Association Between Ambient Air Pollution and Elevated Risk of Tuberculosis Development

Figures & data

Figure 1 Schematic representation of the probabilistic risk assessment framework used in this study: (A) problem formulation, (B) exposure/effect analyses, and (C) risk characterization (see text for the symbolic meaning).

Figure 2 Probability distributions of concentrations of air pollutants for (A) NO₂, (B) NO_X, (C) CO, (D) PM_2.5, and (E) PM₁₀ in Taipei City, Hualien, Taitung, and Pingtung Counties, respectively.

Table 1 Fitted Coefficients of Three-Parameter Hill Models for Describing the Relationships Between Air Pollutant Concentrations and Population Attributable Fraction (PAF) of Incident TB

Scenario	Air Pollutant	Fitted Coefficients (mean ± se)			r^2
		PAFmax	AC50	n
Severe	NO2 (Figure 3A)	1	31.59 ± 1.15 ppb***	4.43 ± 0.80***	0.86***
	NOX (Figure 3B)	1	46.33 ± 1.88 ppb***	4.03 ± 0.73***	0.86***
	CO (Figure 3C)	1	1.12 ± 0.07 ppm***	2.70 ± 0.38***	0.89***
	PM2.5 (Figure 3D)	0.34 ± 0.01***	28.46 ± 0.11 μg/m^3***	26.72 ± 2.53***	0.99***
	PM10 (Figure 3E)	0.36 ± 0.04***	53.32 ± 1.39 μg/m^3 ***	12.82 ± 3.04**	0.96***
Moderate	NO2 (Figure 3F)	0.29 ± 0.11*	29.78 ± 4.18 ppb***	5.46 ± 1.40*	0.98***
	NOX (Figure 3G)	0.49 ± 0.09**	44.05 ± 2.68 ppb***	7.48 ± 1.49**	0.98***
	CO (Figure 3H)	0.21 ± 0.17	0.84 ± 0.25 ppm*	5.75 ± 3.40	0.91**

Notes: *p-value < 0.05, **p-value < 0.01, ***p-value < 0.001.

Figure 3 Population dose–response profiles describing the relationship between air pollutant concentrations and population attributable fraction (PAF) of incident TB for (A–E) NO₂, NO_X, CO, PM_2.5, and PM₁₀ under the severe scenario; and for (F–H) NO₂, NO_X, and CO under the moderate scenario.

Table 2 Estimated the Highest Population Attributable Fraction (PAF) of TB Associated with Air Pollution Exposure at Exceedance Risk Probabilities of 0.8 (Most Likely) and 0.5 (likely) for Taipei City, Hualien, Taitung, and Pingtung, Respectively

Scenario	Study Site	Exceedance Risk (ER)
		0.8			0.5
		Air Pollutant	PAF	Contribution to Incidence^b	Air Pollutant	PAF	Contribution to Incidence
Severe	Taipei	CO	12.2% (9.0–15.4%)^a	4.72 (2.80–6.96)	CO	16.6% (13.7–19.5%)	6.41 (4.07–9.05)
	Hualien	CO	2.7% (0.7–4.8%)	2.58 (0.60–4.96)	CO	5.2% (2.4–8.0%)	4.88 (2.01–8.27)
	Taitung	CO	2.6% (0.6–4.5%)	2.29 (0.49–4.41)	CO	3.8% (1.4–6.3%)	3.42 (1.14–6.09)
	Pingtung	CO	1.6% (0.1–3.1%)	1.46 (0.10–2.94)	PM10	11.9% (9.0–14.8%)	10.74 (7.26–14.75)
Moderate	Taipei	NO2	1.7% (0.1–3.2%)	0.64 (0.03–1.35)	NOX	3.3% (1.0–5.6%)	1.28 (0.34–2.36)
	Hualien	CO	0.1% (0.0–0.5%)	0.11 (0.0–0.46)	CO	0.2% (0.0–1.4%)	0.35 (0.0–1.37)
	Taitung	CO	0.05% (0.0–0.4%)	0.09 (0.0–0.39)	CO	0.1% (0.0–0.8%)	0.19 (0.0–0.78)
	Pingtung	CO	0.02% (0.0–0.18%)	0.04 (0.0–0.17)	CO	0.09% (0.0–0.7%)	0.15 (0.0–0.63)

Notes: ^aMedian (95% CI). ^bContribution of specific air pollutant i to TB incidence rate (per 100,000 population) = Φ_ER(PAF_i)×site-specific incidence rate. The site-specific incidence rates in the period 2006–2016 were 38.6 ± 6.8 (mean ± sd), 93.5 ± 17.1, 89.1 ± 15.3, and 90.3 ± 11.5 per 100,000 populations in Taipei City and Hualien, Taitung, and Pingtung Counties, respectively (Table S1).

Figure 4 Population-level exceedance risk curve of air pollutants-associated population attributable fraction (PAF) of incident TB under severe and moderate scenarios for (A, B) Taipei City, (C, D) Hualien County, (E, F) Taitung County, and (G, H) Pingtung County, respectively.