Risk Factors and Comorbidities Associated With Vitamin B₁₂ Deficiency in an Adult Population

Abstract

Introduction/Objectives:

Vitamin B₁₂ deficiency (B12D) is associated with multiple risk factors and comorbidities; however, there are no firm guidelines regarding screening for B12D in the population at risk. We aimed to identify the risk factors and comorbidities associated with B12D in an adult population.

Methods:

Retrospective review of entire cohort of adult patients who received outpatient medical care in our large urban tertiary healthcare system between January 1, 2011, and December 31, 2020. Variables were compared between the group who had B12D and those who did not (NoB12D).

Results:

Patients with B12D (n = 2666) were younger than NoB12D group (n = 2334; 62.2 ± 18.5 vs 76.5 ± 7.6 year; P < .001). There were significantly higher associations of certain factors in the B12D group compared to NoB12D group, such as female sex, recreational drug use, congestive heart failure (CHF), cerebrovascular accident (CVA), chronic obstructive pulmonary disease (COPD), asthma, peripheral neuropathy (PN), subacute combined degeneration of spinal cord (SCD), dementia, mild cognitive impairment (MCI), gastroesophageal reflux disorder (GERD), gastritis, inflammatory bowel disease (IBD), small bowel resection (SBR), colorectal surgery (CRS), bariatric surgery (BS), depression, bipolar disorder, anxiety disorder, schizophrenia, anemia, use of non-steroidal anti-inflammatory drug (NSAID), proton pump inhibitor (PPI), histamin-2 receptor antagonist (H2RA), and lithium. Compared to White race, Black and Hispanic races had higher odds of B12D (OR = 1.43, 95% CI = 1.02-2.01; P = .040, and OR = 3.15, 95% CI = 2.23-4.45; P < .001, respectively). Additional comorbidities with greater odds of B12D included respiratory disorders (COPD and asthma; OR = 1.41, 95% CI = 1.12-1.78; P = .003), anemia (OR = 253.72, 95% CI = 164.66-390.93; P < .001), neurological diseases (PN and SCD; OR = 2.78, 95% CI = 2.16-3.58; P < .001), cognitive disorders (dementia and MCI; OR = 3.37, 95% CI = 2.40-4.74; P < .001), gastrointestinal disorders (GERD, gastritis, IBD, SBR, CRS, and BS; OR = 2.38, 95% CI = 1.95-2.90; P < .001), PPI use (OR = 2.81, 95% CI = 1.88-4.20; P < .001), and NSAID use (OR = 1.40, 95% CI = 1.08-1.81; P = .011).

Conclusion:

Younger age, female sex, Black and Hispanic races, recreational drug use, CHF, CVA, COPD, asthma, PN, SCD, dementia, MCI, GERD, IBD, SBR, colorectal surgery, bariatric surgery, depression, bipolar disorder, anxiety disorder, schizophrenia, anemia, use of PPI, H2RA, NSAIDs, and lithium are associated with B12D.

Introduction

Vitamin B₁₂ deficiency (B12D) results in a myriad of disorders, which include neurological, psychiatric, hematological, gastrointestinal, dermatological, and other diseases. B12D is biochemically defined as a serum B₁₂ level below 232 pg/mL, or between 232 and 300 pg/mL with evidence of impaired cobalamin-dependent metabolism, such as elevated serum methylmalonic acid and/or homocysteine levels. B12D is believed to occur because of dietary deficiency, which happens to be an uncommon cause. Human body stores of vitamin B₁₂ are able to sustain a person for 3 to 4 years even if vitamin B₁₂ is completely removed from the diet. ¹ The body’s capacity to store about 1000 to 2000 times as much as the amount typically consumed in a day creates a unique ability of B12D to take several years to appear, if at all. ² The data from the 2013 to 2016 National Health and Nutrition Examination Survey (NHANES) showed that only a small fraction of the American population lacks vitamin B₁₂ in their diet. As per the data, less than 8% of women and less than 3% of men had dietary intake of vitamin B₁₂ below the estimated average requirement (EAR). ³ However, in the United States, approximately 20% of the adults older than 60 years have B12D. ⁴

Some of the risk factors for B12D include gastric or small intestine resections, inflammatory bowel disease, pernicious anemia, use of metformin for more than 4 months, use of proton pump inhibitors or histamine H₂ blockers for more than 12 months, being vegan or strict vegetarian, and adults older than 75 years.^2,4 Other factors, including psychiatric illness, ⁵ substance use disorder, ⁵ obesity, ⁶ metabolic syndrome, ⁷ heart failure, ⁸ and gender^9,10 have all been suggested as increasing risk. Although these factors have been identified across various studies, nevertheless there is a lack of a standard guideline to screen for B12D in whom and how often in the absence of symptoms. ¹¹

A review of the existing literature highlights the inconsistencies in the frequencies of associations of even the known risk factors. For example, the Diabetes Prevention Program Outcomes Study found that the prevalence of B12D and borderline-low vitamin B₁₂ levels was 2% to 10% higher in individuals who were on metformin compared to the ones who were not. ¹² On the other hand, there are studies that have reported a prevalence closer to 30%.^13,14 Based on the available data, the American Diabetes Association proposed periodic measurement of vitamin B₁₂ level in metformin-treated patients, especially in those who have anemia or peripheral neuropathy. ¹⁵ This still does not offer a firm recommendation.

The variation in protocols and data demonstrates a possible area of improvement to efficiently identify and treat B12D in individuals who happen to have certain risk factors, or comorbidities. With this in mind, we aimed to identify the risk factors and comorbidities associated with B12D in an adult population.

Materials and Methods

Study Design and Setting

Our study was a retrospective cohort analytic study conducted across all adult patients who visited our internal medicine primary care offices of a large tertiary healthcare system in both urban and suburban locations.

Participants

Patients aged 18 years or older who received care between January 1, 2011, and December 31, 2020. The inclusion criteria were adult patients aged 18 years or older, who had a serum vitamin B₁₂ test result documented in their medical records. We included patients who had a diagnosis of B12D and patients who had normal serum vitamin B₁₂ levels. The exclusion criteria were patients under the age of 18 years, or patients aged 18 years and older who did not have documentation of serum vitamin B₁₂ test result in their medical records.

Variables

We collected the following data on each patient: age, sex, race, alcohol use, and tobacco use; comorbid medical conditions, such as hypertension, hyperlipidemia, hypothyroidism, coronary artery disease (CAD), cerebral vascular attack (CVA), transient ischemic attack (TIA), peripheral neuropathy (PN), transverse myelitis (TM), subacute combined degeneration of spinal cord (SCD), gait disorder, frequent falls, dementia, mild cognitive impairment (MCI), congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), asthma, chronic kidney disease (CKD), liver disease, gastroesophageal reflux disorder (GERD), gastritis, inflammatory bowel disease (IBD), small bowel resection (SBR), colorectal surgery, bariatric surgery, diabetes mellitus (DM), hypothyroidism, other endocrine disorders, psychiatric conditions (depression, bipolar disorder, anxiety disorder, and schizophrenia), osteoarthritis and other rheumatologic diseases, anemia, and cancer. We also collected the data of systolic and diastolic blood pressures, body mass index (BMI), serum vitamin B₁₂ level, Folate, homocysteine, methylmalonic acid, estimated glomerular filtration rate (eGFR), hemoglobin; use of medications, such as proton pump inhibitor (PPI), histamine type 2 receptor antagonists (H₂RA), metformin, aspirin, non-steroidal anti-inflammatory drugs (NSAIDs), folic acid, and lithium. We recorded the collected data into REDCap (Research Electronic Data Capture; 2021, Vanderbilt University, USA) spreadsheet.

Data Source and Access

This study was reviewed and approved by the Institutional Review Board (IRB) of our healthcare system (IRB Study: 21-202). Permission was granted to use materials that were collected solely for research study purposes as per the Health Insurance Portability and Accountability Act (HIPPA) requirements, and the informed consent waivers were granted by the IRB. This study was fully compliant with the ethical standards set forth by the IRB. All investigators had full access to the data available only in the electronic medical records (Epic healthcare software, Epic Systems Corporation, Wisconsin, USA) of the list of patients approved by the medical informatics department, who were selected based on the selection criteria of the study.

Bias

To address the potential for inappropriate or presumptive diagnosis of vitamin B₁₂ deficiency, we excluded patients from grouping into the B12D group if they had a serum B₁₂ level between 232 and 300 pg/mL and a normal serum methylmalonic acid level or had no documentation of methylmalonic acid level in their electronic medical records.

Study Size

We selected a sample population of 5000 adult patients who visited our internal medicine primary care offices who fitted into our inclusion criteria. Based on the national data suggesting 3% to 6% of adults diagnosed as having B12D, we estimated that based on a population of approximately 25 000 patients, we needed a sample of at least 1000 patients with vitamin B12 deficiency.

Statistical Methods

Statistical analysis was computed by using SPSS (Statistical Package for the Social Sciences, version 15.01, IBM, Armonk, New York, USA) software. Patients were divided into 2 groups: first, patients who had B12D (B12D group) and second, who did not have B12D (NoB12D). We applied univariate analysis using independent t test, Mann Whitney U test, and Chi Square tests. We applied multivariate analysis by using Random Forest analysis to select the most important factors in B12D as shown in Table 2. Selected variables were entered using categorization in the logistic regression to model. The dependent variable was B12D, and the independent variables were the variables that were produced by the multivariate analysis. We placed the data elements into the model using the “Enter” method. The ability to appropriately classify the data was verified by use of area under the curve (AUC) which demonstrated an AUC value of 0.702 which was considered acceptable. Additional validation of the model was conducted by splitting the data, in which 80% was training data and 20% was testing data. The AUC for the training data was 0.719 while the AUC for the testing data was 0.726 which supported an acceptable ability to classify data. In this study, significance was defined as a P < .05.

Table 2.

Influence of Risk Factors on Vitamin B12 Deficiency.

Risk factor	B	P	Exp(B)	95% C.I. for Exp(B)
				Lower	Upper
Age	−0.103	<.001	0.902	0.894	0.91
Sex male	−0.52	<.001	0.594	0.492	0.719
White vs Black	0.357	.040	1.428	1.016	2.008
White vs Hispanic	1.147	<.001	3.149	2.229	4.450
White vs Other races	−0.109	.572	0.897	0.614	1.309
Recreational drug use	−0.329	.001	0.719	0.591	0.875
Cardiovascular	−0.252	.038	0.777	0.612	0.987
Respiratory	0.346	.003	1.413	1.124	1.777
Liver disease	−0.115	.389	0.891	0.685	1.159
Anemia	5.536	<.001	253.716	164.661	390.934
Cancer	−0.222	.044	0.801	0.645	0.994
OA/rheumatological disease	−0.869	.940	0.424	0.326	0.550
Diabetes mellitus	0.003	.973	1.003	0.834	1.207
Neurological disease	1.022	<.001	2.779	2.157	3.579
Cognitive disorders	1.216	<.001	3.373	2.402	4.737
GI disorders	0.865	<.001	2.375	1.945	2.900
Mental health disorders	0.023	.817	1.024	0.840	1.247
PPI	1.034	<.001	2.813	1.884	4.200
H2RA	0.139	.386	1.149	0.840	1.572
Metformin	−0.259	.162	0.771	0.536	1.110
Aspirin	0.058	.880	1.059	0.499	2.249
NSAIDs	0.336	.011	1.399	1.079	1.813

Abbreviations: GI, Gastrointestinal; H2RA, Histamine-2 receptor antagonist; NSAIDs, non-steroidal anti-inflammatory drugs; OA, Osteoarthritis; PPI, Proton pump inhibitor.

Results

A total of 5000 patients were included in this study. There were 2666 (53.3%) patients in the B12D group and 2334 (46.7%) patients in the NoB12D group (Table 1). The mean age of the B12D group was significantly younger than the NoB12D group (62.2 ± 18.5 vs 76.5 ± 7.6 year; P < .001). Women had a greater frequency of association with B12D compared to NoB12D group (68.6% vs 56.7%; P < .001). Most patients in both the groups identified as White; however, there were more patients who identified as African American or Hispanic in the B12D group compared NoB12D group (18.8% vs 11.7%, P < .001; and 16.0% vs 3.8%, P < .001; respectively). Use of tobacco products was comparable between the 2 groups; however, the use of alcohol was significantly less, and use of recreational drug was significantly more in the B12D group compared to the NoB12D group (39.3% vs 53.0%, P < .001; and 6.6% vs 2.6%, P < .001; respectively; Table 1). Among the laboratory parameters, the mean hemoglobin level and mean vitamin B12 level were significantly lower in the B12D group compared to the NoB12D group (12.0 ± 3.7 vs 13.2 ± 1.2 g/dl, P < .001; and 161.9 ± 33.0 vs 420.1 ± 16.0 pg/mL, P < .001; respectively), while there was no significant difference in the mean folate level (Table 1).

Table 1.

Baseline Characteristics.

Variable	Variable	B12 deficiency (n = 2666)	No B12 deficiency (n = 2334)	P
Age	Years, mean (SD)	62.2 (18.5)	76.5 (7.6)	<.001
Sex	Male, n (%)	840 (31.4)	1010 (43.3)	<.001
	Female, n (%)	1826 (68.6)	1324 (56.7)
Race	White, n (%)	1584 (59.4)	1828 (78.4)	<.001
	African American, n (%)	500 (18.8)	272 (11.7)
	Hispanic, n (%)	426 (16.0)	89 (3.8)
	Other, n (%)	156 (5.9)	144 (6.2)
Social	Tobacco use, n (%)	1159 (43.5)	979 (42.1)	.312
	Alcohol use, n (%)	1048 (39.3)	1236 (53.0)	<.001
	Recreational drug use, n (%)	175 (6.6)	61 (2.6)	<.001
Lab values	Hb (g/dl), mean (SD)	12.0 (3.7)	13.2 (1.2)	<.001
	B12 (pg/mL), mean (SD)	161.9 (33.0)	420.1 (16.0)	<.001
	Folate Low, n (%)	79 (2.9)	93 (3.9)	.310
Comorbidities	Hypertension, n (%)	1512 (56.7)	1664 (71.3)	<.001
	Hyperlipidemia, n (%)	1279 (48.0)	1541 (66.0)	<.001
	CHF, n (%)	221 (8.3)	128 (5.5)	<.001
	CAD, n (%)	458 (17.2)	478 (20.5)	.003
	CVA, n (%)	260 (9.8)	165 (7.1)	<.001
	COPD, n (%)	224 (8.4)	154 (6.6)	.016
	Asthma, n (%)	494 (18.5)	240 (10.3)	<.001
	DM, n (%)	714 (26.8)	650 (27.9)	.388
	Hypothyroidism, n (%)	529 (19.8)	444 (19.0)	.465
	OED, n (%)	152 (5.7)	163 (7.0)	.063
	Osteoarthritis, n (%)	680 (25.5)	713 (30.5)	<.001
	Other Rheum Dis, n (%)	234 (8.8)	630 (27.0)	<.001
	Peripheral Neuropathy, n (%)	318 (11.9)	0 (0.0)	<.001
	Transverse Myelitis, n (%)	4 (0.2)	0 (0.0)	.128
	SCD, n (%)	14 (0.5)	0 (0.0)	<.001
	Gait disorder, n (%)	167 (6.3)	145 (6.2)	.940
	Frequent falls, n (%)	169 (6.3)	124 (5.3)	.123
	Dementia, n (%)	161 (6.0)	79 (3.4)	<.001
	MCI, n (%)	116 (4.4)	56 (2.4)	<.001
	GERD, n (%)	802 (30.1)	492 ( 21.1)	<.001
	Gastritis, n (%)	165 (6.2)	15 (0.6)	<.001
	IBD, n (%)	88 (3.3)	16 (0.7)	<.001
	Small bowel resection, n (%)	67 (2.5)	11 (0.5)	<.001
	Colorectal surgery, n (%)	121 (4.5)	4 (0.2)	<.001
	Bariatric surgery, n (%)	264 (9.9)	3 (0.1)	<.001
	Depression, n (%)	645 (24.2)	396 (17.0)	<.001
	Bipolar Ds, n (%)	82 (3.1)	22 (0.9)	<.001
	Anxiety Dis, n (%)	658 (24.7)	445 (19.1)	<.001
	Schizophrenia, n (%)	25 (0.9)	4 (0.2)	<.001
	CKD, n (%)	330 (12.4)	314 (13.5)	.257
	Liver Dis, n (%)	95 (3.6)	141 (6.0)	<.001
	Anemia, n (%)	221 (14.9)	28 (1.2)	<.001
	Cancer, n (%)	566 (21.2)	653 (28.0)	<.001
Medications	Aspirin, n (%)	997 (37.4)	994 (42.6)	<.001
	NSAIDs, n (%)	1163 (43.6)	641 (27.5)	<.001
	PPI, n (%)	1231 (46.2)	492 (21.1)	<.001
	H2RA, n (%)	955 (35.8)	489 (21.0)	<.001
	Metformin, n (%)	503 (18.9)	654 (28.0)	<.001
	Folic acid, n (%)	462 (17.3)	11 (0.5)	<.001
	Vitamin B12, n (%)	1591 (59.7)	17 (0.7)	<.001
	Lithium, n (%)	23 (0.9)	0 (0.0)	<.001

SD = Standard deviation, Hb = Hemoglobin, DM = Diabetes mellitus, OED = Other Endocrine Disorders, CAD = Coronary artery disease, CVA = Cerebrovascular accident, SCD = Subacute combined degeneration of spinal cord, MCI = Mild Cognitive Impairment, CHF = Congestive heart failure, GERD = Gastroesophageal reflux disorder, IBD = Inflammatory bowel disease, COPD = Chronic obstructive pulmonary disease, Other Rheum Dis = Other rheumatological disorders, CKD = Chronic kidney disease, PPI = Proton Pump Inhibitor, H2RA = Histamine-2 receptor antagonist.

We found significantly higher frequencies of association of several comorbid medical conditions in the B12D group compared to NoB12D group, such as CHF (8.3% vs 5.5%; P < .001), CVA (9.8% vs 7.1%; P < .001), COPD (8.4% vs 6.6%; P = .016), asthma (18.5% vs 10.3%; P < .001), PN (11.9% vs 0.0%; P < .001), SCD (0.5% vs 0.0%; P < .001), dementia (6.0% vs 3.4%; P < .001), MCI (4.4% vs 2.4%; P < .001), GERD (30.1% vs 21.1%, P < .001)), gastritis (6.2% vs 0.6%; P < .001), IBD (3.3% vs 0.7%; P < .001), SBR (2.5% vs 0.5%; P < .001), colorectal surgery (4.5% vs 0.2%; P < .001), bariatric surgery (9.9% vs 0.1%; P < .001), depression (24.2% vs 17.0%; P < .001), bipolar disorder (3.1% vs 0.9%; P < .001), anxiety disorder (24.7% vs 19.1%; P < .001), schizophrenia (0.9% vs 0.2%; P < .001), and anemia (14.9% vs 1.2%; P < .001; Table 1).

Analysis of medication usage showed that there were higher associations of use of certain medications in the B12D group compared to NoB12D group, such as NSAIDs (0.9% vs 0.2%; P < .001), PPI (46.2% vs 21.1%; P < .001), H2RA (35.8% vs 21.0%; P < .001), and lithium (0.9% vs 0.0%; P < .001). We found significantly less use of metformin in the B12D group compared to NoB12D group (18.9% vs 28.0%; P < .001; Table 1).

Logistic regression analysis showed that compared to White race, Black and Hispanic races had higher odds of B12D (OR = 1.43, 95% CI = 1.02-2.01; P = .040 and OR = 3.15, 95% CI = 2.23-4.45; P < .001, respectively). Additional comorbidities with greater odds of B12D included respiratory disorders (COPD and asthma; OR = 1.41, 95% CI = 1.12-1.78; P = .003), anemia (OR = 253.72, 95% CI = 164.66-390.93; P < .001), neurological diseases (PN and SCD; OR = 2.78, 95% CI = 2.16-3.58; P < .001), cognitive disorders (dementia and MCI; OR = 3.37, 95% CI = 2.40-4.74; P < .001), gastrointestinal disorders (GERD, gastritis, IBD, SBR, colorectal surgery, and bariatric surgery; OR = 2.38, 95% CI = 1.95-2.90; P < .001), PPI use (OR = 2.81, 95% CI = 1.88-4.20; P < .001), and NSAID use (OR = 1.40, 95% CI = 1.08-1.81; P = .011; Table 2).

Discussion

The finding that patients in our B12D group were significantly younger than those in the NoB12D group contrasts with the prevailing literature, which consistently links B12D with increasing age. Epidemiological studies have long shown that the prevalence of B12D increases with age due to factors, such as reduced gastric acid secretion, malabsorption, and dietary inadequacies among older adults. ¹⁶ In another study, the elderly population was found to be at a heightened risk, primarily due to intrinsic factor deficiency and poorer B₁₂ absorption rates. ¹⁷ However, our data suggests that additional risk factors may contribute to B12D among younger patients, potentially related to dietary habits, chronic illnesses, or medications that impair absorption, even before advanced age becomes a predominant factor. Supporting this hypothesis, 1 study found that the relationship between low vitamin B₁₂ levels and age was significant only when dietary diversity was also limited, indicating that non-age-related factors, such as nutrition play a crucial role. ¹⁸ Some studies have suggested that certain younger populations, such as individuals with gastrointestinal disorders (eg, Crohn’s disease) or vegan individuals, have high B12D rates regardless of age. ¹⁹ These studies highlight that malabsorption, or dietary inadequacies can sometimes overshadow age as a risk factor, providing counterpoints to age-centric models of deficiency. Our results may therefore highlight a sub-cohort of younger patients with compounded risk factors, emphasizing the complexity of B12D and the need to consider non-age-related risk factors in clinical assessments.

Several studies support the finding that women are at increased risk of B12D.^19,20 This association could be driven by dietary factors, as women are more likely to adopt restrictive diets, such as vegetarian or vegan diets, which can lead to lower vitamin B₁₂ intake. ¹⁹ Additionally, hormonal differences, particularly during pregnancy and breastfeeding, may increase the need for vitamin B₁₂, thereby putting women at a higher risk of deficiency. ²⁰ Furthermore, studies have shown that conditions like autoimmune diseases, which are more prevalent in women, could contribute to an impaired ability to absorb vitamin B₁₂, particularly through mechanisms like pernicious anemia. ²¹ However, studies have also reported that men, particularly older men, may be more susceptible than women due to dietary habits, such as lower intake of fortified foods and higher consumption of processed meats, which can impact vitamin B₁₂ absorption. ²² Besides, disorders, such as peptic ulcers and alcohol-related liver disease, are more prevalent in men, which can also impair vitamin B₁₂ absorption. ²³ This suggests that while gender plays a role in B12D, sub-factors, such as lifestyle, diet, and comorbidities significantly influence whether men or women are more affected.

There is evidence in the literature that suggests race as a contributing factor to vitamin B₁₂ level in the body. A study found that people of the Black race had higher serum B₁₂ levels than their White counterparts. ²⁴ Similarly, Latin-American patients had levels intermediate to those of White and Black patients. The reason was attributed due to a combination of genetic and environmental factors. ²⁵ Another study reported that Black patients had higher serum vitamin B₁₂ levels across all age groups and sexes, compared to Asian and White patients. ²⁶ As per a report from the American Society of Hematology, Hispanic patients had higher prevalence of B12D compared to the Black patients. ²⁷ These differences were linked to medical comorbidities, nutritional practices and other social determinants of health. ²⁷ Race specific genetic factors, such as the Fucosyltransferase 2 (FUT2) and transcobalamin-II (TCN2) genes, have shown to have the strongest associations with serum B₁₂ levels.^28,29 The FUT2 gene also produces antigens that Helicobacter pylori target to cause infection, which then interferes with the release of intrinsic factors leading to vitamin B₁₂ malabsorption. Patients of Black and Hispanic races in our study had B12D which adds another dimension to the existing literature.

Several studies have associated recreational drug use with B12D. Chronic alcohol consumption impairs the absorption of vitamin B₁₂ by reducing intrinsic factor secretion, essential for vitamin B₁₂ absorption in the intestines. ³⁰ Additionally, opioids have been shown to disrupt gastrointestinal function, further affecting nutrient absorption, including vitamin B₁₂. ³¹ Alterations in gut microbiota due to recreational drug use can also impact vitamin B₁₂ metabolism, highlighting the complex relationship between recreational drugs and nutrient deficiencies. ³² These studies collectively underscore the need for monitoring and addressing vitamin B₁₂ levels in individuals with a history of recreational drug use to mitigate related health issues.

We found a strong association of CHF with B12D. Studies regarding B12D and CHF are limited. Several studies argue that without the evidence of macrocytosis, B12D had not been adequately considered. One possible etiology is that patients might have a mixed form of anemia, and B12D should be considered as treatment showed improved quality of life and exercise capacity. ³³ Also, elevations in red cell distribution width (RDW) have been associated with increased risk of symptomatic HF and are associated with increased morbidity and mortality in patients with HF. ³⁴ Elevated methylmalonic acid (MMA) levels, which may indicate subclinical B12D, are seen in HF patients and may be a useful biomarker for oxidative stress and to assess the prognosis as a predictor of all-cause mortality due to its role in mitochondrial-lysosomal interplay and decrease in mitochondrial respiration with B12D.^9,35 Elevated homocysteine, which is present in B12D, is also associated with arterial endothelial dysfunction and an independent risk factor for cardiovascular disease and CHF. ³⁵ One case report showed reversibility of dilated cardiomyopathy with treatment of significant B12D. ³⁶ Other studies have determined that vitamin B₁₂ levels have no relationship to prognosis. ³⁷ One study showed increased vitamin B₁₂ and direct bilirubin levels were associated with death in stable HF patients, though there is no independent association with mortality. ³⁸ More studies are needed to further elucidate the importance of B12D in patients with CHF.

Association of CVA with B12D can be explained by the fact that CVA occurs when there is a blockage or reduction in the blood supply to the brain that results in cell death. Previous studies have shown that reduced levels of B-vitamins, such as vitamin B₁₂, are a modifiable risk factor due to their role in the metabolism of homocysteine. ³⁹ One proposed mechanism suggests that increased levels of homocysteine lead to decreased levels of nitric oxide. This causes changes in vasodilation leading to vascular damage due to free radical production and lipid peroxidation. ⁴⁰ Since vitamin B₁₂ metabolizes homocysteine, increasing levels of vitamin B₁₂ have been shown to be beneficial for patients with elevated homocysteine levels to reduce risk of stroke.^41,42

Association of COPD with B12D has also been reported. ⁴³ Etiology of B12D in patients with COPD is multifactorial, which includes increased basal metabolic rate, alterations in diet and caloric intake, and loss of skeletal mass. ⁴⁴ Although few studies have reported that even with vitamin B₁₂ supplementation, the pulmonary and cardiovascular risk do not decrease.^{45 -47} Till date, there is no link reported between vitamin B₁₂ levels and its effect on lung function and overall effect on asthma. ⁴⁸ Hence, our finding of association of asthma with B12D is unique.

Our study found a strong association of peripheral neuropathy with B12D. Peripheral neuropathy is a result of damage to small or large nerve fibers through various pathophysiologies. ⁴⁹ Vitamin B₁₂ acts as a cofactor for the conversion of methylmalonyl-CoA to succinyl-CoA and homocysteine to methionine, crucial products for lipid, carbohydrate, and nucleic acid synthesis. ⁵⁰ In B12D, damage to the myelin sheath resulting in peripheral neuropathy occurs due to the accumulation of methylmalonic acid (MMA) and homocysteine. It presents as peripheral neuropathy and ataxia, and more chronically as subacute combined degeneration of the spinal cord (SCD). ⁵¹ SCD involves degeneration of white matter tracts of the posterior and lateral columns of the lower cervical and upper thoracic spinal cord segments. Vitamin B₁₂ plays a crucial role in DNA synthesis and odd-chain fatty acid metabolism, which are necessary for the maintenance of the myelin sheath surrounding neurons. Vitamin B₁₂ is a cofactor for homocysteine methyltransferase, an enzyme which converts homocysteine to methionine. Methionine is needed to form S-adenosylmethionine, which is a co-substrate involved in methyl group transfers required for the integrity of the myelin sheath. It is thought that a deficiency of cobalamin results in methyl group deficiency and the subsequent demyelination seen in SCD.^52,53

B12D is associated with cognitive decline and dementia.^54,55 Several studies have demonstrated an association with B12D and elevated homocysteine levels with dementia.^56,57 Larger studies have found low-quality evidence supporting elevated homocysteine levels, a byproduct of low vitamin B₁₂, to be associated with the onset of dementia, with treatment slowing the rate of brain atrophy. ⁵⁸ However, there are studies that found no clear causal link between B12D and cognitive impairment or dementia. ⁵⁹ We found a strong association between B12D and dementia.

GERD was found to have higher levels of association with B12D. Current literature cites GERD in association with B12D in the context of long-term PPI use. ⁶⁰ There is a well-documented association between GERD treatment with PPIs and B12D, primarily due to impaired absorption of the vitamin in the absence of adequate gastric acid. Regular monitoring of vitamin B₁₂ levels may be warranted in patients on long-term PPI therapy, especially those with additional risk factors for B12D.^61,62

We found that IBD was associated with B12D. A potential explanation for these results is the impact IBD has on intestinal absorption of vitamin B₁₂. The B₁₂-intrinsic factor complex is taken up into the enterocytes within the terminal ileum. ⁶³ It has been reported that only ileal resection of greater than 20 cm in patients with Crohn’s disease was associated with B12D. ⁶⁴ Patients with ulcerative colitis and a history of restorative proctocolectomy were also found to be at an increased risk for B12D, although the mechanism of this relationship is not well understood. ⁶⁴

In our study, B12D was associated with small bowel resection (SBR), colorectal surgery, and bariatric surgery. In this group of patients, B12D is believed to be due to malabsorption from a reduced absorptive surface area following the surgery, as well as reduced oral intake. ⁶⁵ A meta-analysis reported an association of B12D in patients who underwent Roux-en-Y Gastric Bypass (RYGB) or sleeve gastrectomy (SG), with the risk being greater in RYGB group.^{66 -68} In comparison to SG, RYGB involved multiple components of the small intestine, where vitamin B₁₂ is absorbed. ⁶⁵ Given the extensive nature of vitamin B₁₂ absorption associated with the duodenum, ileum, and jejunum, SBR involves increased risk of B12D due to malabsorption, and in some cases, short bowel syndrome. ⁶⁹ Data regarding B12D in colorectal surgery is somewhat limited. B12D was noted in 1 study that followed patients who underwent restorative proctocolectomy, which involved resection of the terminal ileum involved in B12 resorption. ⁷⁰

Research has consistently highlighted a significant association between B12D and various mental health disorders. For instance, studies indicate that low levels of vitamin B₁₂ are linked to increased risks of depression and bipolar disorder, potentially due to its role in synthesizing neurotransmitters such as serotonin and norepinephrine, which are involved in mood regulation and maintaining neuronal health. ⁷¹ Additionally, B12D can lead to elevated levels of homocysteine, an amino acid that has been implicated in the pathogenesis of mood disorders. ⁷² Furthermore, evidence suggests that B12D may exacerbate symptoms of anxiety disorders and schizophrenia, possibly by impairing cognitive function and altering brain metabolism.^73,74 The role of vitamin B₁₂ in myelin formation is crucial for proper neuronal function. B12D can result in neurological changes such as peripheral neuropathy and cognitive disturbances, which may manifest as anxiety or exacerbate existing anxiety disorders. ⁷⁴ Elevated homocysteine levels associated with B12D have also been linked to increased anxiety and cognitive impairment. ⁷⁵ Schizophrenia has been linked to disturbances in neurotransmitter systems, particularly those involving glutamate and dopamine. Studies have shown that patients with schizophrenia often have lower vitamin B₁₂ levels compared to the healthy controls. ⁷⁶ Additionally, B12D can impair cognitive function and exacerbate symptoms of psychosis, potentially through disruptions in methylation processes and neurotransmitter imbalances. Overall, adequate level of vitamin B₁₂ appears to be crucial for mental well-being, and having B12D is associated with mental health disorders.

The association of B12D with anemia has been widely established, which supports our findings.

Aspirin (ASA) use has been reported to be associated with B12D in 1 study. ⁷⁷ While the mechanism has not yet been elucidated, ASA and NSAIDs are known to impact gut mucosa through cyclooxygenase (COX) inhibition. This, in turn, may lead to a malabsorptive effect on vitamin B₁₂ in the gut; however, more studies are required to further understand the association. ⁷⁷ They found that patients on ASA were more likely to have B12D even when adjusted for age, gender, and H. pylori infection. ⁷⁸ We found similar association of ASA use with B12D in our patients.

The mechanisms leading to B12D in patients taking H2RA and PPI are similar and are likely driven by an increasingly alkaline environment in the stomach. Reduction of hydronium ions in the stomach leads to impaired conversion from pepsinogen to pepsin. Consequently, the resultant achlorhydria from PPIs and H2RAs is hypothesized to reduce vitamin B₁₂ availability, which would subsequently be bound to R-binder to complete the downstream pathways necessary for adequate vitamin B₁₂ absorption in the distal small intestines.^61,79,80 A study found that there was an overall increased risk of PPI use amongst B12D patients, however in the populations assessed, there was still a more clear association to be made. ⁷⁹ Previous literature has suggested increased risk of vitamin B₁₂ deficiency in patients on a PPI over H2RA, however further studies are needed. ⁸⁰

We found that lithium treatment was associated with B12D. One study reported that patients who received lithium had mean serum B₁₂ concentrations approximately 20% lower than those who did not receive lithium. ⁸¹ However, the clinical significance of this finding remains unclear, as there was only a nonsignificant trend toward an increased prevalence of assay-defined B12D in the lithium group at 1 center, and no cases at the other center. ⁸¹ Additionally, a statistically significant direct association was reported between hair lithium and cobalt concentrations, suggesting a role of lithium in the transport and distribution of vitamin B₁₂. ⁸² This interaction may contribute to the observed decrease in serum B₁₂ levels in patients on lithium therapy. While the exact mechanism by which lithium affects vitamin B₁₂ metabolism is not fully understood, it is important to monitor vitamin B₁₂ levels in patients undergoing long-term lithium treatment, especially if they present with symptoms of B12D, such as megaloblastic anemia or neurological impairments.^81,82

Association of Metformin use and B12D has been widely reported.^{12,83 -87} While the mechanism is still not well delineated, Metformin is believed to induce malabsorption of vitamin B₁₂ and intrinsic factor in the ileum through calcium dependent disruption of B₁₂-IF uptake on ileal cell membranes.^{84 -86} In our study, almost 19% of patients who had B12D reported taking Metformin which was significantly less than the group who were not taking Metformin. Hence, we found no association of Metformin use with B12D.

Our study had some limitations. Being a retrospective study, we collected the data on the comorbidities based on the date of documentation for which we had to rely solely on the entry of the conditions made by the care teams. A few variables had low frequencies but significantly higher associations which might have influenced the logistic regression analysis. Although the study patients belonged to our medical offices in both urban and suburban locations, nevertheless our results cannot be generalized. The major strength of this study was a large database of patients who consistently followed the clinicians in the care teams for greater than 2 decades, which allowed the care teams to document each comorbidity and other variables in a chronological manner.

Conclusion

We conclude that younger age, female sex, Black and Hispanic races, recreational drug use, CHF, CVA, COPD, asthma, PN, SCD, dementia, MCI, GERD, IBD, SBR, colorectal and bariatric surgery, depression, bipolar disorder, anxiety disorder, schizophrenia, anemia, use of PPI, H2RA, NSAIDs, and lithium are associated with B12D. Further studies at different settings are needed to identify similar, or additional risk factors.