Open Access

Short-term effects of air pollution on hospital admissions for cardiovascular diseases and diabetes mellitus in Sofia, Bulgaria (2009–2018)

Angel M. Dzhambov
Angel M. Dzhambov
, 
Krasimira Dikova
Krasimira Dikova
, 
Tzveta Georgieva
Tzveta Georgieva
, 
Teodor I. Panev
Teodor I. Panev
, 
Plamen Mukhtarov
Plamen Mukhtarov
 and 
Reneta Dimitrova
Reneta Dimitrova
   | Apr 04, 2023
Archives of Industrial Hygiene and Toxicology's Cover Image
About this article
Previous Article
Next Article
Cite
Article Category: Original article
Published Online: Apr 04, 2023
Page range: 48 - 60
Received: Jan 01, 2023
Accepted: Mar 01, 2023
DOI: https://doi.org/10.2478/aiht-2023-74-3704
© 2023 Angel M. Dzhambov et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Figure 1

Daily hospital admission counts for ischaemic heart disease (IHD), cerebral infarction (CI), and type 2 diabetes mellitus (T2DM) in Sofia, Bulgaria from 2009 to 2018
Daily hospital admission counts for ischaemic heart disease (IHD), cerebral infarction (CI), and type 2 diabetes mellitus (T2DM) in Sofia, Bulgaria from 2009 to 2018

Figure 2

Daily concentrations of air pollutants in Sofia, Bulgaria from 2009 to 2018. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Daily concentrations of air pollutants in Sofia, Bulgaria from 2009 to 2018. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Figure 3

Risk of hospital admissions for ischaemic heart disease associated with air pollution levels over seven days. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence intervals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Risk of hospital admissions for ischaemic heart disease associated with air pollution levels over seven days. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence intervals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Figure 4

Risk of hospital admissions for cerebral infarction associated with air pollution levels over seven days. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence intervals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Risk of hospital admissions for cerebral infarction associated with air pollution levels over seven days. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence intervals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Figure 5

Risk of hospital admissions for type 2 diabetes mellitus associated with air pollution levels over seven days. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Risk of hospital admissions for type 2 diabetes mellitus associated with air pollution levels over seven days. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Figure 6

Risk of hospital admissions for ischemic heart disease associated with air pollution levels over seven days stratified by gender and age. Legend: circles – male <65 yrs.; squares – female <65 yrs.; triangles – male ≥65 yrs.; diamonds – female ≥65 yrs. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Risk of hospital admissions for ischemic heart disease associated with air pollution levels over seven days stratified by gender and age. Legend: circles – male <65 yrs.; squares – female <65 yrs.; triangles – male ≥65 yrs.; diamonds – female ≥65 yrs. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Figure 7

Risk of hospital admissions for cerebral infarction associated with air pollution levels over seven days stratified by gender and age. Legend: circles – male <65 yrs.; squares – female <65 yrs.; triangles – male ≥65 yrs.; diamonds – female ≥65 yrs. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Risk of hospital admissions for cerebral infarction associated with air pollution levels over seven days stratified by gender and age. Legend: circles – male <65 yrs.; squares – female <65 yrs.; triangles – male ≥65 yrs.; diamonds – female ≥65 yrs. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Figure 8

Risk of hospital admissions for type 2 diabetes mellitus associated with air pollution levels over seven days stratified by gender and age. Legend: circles – male <65 yrs.; squares – female <65 yrs.; triangles – male ≥65 yrs.; diamonds – female ≥65 yrs. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide
Risk of hospital admissions for type 2 diabetes mellitus associated with air pollution levels over seven days stratified by gender and age. Legend: circles – male <65 yrs.; squares – female <65 yrs.; triangles – male ≥65 yrs.; diamonds – female ≥65 yrs. Lag models for the effects of pollutant concentrations include mutually adjusted 0–7 day lags, averaged over 3 and 7 days (lags 0–3 and 0–7), and the cumulative effect of lags 0–7 (net sum of lagged effects). All models are adjusted for time trend, day of the week, deciles of temperature, and relative humidity. Pollutants are tested one-at-a-time (single-pollutant models). Coefficients shown are incidence rate ratios (IRR) with 95 % confidence inter vals, where intervals not crossing the horizontal reference line indicate statistically significant estimates. CO – carbon monoxide; NO2 – nitrogen dioxide; O3 – ozone; PM10 – particulate matter ≤10 µm; PM2.5 – particulate matter ≤2.5 µm; SO2 – sulphur dioxide

Risk of hospital admissions associated with air pollution on the same day (lag0), stratified by time of year

Pollutant April — September October — March
IHD(IRR) Cl(IRR) T2DM(IRR) IHD(IRR) Cl(IRR) T2DM(IRR)
PM10 1.020 (1.000, 1.041)* 1.004 (0.984, 1.024) 0.999 (0.970, 1.029) 0.995 (0.991, 0.999)* 0.997 (0.995, 1.000) 0.994 (0.990, 0.997)*
PM2.5 1.010 (0.983, 1.038) 1.011 (0.981, 1.041) 1.026 (0.980, 1.073) 0.992 (0.986, 0.999)* 0.997 (0.993, 1.001) 0.993 (0.987, 0.999)*
SO2 0.979 (0.909, 1.055) 1.056 (0.990, 1.126) 1.001 (0.903, 1.111) 0.988 (0.969, 1.008) 0.982 (0.960, 1.004) 0.964 (0.934, 0.996)*
NO2 1.038 (1.018,1.057)* 1.022 (1.006,1.041)* 1.026 (1.000,1.052)* 0.996 (0.988, 1.003) 0.996 (0.990, 1.003) 0.997 (0.988, 1.006)
O3 0.984 (0.970, 0.997)* 0.994 (0.982, 1.007) 0.987 (0.968, 1.007) 0.995 (0.983, 1.006) 1.003 (0.991, 1.015) 0.993 (0.976, 1.011)
CO 1.071 (0.978, 1.175) 0.978 (0.887, 1.078) 1.010 (0.874, 1.169) 0.964 (0.937, 0.992)* 0.978 (0.961, 0.996)* 0.969 (0.943, 0.996)*

Risk of hospital admissions associated with above-threshold air pollution levels on the same day (lag0)

Pollutant IHD(IRR) CI(IRR) T2DM(IRR)
PM10 ≥45 μg/m3 1.007 (0.980, 1.034) 1.000 (0.973, 1.028) 0.991 (0.950, 1.034)
PM 5 ≥15 μg/m3 1.005 (0.982, 1.029) 1.004 (0.979, 1.030) 0.997 (0.959, 1.036)
SO ≥40 μg/m3 0.952 (0.858, 1.056) 0.915 (0.796, 1.053) 1.003 (0.827, 1.216)
NO ≥25 μg/m3 1.039 (1.013, 1.066)* 1.019 (0.993, 1.045) 1.029 (0.991, 1.069)
O ≥60 μg/m3 0.971 (0.935, 1.008) 0.976 (0.941, 1.012) 0.944 (0.893, 0.997)*
CO ≥4 mg/m3 0.929 (0.810, 1.064) 0.995 (0.912, 1.085) 0.875 (0.771, 0.994)*

Descriptive statistics for the exposure variables in the study

Variable Missing data(N, %) Percentiles Min Max
25th 50th 75th
PM10 45 (1.23) 22.32 31.29 45.38 3.60 601.04
PM2.5 204 (5.59) 12.22 17.76 27.10 0.49 485.77
NO2 23 (0.63) 21.40 30.03 41.70 0.01 184.89
SO2 25 (0.68) 4.62 6.72 10.68 0.01 123.31
O3 26 (0.71) 22.82 38.47 52.28 0.39 97.27
CO 45 (1.23) 0.42 0.61 0.90 0.00 7.83
Temperature 20 (0.55) 4.70 12.44 19.19 -14.39 31.375
Relative humidity 20 (0.55) 56.29 65.80 76.21 30.50 98.26

Risk of hospital admissions associated with air pollution levels on the same day (lag0)

Pollutant IHD(IRR) CI(IRR) T2DM(IRR)
PM10 0.996 (0.992, 1.000) 0.998 (0.995, 1.001) 0.995 (0.991, 0.999)*
PM2.5 0.995 (0.989, 1.001) 0.998 (0.994, 1.003) 0.993 (0.987, 1.000)
SO2 0.990 (0.967, 1.013) 0.985 (0.960, 1.011) 0.991 (0.957, 1.003)
NO2 1.003 (0.996, 1.010) 0.999 (0.993, 1.005) 1.002 (0.993, 1.011)
O3 0.991 (0.982, 1.001) 0.995 (0.985, 1.004) 0.992 (0.978, 1.006)
CO 0.974 (0.948, 1.001) 0.982 (0.962, 1.002) 0.974 (0.947, 1.003)

Risk of hospital admissions associated with deciles of air pollutant levels on the same day (lag0)

Deciles PM10 PM2.5 SO2 NO2 O3 CO
Ischaemic heart disease (IRR)
D2 0.993 (0.948, 1.041) 1.007 (0.959, 1.058) 1.005 (0.959, 1.054) 1.019 (0.970, 1.070) 1.022 (0.975, 1.072) 0.996 (0.953, 1.041)
D3 0.976 (0.934, 1.019) 1.021 (0.977, 1.067) 1.001 (0.952, 1.053) 0.995 (0.945, 1.048) 1.057 (1.004, 1.114)* 0.997 (0.949, 1.047)
D4 0.976 (0.933, 1.022) 0.989 (0.945, 1.036) 1.014 (0.962, 1.068) 1.031 (0.980, 1.086) 1.022 (0.968, 1.078) 0.981 (0.929, 1.035)
D5 0.953 (0.908, 1.000) 1.004 (0.959, 1.050) 1.012 (0.962, 1.065) 1.027 (0.977, 1.080) 1.02 (0.965, 1.078) 0.998 (0.944, 1.054)
D6 1.004 (0.958, 1.053) 1.015 (0.966, 1.067) 0.990 (0.937, 1.045) 1.046 (0.994, 1.100) 1.033 (0.977, 1.092) 1.009 (0.954, 1.067)
D7 0.983 (0.941, 1.027) 1.000 (0.954, 1.048) 1.011 (0.957, 1.069) 1.078 (1.025,1.134)* 0.978 (0.923, 1.037) 1.012 (0.960, 1.067)
D8 1.008 (0.965, 1.053) 0.998 (0.956, 1.042) 0.997 (0.943, 1.053) 1.055 (1.003,1.109)* 0.982 (0.924, 1.043) 1.007 (0.951, 1.067)
D9 1.022 (0.977, 1.070) 1.011 (0.965, 1.060) 0.999 (0.940, 1.061) 1.067 (1.012,1.125)* 0.981 (0.921, 1.045) 1.048 (0.994, 1.106)
D10 0.944 (0.895, 0.997)* 1.006 (0.953, 1.062) 0.962 (0.895, 1.035) 1.039 (0.983, 1.098) 0.960 (0.897, 1.027) 0.981 (0.922, 1.044)
Cerebral infarction (IRR)
D2 0.973 (0.929, 1.020) 1.023 (0.977, 1.072) 0.968 (0.915, 1.024) 1.032 (0.981, 1.086) 1.027 (0.980, 1.076) 0.957 (0.913, 1.003)
D3 1.016 (0.967, 1.068) 0.992 (0.945, 1.042) 0.977 (0.924, 1.033) 1.033 (0.984, 1.085) 1.035 (0.984, 1.089) 0.957 (0.908, 1.009)
D4 0.988 (0.939, 1.040) 1.034 (0.984, 1.087) 0.942 (0.889, 0.998)* 1.042 (0.991, 1.097) 1.058 (1.005,1.114)* 0.965 (0.913, 1.020)
D5 1.004 (0.954, 1.057) 1.038 (0.987, 1.092) 0.971 (0.915, 1.030) 1.060 (1.008,1.115)* 1.023 (0.970, 1.079) 0.957 (0.905, 1.011)
D6 0.975 (0.927, 1.027) 1.018 (0.966, 1.073) 0.960 (0.903, 1.020) 1.038 (0.988, 1.090) 0.986 (0.931, 1.045) 0.976 (0.922, 1.034)
D7 0.994 (0.943, 1.046) 1.022 (0.967, 1.080) 0.973 (0.914, 1.036) 1.053 (1.001,1.107)* 1.010 (0.952, 1.073) 0.995 (0.936, 1.057)
D8 1.004 (0.953, 1.058) 1.023 (0.971, 1.079) 0.982 (0.921, 1.047) 1.050 (0.995, 1.108) 1.011 (0.950, 1.075) 0.963 (0.905, 1.024)
D9 0.983 (0.931, 1.038) 0.981 (0.926, 1.039) 0.975 (0.913, 1.040) 1.072 (1.018,1.128)* 0.987 (0.927, 1.052) 0.961 (0.903, 1.023)
D10 0.965 (0.913, 1.020) 1.000 (0.942, 1.060) 0.936 (0.869, 1.009) 0.997 (0.946, 1.052) 0.995 (0.929, 1.067) 0.929 (0.872, 0.989)*
Type 2 diabetes mellitus (IRR)
D2 0.986 (0.918, 1.058) 1.019 (0.949, 1.095) 0.968 (0.889, 1.055) 1.050 (0.971, 1.135) 0.992 (0.924, 1.065) 0.986 (0.914, 1.063)
D3 0.933 (0.868, 1.002) 0.984 (0.914, 1.059) 0.958 (0.876, 1.048) 1.009 (0.934, 1.090) 1.010 (0.938, 1.087) 0.999 (0.921, 1.083)
D4 0.930 (0.865, 1.000) 0.951 (0.883, 1.025) 0.991 (0.905, 1.086) 1.003 (0.928, 1.085) 0.985 (0.913, 1.063) 0.947 (0.870, 1.031)
D5 0.925 (0.858, 0.997)* 0.988 (0.916, 1.065) 0.971 (0.887, 1.063) 1.058 (0.979, 1.144) 0.965 (0.890, 1.045) 0.944 (0.865, 1.030)
D6 0.987 (0.916, 1.063) 0.988 (0.916, 1.066) 0.941 (0.858, 1.033) 1.006 (0.930, 1.087) 0.985 (0.907, 1.071) 0.959 (0.877, 1.048)
D7 0.952 (0.884, 1.025) 0.978 (0.905, 1.056) 0.977 (0.888, 1.074) 1.094 (1.012,1.182)* 0.990 (0.907, 1.080) 0.960 (0.877, 1.051)
D8 0.961 (0.891, 1.036) 0.999 (0.924, 1.080) 0.956 (0.867, 1.053) 1.023 (0.944, 1.108) 0.997 (0.910, 1.092) 0.977 (0.890, 1.071)
D9 0.963 (0.891, 1.041) 0.986 (0.909, 1.068) 1.010 (0.914, 1.116) 1.042 (0.961, 1.130) 0.982 (0.893, 1.080) 0.985 (0.898, 1.081)
D10 0.941 (0.868, 1.020) 0.975 (0.895, 1.063) 0.919 (0.823, 1.026) 1.059 (0.974, 1.151) 0.926 (0.837, 1.025) 0.971 (0.882, 1.068)

Spearman’s correlations between the key variables in the study

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
1. IHD 1.00
2. CI 0.52* 1.00
3. T2DM 0.83* 0.50* 1.00
4. PM10 0.05* 0.02 0.02 1.00
5. PM25 0.03 -0.01 0.01 0.88* 1.00
6. SO2 0.07* -0.01 0.04* 0.50* 0.55* 1.00
7. NO2 0.15* 0.08* 0.13* 0.77* 0.69* 0.42* 1.00
8. O3 -0.15* -0.07* -0.07* -0.48* -0.43* -0.27* -0.62* 1.00
9. CO 0.09* -0.01 0.07* 0.66* 0.69* 0.55* 0.66* -0.52* 1.00
10. Temperature -0.12* -0.02 -0.03 -0.20* -0.34* -0.53* -0.21* 0.49* -0.42* 1.000
11. Relative humidity 0.09* -0.02 0.03 0.08* 0.15* 0.12* 0.07* -0.50* 0.30* -0.53* 1.00
eISSN:
1848-6312
Languages:
English, Slovenian
Publication timeframe:
4 times per year
Journal Subjects:
Medicine, Basic Medical Science, other
Journal RSS Feed