Pulmonary nontuberculous mycobacterial infection is associated with urinary molybdenum concentrations in a bronchiectasis cohort

Abstract

Background:

Epidemiologic studies have demonstrated an association between trace metals in source water and water distribution systems and nontuberculous mycobacterial (NTM) pulmonary infection and disease. Specifically, molybdenum has been strongly associated with Mycobacterium abscessus and Mycobacterium avium complex infection and disease in multiple epidemiologic studies. We aimed to determine if urinary molybdenum, chromium, and manganese concentrations were associated with pulmonary NTM infection in a bronchiectasis patient cohort.

Methods:

We enrolled 120 consecutive adult bronchiectasis patients participating in a bronchiectasis natural history study. We collected one urine sample per patient during routine visits and sent samples to the Centers for Disease Control and Prevention for trace metal testing. We defined three patient subgroups: (1) current NTM infection, (2) previous NTM infection, and (3) never NTM infection. Logistic regression models adjusted for age, racial/ethnic group, and cavitary disease status were used to test for differences between metal concentrations across infection statuses.

Results:

Forty-six (38%) patients were currently infected with NTM, 46 (38%) were previously infected, and 28 (23%) were never infected. For each increase of one standard deviation of the logarithm of molybdenum concentration, the odds of current versus never NTM infection increased significantly by 14% [odds ratio (OR) = 1.14, 95% confidence interval (CI) = 1.02, 1.28], while manganese and chromium did not exhibit significant associations.

Conclusions:

This clinical study is the first to examine the relationship between trace metals and pulmonary NTM infections. These findings support those from prior population-based studies.

What this study adds

This original research is the first to examine the association between urinary trace metal concentrations and nontuberculous mycobacterial (NTM) pulmonary infection within a bronchiectasis cohort. Exposure to trace metals occurs through drinking water and other environmental factors. This pilot study expands the scope and findings of prior ecologic epidemiologic studies to better understand how trace metals from environmental exposure affect the risk of NTM infection. We found that an increase of one natural logarithm in molybdenum was independently associated with a 14% increase in the odds of current NTM infection.

Introduction

The incidence of nontuberculous mycobacterial pulmonary disease (NTM PD) has been increasing in the United States and globally.^1–3 In multiple ecologic epidemiologic studies, trace metals, particularly molybdenum, in source water and public drinking water have been associated with NTM pulmonary infection in a variety of populations across the United States.^4–8 To build on findings from these population studies, we conducted a pilot clinical study in a cohort of bronchiectasis patients, who are at high risk of NTM infection and disease.⁹ These patients were enrolled in a clinical protocol at the National Institutes of Health (NIH).^1,10 Our aim was to investigate the relationship between environmental trace metal exposure, measured through patient urine samples, and the risk of NTM infection in susceptible individuals. We hypothesize that increasing urinary trace metal concentrations, namely, molybdenum, will increase the risk of having current or past NTM infection compared with patients who were never infected.

Methods

Study population

Our study population comprised 120 adult patients in the NIH Clinical Center’s Natural History of Bronchiectasis study (09-H-0172) (www.clinicaltrials.gov identifier NCT00943514). All participants provided written informed consent, and all methods were carried out in accordance with relevant ethics guidelines and regulations. For this pilot study, the sample size was determined by the availability of funding, which restricted patient enrollment. In 2024, consecutive patients on this protocol seen at the NIH Clinical Center for routine follow-up were recruited to provide urine samples for trace metal testing as described below.

Data abstraction

Demographic, clinical, microbiological, and radiographic data were abstracted from the NIH Clinical Center’s electronic medical records. Clinical variables selected for analysis included those previously identified as measures of disease severity, to assess whether these factors might confound the relationship between NTM infection and trace metal concentrations:^10,11 BMI, 6-minute walk test distance, saturation product,¹² forced expiratory volume in 1 second (FEV₁% predicted), and presence of cavitary disease. We used measurements obtained within 6 months of each patient’s sample collection date (±6 months).

Trace metal testing

We chose urinary samples for trace metal testing, as this sample type is considered to be a reliable measure of environmental exposure.^13,14 Urine samples were tested at the Centers for Disease Control and Prevention for molybdenum, manganese, and chromium by inductively coupled plasma mass spectrometry and corrected for creatinine concentration.^14–23

Nontuberculous mycobacterial infection status definition

A study subject with current NTM infection was defined as someone who had an NTM pulmonary isolate within 6 months (±6) of the urine sample collection. Past NTM infection was defined as someone without current infection, but with an NTM pulmonary isolate at least 6 months before urine sample collection. Never NTM infection was defined as a study subject with no documented positive NTM cultures. Isolates of M. gordonae and M. paragordonae were excluded from our analyses.

Data analysis

We first evaluated the relationship between urinary molybdenum, manganese, and chromium concentrations and NTM infection status, using global Kruskal-Wallis heterogeneity tests to assess statistical significance, with P < 0.20 as a screening criterion. Creatinine-corrected trace metal measurements were natural-logarithm transformed, centered, and scaled for analysis, as were corresponding continuous covariates. The most marked difference was observed for molybdenum when comparing never-infected with currently infected groups, therefore we focused the remainder of our analysis on molybdenum and NTM in these infection groups.

To evaluate confounding by demographic and clinical factors, we assessed the relationships between those factors and molybdenum concentrations, and between those covariates and NTM infection status, using heterogeneity and correlation tests. Using a reduced set of variables deemed to be potential confounders, we then constructed a logistic regression model with NTM infection status as the outcome variable. R (Version 4.3.0, R Foundation for Statistical Computing, Vienna, Austria) was used for analysis. We used a backward elimination model with the criterion of minimizing the Akaike information criterion to generate the final model.

Results

Overall, 76.7% of patients had a history of NTM pulmonary infection, with 50.8% having ever been infected with Mycobacterium avium-intracellulare complex and 27.5% of patients with Mycobacterium abscessus; demographic and clinical features were similar to those previously described for this cohort (Table 1).¹⁰ All patients with NTM isolated met American Thoracic Society disease criteria for NTM PD, with at least 2 sputum samples or 1 bronchoalveolar lavage. The median molybdenum concentration was marginally significantly higher among patients with current NTM infection (60.9 µg/g, n = 46), compared with those patients never infected with NTM (42.5 µg/g, n = 28) or those previously infected with NTM (47.5 µg/g, n = 46) (Figure 1; p_{current vs. never} = 0.09). Median concentrations of manganese and chromium were not significantly different among groups (Figure 1). In our final model adjusted for age, race/ethnicity, and cavitary disease, we found that an increase of one natural-logarithm in the concentration of creatinine-corrected molybdenum was independently associated with a 14% increase in the odds of current NTM infection [odds ratio (OR) = 1.14, 95% confidence interval (CI) = 1.02, 1.28) (Table 2).

Table 1.

Demographic, clinical, microbiological, and radiographic features of the bronchiectasis cohort

Characteristic	Current NTM infection (n = 46)	Past NTM infection (n = 46)	Never NTM infection (n = 28)	Overall (n = 120)
Age
Median (IQR)	67.5 (21.0)	64.2 (18.0)	57.5 (28.3)	66.0 (22.3)
Sex (%)
Female	33 (71.7)	34 (73.9)	22 (78.6)	89 (74.2)
BMI
Median (IQR)	20.8 (4.9)	22.1 (3.6)	22.5 (4.2)	22.0 (4.3)
FEV1% predicted
Median (IQR)	79.0 (35.3)	83.0 (33.0)	81.0 (36.8)	82.0 (32.0)
6MWT DSP
Median (IQR)	514 (149.3)	514 (163.2)	516 (144.4)	516 (152.9)
Race/ethnicity (%)
White	38 (86.4)	29 (64.4)	21 (80.8)	88 (76.5)
Asian	6 (13.6)	13 (28.9)	3 (11.5)	22 (19.1)
Black	0 (0.0)	3 (6.7)	2 (7.7)	5 (4.4)
Species (%)a
M. abscessus b	25 (26.3)	16 (19.0)	NA	41 (22.9)
MAC	56 (59.0)	55 (65.5)	NA	111 (62.0)
Other NTM species	14 (14.7)	13 (15.5)	NA	27 (15.1)
Cavitary disease (%)	4 (9.1)	6 (14.0)	1 (4.0)	11 (9.8)
Comorbid conditions (%)
CF	1 (2.2)	2 (4.3)	0 (0.0)	3 (2.5)
PCD	4 (8.7)	4 (8.7)	1 (3.6)	9 (7.5)
Asthma	6 (13.0)	6 (13.0)	7 (25.0)	19 (15.8)
Region of residence (%)
Midwestern	1 (2.2)	4 (8.7)	3 (10.7)	8 (6.7)
Northeastern	9 (19.6)	6 (13.0)	7 (25.0)	22 (18.3)
Southern	32 (69.6)	31 (67.4)	16 (57.1)	79 (65.8)
Western	4 (8.7)	5 (10.9)	2 (7.1)	11 (9.2)

Missing data: FEV₁% predicted: n = 7; 6-minute walk test distance saturation product: n = 5; presence of cavitary disease: n = 8; semiquantitative culture result: n = 88. Includes M. avium, M. intracellulare/chimaera, M. chimaera, M. intracellulare. Since many patients had isolates from multiple NTM species, species results are represented as the number and percentage of isolates, not participants.

^aExcludes M. gordonae and M. paragordonae.

^bIncludes M. abscessus, M. abscessus group, M. chelonae, M. chelonae group, M. massiliense.

MAC indicates Mycobacterium avium complex.

Figure 1.

Trace metal concentrations by NTM infection status. A, molybdenum concentrations by NTM infection status*. Microgram of molybdenum per gram of creatinine is abbreviated as μg/g. Median molybdenum concentrations are shown above the median line inside the boxplot corresponding to each NTM infection status group. Global Kruskal-Wallis P-value = 0.09; Tukey-Welsch post hoc tests: current versus never: P = 0.09, previous versus never: P = 0.37. ⁺Comparison is statistically significant at the α = 0.20 significance threshold level using the Kruskal-Wallis test and Tukey post hoc tests. Values are untransformed, creatinine-corrected concentrations. *Creatinine-corrected molybdenum results range from 4.78 to 652 μg/g. Values above 200 μg/g (n = 2 current NTM infection: 226.0 μg/g, 216.0 μg/g; and 3 past NTM infection: 652.0 μg/g, 627.0 μg/g, and 347.0 μg/g) were excluded from this figure to assist with visibility between boxplots by NTM infection status. B, manganese concentrations by NTM infection status**. Microgram of manganese per gram of creatinine is abbreviated as μg/g. Values below the limit of detection threshold: n = 46. Median manganese concentrations are shown above the median line inside the boxplot corresponding to each NTM infection status group. Global Kruskal-Wallis P-value = 0.58. Heterogeneity tests were not significant between groups. **Creatinine-corrected manganese results range from 0.029 to 6.78 μg/g. Values above 3.5 μg/g (n = 3 current NTM infection: 6.74 μg/g, 6.73 μg/g, 4.58 μg/g; 1 never NTM infection: 6.78 μg/g) were excluded from this figure to assist with visibility between boxplots by NTM infection status. C, chromium concentrations by NTM infection status***. Microgram of chromium per gram of creatinine, abbreviated as μg/g. Values below the limit of detection threshold: n = 52. Median chromium concentrations are shown above the median line inside the boxplot corresponding to each NTM infection status group. Global Kruskal-Wallis P-value = 0.55. Heterogeneity tests were not significant between groups. Values are untransformed, creatinine-corrected concentration measurements. ***Creatinine-corrected chromium results range from 0.043 to 23.98 μg/g. Values above 2 μg/g (n = 1 current NTM infection: 6.58 μg/g; 1 never NTM infection: 2.65 μg/g; 3 past NTM infection: 23.98 μg/g, 2.83 μg/g, and 2.72 μg/g) were excluded from this figure to assist with visibility between boxplots by NTM infection status.

Table 2.

Current versus never NTM infection logistic regression results for log-transformed molybdenum

Measure	Adjusted odds ratio (95% CI)a	P-value
Molybdenum	1.14 (1.02, 1.28)	0.02

^aOdds ratios are represented as the change in odds of being currently NTM infected versus never NTM infected for each increase of one logarithm unit of each measure. Model adjusted for: age, race/ethnicity, and cavitary disease.

Discussion

We found that higher urinary molybdenum concentrations were significantly and independently associated with an increased risk of current versus never NTM infection. Our study is the first clinical study to examine the relationship between trace metals and pulmonary NTM infections. These findings support results from multiple ecological epidemiologic studies that demonstrated a significant association between environmental molybdenum and pulmonary NTM infections and disease.^4–8 In particular, the findings of this study (OR = 1.14; 95% CI = 1.02, 1.28) show remarkable similarity to the findings of a recent population-based study (OR = 1.12; 95% CI = 1.01, 1.25), which examined trace metal concentrations in public drinking water and risk of NTM pulmonary infection among older adults in the United States.²⁴ The mechanism by which molybdenum may influence NTM infection remains unclear, but several hypotheses have been discussed in previous studies.^25–28

Our study has several limitations. First, this study was a small pilot study with limited statistical power, which precluded statistical analysis at the species level. Second, the source of molybdenum is unknown, as exposure can occur through air, water, foods, dietary supplements, or medications. Thus, although urinary trace metal concentrations provide an excellent measure of a patient’s entire exposome,²⁹ we cannot ascertain environmental exposure sources without a more detailed, questionnaire-based investigation. Given the relatively short biological half-life of molybdenum in the human body (estimated to be several weeks),³⁰ we used urinary trace metal concentrations as a measure of household exposure, and assumed that most exposure was primarily from household drinking water. Patients may also be exposed to molybdenum through molybdenum-rich diets, including legumes, cereal grains, leafy vegetables, beef liver, and milk,^31,32 and from dietary supplements. While we do not know if diet exposures vary across patient subgroups, future studies could ascertain dietary and water exposures in more detail and could also test home water sources. Finally, this study was cross-sectional, such that we could not disentangle cause and effect, specifically whether molybdenum exposure is a risk factor for NTM infection, or a result of such infection. Future longitudinal studies are needed to address this question. We hypothesize that home drinking water is the primary source of molybdenum exposure in these patients (Supplemental Table 1; https://links.lww.com/EE/A415).

In summary, we have demonstrated a significant association between urinary molybdenum concentrations and pulmonary NTM infection in a high-risk cohort. Future larger longitudinal studies are needed to fully evaluate these findings.

Conflicts of interest statement

The authors declare that they have no conflicts of interest with regard to the content of this report.

ACKNOWLEDGMENTS

We would like to gratefully acknowledge the nurses in the NIH Clinical Center Outpatient Clinic 7 for coordinating, leading, and facilitating clinical sample collection. Additionally, we would like to thank the patients who participated in this study. We would also like to acknowledge the CDC Sample Logistics Laboratory within the Inorganic and Radiation Analytical Toxicology Branch for conducting the trace metal testing, especially Jeff Jarrett, Joaudimir Castro Georgi, Jessica Wright, and Cynthia Ward.