Skip to main content
. 2021 May 10;13(5):1589. doi: 10.3390/nu13051589

Table 4.

COVID-19 Severity defined by without study-specific definitions.

Author [Ref] Micronutrient in Question Reported Summary Risk Estimate: Odds Ratio (OR) (95% Confidence Interval)/Mean (SD); % Population Infected (Infected/Population Size) Key Findings
Supplementation Deficiency
Supplemented Non-Supplemented Quintile 1 Quintile 2 Quintile 3
Outcome: Death due to COVID-19 episode/during hospitalisation due to COVID-19 episode
Annweiler [18] Vitamin D3 0.11 (0.03–0.48) *,a; 17.5% (10/57) Reference; 55.6% (5/9) *

  • 52% to 97% a lowered risk of death in those supplemented with bolus dose, compared to those not supplemented.

  • Significantly lower proportion of deceased participants (66.7%) receiving vitamin D3 bolus dose during or just before COVID-19 compared to survivors (92.2%, p = 0.023).
Carpagnano [22] Vitamin D NR; 3.1% (1/32) # NR; 20% (2/10) #

  • 50% probability of dying in patients with severe vitamin D deficiency after 10 days, compared to those without severe vitamin D deficiency (log-rank test, p = 0.019)

  • 5% mortality risk in patients without severe Vitamin D deficiency.
Castillo [47] Vitamin D NR; 0% (0/50) # NR; 7.7% (2/26) #

  • Lower proportion of deaths amongst patients treated with calcifediol, compared to those not treated with calcifediol (p-value unreported)
Hastie [17] Vitamin D Reference 1.21 (0.83–1.76) b 1.02 (0.75–1.38) b

  • 2% to 21% b higher risk of death in patients with insufficient or deficient Vitamin D levels, compared to those with sufficient Vitamin D levels.

  • 2% decrease in risk of death per 10 nmol/L increase in Vitamin D levels [Hazard ratio: 0.98 (0.91–1.06) b]
Karahan [27] Vitamin D3 NR; 0% (0/12) * NR; 14.7% (5/34) * NR; 62.1% (64/103) *

  • Significantly lower mean ± SD serum 25(OH) vitamin D level among deceased patients (10.4 ± 6.4 ng/mL) compared to surviving patients (19.3 ± 11.2 ng/mL, p < 0.001).

  • Significantly higher mortality rate among patients with severe-critical COVID-19 (66.7%) compared with moderate COVID-19 patients (2.1%, p < 0.001)
Mardani [31] Vitamin D

  • Significantly lower mean vitamin D concentration in deceased positive patients (8.175 ng/mL) compared to those surviving positive (19.25 ng/mL) and negative patients (30.17 ng/mL, p < 0.0001)
Panagiotou [35] Vitamin D Reference; NR 0.97 (0.42, 2.23); NR

  • No significant difference in mortality rates between deficient and normal groups (p > 0.05)
Radujkovic [37] Vitamin D ᶧ Reference, NR For all subjects: 14.73 (4.16–52.19) c; NR
For inpatients only: 11.51 (3.24–40.92) c; NR

  • 10.5 c and 13.7 c times higher risk of death in the Vitamin D deficient group, compared to those with not deficient, amongst the inpatients and all patients respectively (p < 0.0001)
Vitamin D ᶲ Reference, NR For all subjects: 11.27 (1.48–85.55) c; NR
For inpatients only: 7.97 (1.05–60.60) c; NR

  • Approximately 7 and 11 c times higher risk of death in the Vitamin D insufficient group, compared to those with not insufficient, amongst the inpatients (p = 0.004) and all patients (p = 0.04)
Tan [49] Vitamin D, Magnesium & Vitamin B12 NR; 0% (0/17) # NR; 0% (0/26) #

  • No effect of micronutrient supplementation during hospitalisation on mortality
Capone [46] Vitamin C & Zinc

  • Combination of zinc and Vitamin C supplementation inversely correlated with death incidence in 73 out of 103 hospitalised patients (71.6%) (Pearson correlation coefficient = −0.10, p-value not reported)
Jothimani [26] Zinc Reference; 0% (0/20) 5.48 (0.61–49.35); 18.5% (5/27)

  • No significant difference in death rates between zinc-deficient and non-deficient subjects (p > 0.05)
Bellmann-Weiler [21] Iron Reference; NR 0.458 (0.082–2.572); NR 0.418 (0.15–1.165); NR

  • No significant difference in mortality between patients with absolute or functional iron deficiency, compared to those with no iron deficiency (p > 0.05)

  • 13.3 higher odds of mortality in patients with moderate/severe anemia, compared to whose with no anemia (95% CI: 2.14–83.0, p = 0.006)
Wu [41] Iron (Serum Ferritin) 5.28 (0.72–38.48); NR Reference; NR

  • No significant difference in odds of death between those with serum ferritin above or below 300 ng/mL, and in median serum ferritin levels between survivors and non-survivors (p > 0.05)
Zhao [44] Iron (Serum iron, Pre-treatment)

  • No association between pre-treatment serum iron levels and odds of mortality in patients (OR: 1.00 (0.98–1.02)_ d)

  • No significant difference in median pre-treatment serum iron levels in COVID-19 survivors and non-survivors (median [IQR]: 6.2 µmol/L [4.3–8.0] vs. 4.1µmol/L [2.2–7.5], p > 0.05)
Iron (Serum iron, Post-treatment)

  • Up to 2% lowered odds of mortality with higher post-treatment serum iron levels (OR: 0.99 (0.98–1.00_ d)

  • Significantly higher median post-treatment serum iron levels in COVID-19 survivors, compared to non-survivors (median [IQR]: 19.1 µmol/L [13.2–25.6] vs. 5.5 µmol/L [3.5–11.1], p = 0.002)
Zhou [45] Iron (Serum Ferritin) 9.1 (2.04–40.58); 43.1% (44/102) * Reference; 7.7% (2/26) *

  • 8.1 times higher odds of death amongst those with serum ferritin >300 ng/mL, compared to those ≤300 ng/mL (p = 0.0038)

  • Significantly higher median (IQR) serum ferritin levels (μg/L) in survivors (503.2 (264.0–921.5) μg/L) and non-survivors (1435.3 (728.9–2000.0) μg/L, p < 0.0001)
Moghaddam [34] Selenium (Serum) NR; 13.0% (12/92) # NR; 29.7% (22/74) #

  • Significantly higher mean ± SD serum levels in survivors (53.3 ± 16.2 μg/L) than in non-survivors (40.8 ± 8.1 μg/L, p < 0.001)
Outcome: Hospitalisation Duration (Unit: days)
Carpagnano [22] Vitamin D 12.5 (8–20.5) #,f 8 (6–11.25) #,f

  • Shorter median length of stay in respiratory intermediate care unit for patients with severe deficiency (8 days), compared to those with non-severe deficiency (12.5 days, p-value not reported), as the former tend to experience death or transfer to intensive care units.
Maghbooli [30] Vitamin D3 5 (1–19) g 5 (1–23) g

  • No difference in median length of stay between those with and without deficiency or insufficiency (p > 0.05).
Jothimani [26] Zinc 5.7 (NR) * 7.9 (NR) *

  • Significantly longer mean length of stay in patients with deficiency (7.9 days), compared to those with no deficiency (5.7 days, p = 0.048)

  • 239% higher odds of hospitalisation≥7 days for deficient patients, compared to non-deficient patients (OR: 3.39 (95% CI: 0.99–11.57))

  • Significantly higher proportion of patients with deficiency hospitalised≥7 days (59.3%, 16/27), compared to those without deficiency (30%, 6/20; p = 0.047)
Outcome: Intensive care unit (ICU) Admission
Bellmann-Weiler [21] Iron Reference; NR 0.147 (0.017–1.297); NR 0.556 (0.225–1.373); NR

  • No significant difference in ICU admission rates between patients with absolute or functional iron deficiency, and those without iron deficiency
Carpagnano [22] Vitamin D NR; 12.5% (4/32) # NR; 20% (2/10) #

  • Similar ICU admission rates between patients with and without severe deficiency (p-value not reported)
Castillo [47] Vitamin D 0.03 (0.003–0.25) e; 2% (1/50) * Reference; 50% (13/26) *

  • 97% reduction in odds for ICU admission amongst inpatients treated with calcifediol, compared to those not given calcifediol

  • Significantly lower proportion of ICU admissions amongst inpatients treated with calcifediol, compared to those not given calcifediol (p < 0.001)
Maghbooli [30] Vitamin D3 NR; 14.3% (11/77) NR; 20.9% (33/158)

  • No significant difference in ICU admission rates between those with and without deficiency or insufficiency (p > 0.05).
Panagiotou [35] Vitamin D NR; 18.2% (8/44) * NR; 37.8% (34/90) *

  • No significant difference in mean difference for logarithmically transformed-25(OH)D levels between patients admitted and not admitted to the ITU (mean difference: 0.14; 95% CI: −0.15–0.41, p = 0.3), although those admitted to the ITU had lower 25(OH)D levels (33.5 nmol/L ± 16.8) than those not admitted (48.1 nmol/L ± 38.2)

  • Significantly higher vitamin D deficiency prevalence in patients admitted to ITU (81%), compared to those not admitted to the ITU (60.9%, p = 0.02).
Jothimani [26] Zinc Reference; 10% (2/20) 3.15 (0.58–17.67); 25.9% (7/27)

  • No significant difference in odds of ICU admission between patients with and without deficiency

  • No significant difference in ICU admission rates between patients with and without deficiency (p > 0.05)

* between-group p < 0.05; # between group p-value not reported; ᶧ quintiles classification: no deficiency and deficiency; ᶲ quintiles classification: no insufficiency and insufficiency; a Hazard ratio adjusted for recent bolus vitamin D3 supplementation; b Hazard ratio adjusted for age, sex, ethnicity, month of assessment, Townsend deprivation quintile, household income, BMI category, smoking status, diabetes, systolic blood pressure, diastolic blood pressure, self-reported health rating, and long-standing illness, disability or infirmity; c Hazard ratio adjusted for age, gender, and comorbidities for all following; d Adjusted for age, lymphocyte percentage, lymphocyte count, pretreatment serum iron level, and posttreatment serum iron level; e Adjusted for hypertension and type 2 diabetes; f Median (IQR); g Median (range).