (15) | |
Lingular | 0 | 0 | 0 | 0 (0) | |
LLL | 0 | 2 | 6 | 8 (30) |
RUL = right upper lobe, RML = right middle lobe, RLL = right lower lobe, Lingular = lingular segment of LUL, LUL = the rest of left upper lobe, LLL = left lower lobe
Comparison of air trapping in lower lobes and other lobes in all participants
Air trapping | Other lobes | LLL+ RLL | Pearson chi-square test, LLL = left lower lobe, RLL = right lower lobe, |
---|---|---|---|
Positive | 16 | 56 | <0.01 |
Negative | 224 | 64 |
Distribution of bronchiectasis in total participants
RUL | RML | RLL | LUL | Lingular | LLL | |
---|---|---|---|---|---|---|
Positive | 3 | 5 | 5 | 1 | 1 | 6 |
Negative | 57 | 55 | 55 | 59 | 59 | 54 |
Distribution of bronchial wall thickening in total participants
RUL | RML | RLL | LUL | Lingular | LLL | |
---|---|---|---|---|---|---|
Positive | 4 | 4 | 7 | 4 | 0 | 8 |
Negative | 56 | 56 | 53 | 56 | 60 | 52 |
Degree of air trapping in each age group
Air trapping | Age group 1 | Age group 2 | Age group 3 | Total | Pearson chi-square test |
---|---|---|---|---|---|
Lobular | 3 | 4 | 5 | 12 | 0.11 |
Geographic | 2 | 2 | 8 | 12 | 0.09 |
Extensive | 0 | 1 | 5 | 6 | 0.02 |
Degree of bronchiectasis in each age group
Degree | Age group 1 | Age group 2 | Age group 3 | Total |
---|---|---|---|---|
Cylindrical | 0 | 4 | 6 | 10 |
Varicose | 0 | 0 | 0 | 0 |
Cystic | 0 | 0 | 0 | 0 |
We found that asymptomatic elderly participants had high frequency of air trapping, bronchiectasis, and bronchial wall thickening, which were independent of smoking history. These features are clinically important because of their potential as lung disease markers. Using these parameters in a clinical setting may prevent unnecessary investigation and follow up. So the results may reduce the potential confusion with disease, which results in unnecessary follow up investigations or treatment, or both, with potential expense and possible complications.
Air trapping is a term to describe retention of gas in all or part of a lung in the expiratory phase [4]. In one CT study, it was defined as an approximately 111.9 ± 46.3 (mean ± standard deviation) HU increase in mean lung attenuation in all three levels [5]. We showed a higher prevalence of air trapping in the older patients than younger patients (
We were not able to evaluate the cause of the air trapping in the present study because the relationship with histopathological findings could not be examined. However, there hypotheses for the cause of air trapping in lungs of elderly people. Progressive decline in respiratory and cardiac function causes alteration of alveolar dead space and shunts, possibly contributing to air trapping [4].
We found significant relationship between bronchiectasis and age (
Bronchial wall thickening can be defined as a T/D ratio more than 0.2 [17]. In our study, the prevalence of bronchial wall thickening was more frequently present in older patients than younger patients (
Reticulation showed no significant difference between the age groups in our study (
The strength s of our study were that we evaluated and correlated several CT features in one study. Furthermore, we performed a subgroup analysis of the basis of smoking status, and extent of CT features as described.
There are several limitations to our study. First, the sample size of this study was small. This could increase the rate of type 2 error (false negative). To improve this study, more participants should be included by either extending the length of the study or initiating a multicenter trial. Next, the group did not consist only of nonsmoker, but also of active smokers and ex-smokers. Nevertheless, as described we found that smoking status was not a predictor of outcome; so our study might be representative of the general Thai population. Another limitation was an unequal proportion of male and female patients, which was significantly different in the age groups 1 and 2 (
Because participants were prospectively recruited and consisted in part of young patients, our protocol was designed to minimize radiation dose. This limited us to using only 8 sequential slices with fixed 100–120 kVp protocol. The use of noncontiguous slices limited evaluation. We did not fully investigate extension of bronchiectasis. However, our results were nevertheless significant. Nor did our protocol allow acquisition of additional prone scans, limiting identification of atelectasis and a subpleural line was possible. Actually, reticulation was found in only one patient, so a prone scan might have been redundant.
There was unavoidable selection bias because we only recruited patient participants without respiratory symptoms; therefore, it might not truly be representative of the entire Thai population.
Asymptomatic older patient participants had greater prevalence of air trapping, bronchiectasis, and bronchial wall thickening than the younger patient participants in our study population. All of the described CT features were independent of smoking history. Air trapping was most common in both lower lobes. Extensive degree of air trapping also correlated with aging.