Table 2– Concentration-response functions (CRFs) used in the evaluation in relation to coronary heart disease and related acute morbidity

Outcome	Unit of CRF	Location, age of study participants	Definition of outcome and/or frequency outcome in study population	CRF (95% CI)^#	Study
Prevalence of underlying chronic diseases
CHD prevalence	Busy road ≤150 m from home (defined as autobahn and federal highways: 10-110×10³ vehicles per day) Prevalence exposure 31.4%	North-Rhine-Westphalia, Germany, 45–74 years	7.1%, prevalence, defined as self-reported history of MI or application of a coronary stent or angioplasty or bypass surgery	1.85 (1.21–2.84)	[11]
Exacerbations
Non-fatal MI	PM_2.5 per 10 μg·m⁻³	Ausburg 1999–2001, >18 years	WHO algorithm (includes evaluation chest pain, Q waves examination, levels creatine kinase and other)	1.105 (0.987–1.226)	[25]
			Transformed to PM₁₀ per 1 μg·m⁻³	1.005 (0.999–1.012)
Hospital admission for MI	NO₂ per 10 μg·m⁻³	Rome 1995–1997, >18 years	ICD9-410	1.026 (1.002–1.052)	[26]
	NO₂ per 8 μg·m⁻³	5 European studies, >35 years	ICD9: 410	1.005 (1.00–1.01)	[27]
	NO₂ per 57.3 μg·m⁻³	London 1987–1994, all ages	ICD9:410	1.0274 (1.0084–1.0479)	[28]
			Meta-analytic estimate, NO₂ per 1 μg·m⁻³	1.0005(1.0002–1.0008)
Hospital admission for stroke	PM₁₀ per 12 μg·m⁻³	Helsinki 1998–2004, ≥65 years	ICD10: I60-61 and I63–I64	0.9922 (0.069–1.016)	[29]
	PM₁₀ per 10 μg·m⁻³	9 French cities, ≥65 years	ICD10: I60–I64; G45-G46	1.008 (0.991-1.025)	[30]
	PM₁₀ per 10 μg·m⁻³	APHEA, 8 European cities, all ages	ICD9: 430–8	1.000 (0.997–1.003)	[31]
	PM₁₀ per 25 μg·m⁻³	Southwestern Sweden, median age 70 years	Ischaemic stroke	1.13 (1.04–1.22)	[32]
			Meta-analytic estimate, PM₁₀ per 1 μg·m⁻³	1.0007 (0.9993–1.0020)

CHD: coronary heart disease; MI: myocardial infarction; PM_2.5: particulate matter up to 2.5 μm; PM₁₀: particulate matter up to 10 μm; WHO: World Health Organization; ICD: international classification of diseases; APHEA: Air pollution and health, a European approach. ^#: when estimates were reported as odds ratios, odds ratios were corrected with the formula CRF/(1+It(CRF-1)), where It is estimated as the frequency of the outcome in the population [24].