Literature evaluations of drug-induced vitamin deficiencies
| Vitamin depleted | Causative medications | Study design | Patients (or participants) | Interventions | Relevant study outcomes | Results | References |
|---|---|---|---|---|---|---|---|
| Lipid-soluble | |||||||
| Vitamin A | Orlistat | Meta-analysis | 10,631 participants from 16 trials (average BMI: 36.3 kg/m2) | Orlistat 120 mg three times daily | Levels of fat‐soluble vitamins (A, D, E) and beta‐carotene | Orlistat therapy was associated with lower fat‐soluble vitamin and beta‐carotene levels | [2] |
| RCT | Patients aged 30~60 years with BMI ≥30 kg/m2 | Orlistat 120 mg three times daily for 4 years | Changes in plasma levels of fat-soluble vitamins | Significant decreases in vitamin A in the orlistat group compared with the placebo group (−0.22 vs. −0.19 μmol/L, P<0.05) | [3] | ||
| Cholestyramine | RCT | 303 patients with hypercholesterolemia | Dietary advice plus cholestyramine (8 to 16 g/d for 2 months) | Serum concentrations of vitamin E, β-carotene, lycopene, and vitamin A | 40% decrease in mean serum β-carotene (P<0.001) and 5% increase in vitamin A (P<0.001) after 2 months on cholestyramine | [7] | |
| Vitamin D | Corticosteroids | Cross-sectional | NHANES participants 2001~2006 (n=22,650) | – | Prevalence of low 25(OH)D level (<10 ng/mL) | 11% in steroid users vs. 5% in non-users (P=0.009) | [62] |
| AEDs | Cross-sectional | 58 patients on antiepileptic therapy | – | Serum 25(OH)D level Frequency of 25(OH)D <20 ng/dL |
Lower levels of 25(OH)D in AED users vs. control patients (28.2±10.3 ng/mL vs. 34.4±12.7 ng/mL; P=0.02) 20 (34.4%) of AED users vs. 2 (6.8%) of controls had 25(OH)D levels <20 ng/dL (P=0.01) |
[65] | |
| Orlistat | RCT | 30 obese subjects with a mean BMI of 47 kg/m2 | Orlistat 120 mg three times daily for 1 year | Serum levels of calcitriol | 140±39 pmol/L at baseline vs. 111±45 pmol/L after 1 year of orlistat therapy (P<0.05) | [68] | |
| Cholestyramine | RCT | 268 men aged 42~68 years | Cholestyramine vs. placebo for 7~10 years | Plasma levels of calcitriol | 99±190 pmol/L in placebo vs. 91±56 pmol/L in the cholestyramine group (NS) | [67] | |
| Vitamin K | Antibiotics | RCT | Critically ill children who received antibiotic therapy for a minimum of 14 days | Single dose of prophylactic vitamin K on day 1 of antibiotic therapy | Incidence of vitamin K deficiency | 15% of the total study population 13.3% vs. 16.7% in those with or without prophylactic vitamin K treatment, respectively (P=0.79) |
[72] |
| Nested case-control | Patients aged ≥20 years with or without a hemorrhagic event after using cephalosporins | – | Risk of a hemorrhagic event | aOR 1.71 (95% CI, 1.42~2.06) with the use of cephalosporins | [73] | ||
| Orlistat | RCT | Patients aged 30~60 years with BMI ≥30 kg/m2 | Orlistat 120 mg three times daily for 4 years | Changes in plasma levels of the fat-soluble vitamins | Significant decreases in vitamin K1 in the orlistat group compared with the placebo group (−0.08 vs. 0.07 μg/L, P<0.001) | [3] | |
| Water-soluble | |||||||
| Vitamin B1 | Furosemide | Prospective cohort | 50 patients >18 years of age with an ICU stay of at least 48 hours | Diuretic group (furosemide 20~160 mg/d) vs. control group | Mean serum thiamin levels in the baseline and post–ICU admission days 2, 5, and 10 | The diuretic group had significantly lower serum thiamin levels than the control (15.5±10.7 vs. 46.8±29.5 ng/mL; P<0.001 at baseline, 23.2±15.4 ng/mL vs. 49.0±38.0 ng/mL; P< 0.05 on day 2) | [10] |
| Fluorouracil | Retrospective chart review | 18 patients developed Wernicke-Korsakoff Syndrome during cancer treatment (5 of whom were taking fluorouracil) | – | Serum thiamin concentration | All patients who measured serum thiamine level (n=16) had abnormally low levels of serum thiamine (<7 nmol/L) | [17] | |
| Vitamin B3 | Isoniazid, pyrazinamide | Case report | 69-year-old female with cardiac, lung, and cutaneous sarcoidosis | – | – | Decreased serum nicotinic acid levels (4.5 μg/dL) after isoniazid therapy | [21] |
| Vitamin B6 | Isoniazid | Single-arm clinical trial | 20 patients with pulmonary tuberculosis | One week of therapy including isoniazid | Plasma levels of PLP | Decreased PLP levels in 18 patients (15 nmol/L at baseline vs. 11 nmol/L at 1 year, P<0.001) | [24] |
| AEDs | Cross-sectional | Patients with epilepsy | Converted from an inducing AED (phenytoin, carbamazepine) to a non- inducing AED (levetiracetam, lamotrigine, or topiramate) | Prevalence of low vitamin B6 (<5 ng/mL) | 16/33 (48%) with epilepsy vs. 1/11 (9%) of normal subjects (P<0.05) had low vitamin B6 levels Only 21% of patients had a low vitamin B6 level after switching to a non-inducing AED (P<0.05) |
[28] | |
| Levodopa | Case report | A 75-year-old man with advanced Parkinson’s disease | Levodopa/carbidopa 200/20 mg/d for 45 months | Serum vitamin B6 levels were undetectably low (<2.0 ng/mL) but normalized after vitamin B6 supplementation | [29] | ||
| Theophylline | Single-arm clinical trial | 7 healthy males | 15 weeks of theophylline treatment | Mean plasma and erythrocyte PLP levels | Erythrocyte PLP levels declined from 303.3±73.2 pmoL/g to 185.1±69.2 pmoL/g within 52 days of treatment (P=0.015) Plasma PLP levels declined from 62.6±26.8 nmol/L to 29.7±14.1 nmol/L within 32 days of treatment (P=0.015) |
[30] | |
| Vitamin B7 | AEDs | Cross-sectional | 12 adults with seizures requiring AEDs for more than 1 year | AEDs (carbamazepine, phenytoin, phenobarbital, valproic acid, and felbamate) | Urinary excretion of biotin metabolites | There was 2.5-fold greater urinary excretion of biotin metabolites in the anticonvulsant-treated group than the control group | [33] |
| Vitamin B9 | Methotrexate | RCT | 113 patients with rheumatoid arthritis | Methotrexate of 7.5~25 mg/wk with or without folate supplement | Plasma homocysteine and folate levels | A decrease in serum folate level by 4.1 nmol/L (95% CI, –7.6 to –0.6) and a 3.6 μmol/L (95% CI, 1.7 to 5.6) increase in homocysteine | [40] |
| Sulfonamides | Cross-sectional | 23 outpatients with ulcerative colitis treated with sulfasalazine vs. 20 healthy controls | – | Blood concentrations of folate | UC patients had significantly lower folate concentrations than controls (9.8±7.3 pmol/mL vs. 23.5±9.6 pmol/mL, P=0.015) | [44] | |
| AEDs | Prospective | AED-treated patients (n=2,730) vs. AED-untreated patients with epilepsy (n=170) vs. healthy controls (n=200) | AEDs (carbamazepine, gabapentin, oxcarbazepine, phenytoin, primidone, or valproate) | Mean serum folate level | Mean folate level was 6.0±3.5 ng/mL in the AED-treated group vs. 6.6±3.7 ng/mL in untreated patients (P=0.044) | [45] | |
| Vitamin B12 | Acid-suppressive agents (PPIs, H2RAs) | Meta-analysis | Four case–control studies and one observational study | Prolonged acid-suppressive agents use | Risk of vitamin B12 deficiency | HR 1.83 (95% CI, 1.36~2.46) | [48] |
| Case-control | Patients aged ≥20 years | – | Risk of vitamin B12 deficiency | OR 1.65 (95% CI, 1.58~1.73) with ≥2 years use of PPIs OR 1.25 (95% CI, 1.17~1.34) with ≥2 years use of H2RAs |
[49] | ||
| Case-control | Patients aged ≥65 years | – | Risk of vitamin B12 deficiency | OR 4.45 (95% CI, 1.47~13.34) with ≥12 months use of H2RAs and/or PPIs | [50] | ||
| Metformin | RCT | Patients with type 2 diabetes under insulin treatment | Metformin 850 mg or placebo three times a day for 4.3 years | Percentage change in vitamin B12 from baseline | 19% decrease in vitamin B12 level compared with placebo (P<0.001) | [53] | |
| RCT | Impaired glucose tolerance and FBG of 95~125 mg/dL, aged ≥25 years, BMI ≥24 kg/m2 | Metformin 850 mg or placebo twice daily | Incidence of vitamin B12 deficiency (≤203 pg/mL) | 4.3% of patients in the metformin group vs. 2.3% of patients in the placebo group had vitamin B12 deficiency at the 5-year follow-up (P=0.02) | [54] | ||
| Nested case-control | Patients with diabetes receiving metformin treatment | – | Risk of vitamin B12 deficiency | aOR 2.88 (P<0.001) with 1 g/d metformin dose increment aOR 2.39 (P=0.001) for those patients using metformin for ≥3 years |
[55] |
BMI = body mass index; RCT = randomized clinical trial; NHANES = National Health and Nutrition Examination Survey; NS = not significant; 25(OH)D = 25-hydroxyvitamin D3; AEDs = antiepileptic drugs; ICU = intensive care unit; PLP = pyridoxal phosphate; PPIs = proton pump inhibitors; H2RAs = H2-receptor antagonists; FBG = fasting blood glucose; HR = hazard ratio; OR = odds ratio; aOR = adjusted odds ratio; CI = confidence interval.