22.2: Biomarkers of vitamin B12 status (22b.2)

Biomarkers of vitamin B12 status have been reviewed by the expert international panel as part of the Biomarkers of Nutrition for Devel­op­ment (BOND) project; see Allen et al (2018) for more details. Risk of B12 insufficiency can be assessed through the measure­ment of dietary intake of vitamin B12, a biomarker of exposure. Quantitative dietary assessment methods such as 24h recalls and food records must be used; for more details see Chapter 3: Measure­ment of food con­sump­tion of individuals. The variation in day-to-day intake of vitamin B12 is often large when vitamin B12‑rich foods such as liver are consumed, so multiple days of intake should be recorded on each individual and the observed distribution of intakes adjusted to yield usual intake distribution. This is achieved by removing the variability introduced by the day-to-day variation in B12 intakes within an individual. In this way, the proportion of individuals in a population group at risk of B12 insufficiency can be determined. Details of the adjustment process are given in Chapter 3.

In population groups consuming crystal­line B12 in fortified foods and in supple­ments, con­sump­tion is often intermittent necessitating several days of intake or use of a food frequency questionnaire. Details of the B12 content of many dietary supple­ments are available in the U.S. federal Dietary Supple­ment Ingredient Database. No correction is made for differences in the bioavailability of crystal­line B12 from supple­ments or fortified foods compared with B12 from food because the absorption differences between these two B12 sources are assumed to be small.

Alternatively, instead of determining quantitative B12 intakes, estimates of the usual con­sump­tion of animal products per day, week, or month can be used to assess risk of inadequate intakes of B12 for individuals in households and for population groups; see Chapter 3 for more details. Several studies have reported significant positive associations between animal source food intake and vitamin B12 status (McLean et al., 2007).

A combination of biochemical biomarkers is recom­mended to assess actual B12 status, together with knowledge of factors such as age, gender, pregnancy, folate status, and bacterial over­growth, all known to impact some of the B12 biochemical biomarkers. The four priority bio­chem­ical biomarkers recom­mended by the Biomarkers of Nutrition for Devel­op­ment (BOND) project to detect subclinical vitamin B12 defi­ciency are total serum vitamin B12, serum holotrans­cobalamin (holo TC), plasma homo­cysteine, and serum methyl­malonic acid; see Allen et al., 2018 for more details. Biochemical assays for each of these biomarkers together with the recom­mended tests to determine the cause of the vitamin B12 defi­ciency are outlined below.

The BOND project caution that currently no gold standard exists to define vitamin B12‑deficiency using these biochemical biomarkers. If subclinical vitamin B12 defi­ciency is suspected, then total serum vitamin B12 or serum holoTC is frequently measured first, followed by methyl­malonic acid (MMA) or plasma homo­cysteine (tHcy) concen­trations.

22b.2.1 Serum vitamin B12

Of the vitamin B12 in the serum, 20‑30% is attached to the transport protein trans­cobalamin (previously known as trans­cobalamin II), and the remaining 70‑80% is bound and carried by hapto­corrin, the latter historically designated as trans­cobalamin I and trans­cobalamin III. Hence, total plasma or serum B12 concen­tration includes B12 (cobalamin) bound to trans­cobalamin (holoTC) and B12 bound to hapto­corrin. Assessment of total B12 in plasma or serum provides information on intake of B12, long-term B12 status of the individual, and liver stores.

Of these two circulating transport proteins, only trans­cobalamin bound to B12 to form holotrans­cobalamin (holoTC) in serum is meta­bolically active, delivering vitamin B12 to receptors on cell membranes. In contrast, because there are no hapto­corrin receptors located on most cells, B12 cannot be delivered to extrahepatic tissues by hapto­corrin. Moreover, hapto­corrin binds not only B12 but also B12 analogs. However, because gastric intrinsic factor does not bind the B12 analogs (only B12), the B12 analogs are not reabsorbed but ultimately excreted in the stool (Renz, 1999).

In early vitamin B12 defi­ciency, when individuals are in negative balance, the amount of vitamin B12 attached to holoTC falls, but there is no concomitant decline in the total serum vitamin B12 concen­trations. The latter often remain normal for weeks or months despite low serum holoTC levels. Serum vitamin B12 concen­trations only decline when the percentage saturation of total trans­cobalamin with vitamin B12 falls below 5%(Herbert, 1987).

Total serum vitamin B12 concen­trations are expressed aspmol/L, or aspg/mL in clinical practice with 1.0pg/mL=0.7378pmol/L. Concen­trations above the normal level of 221pmol/L are indicative of an adequate B12 status, whereas those between 150 and 221pmol/L, or often below 150pmol/L, are indicative of B12 depletion or deficiency, respectively.

Traditionally, total serum vitamin B12 is often the first-line biochemical test used for routine screening for vitamin B12 defi­ciency; concen­trations reflect both the vitamin B12 intake (Dullemeijer et al., 2013) and body stores. However, to date a single accepted cut‑value for serum B12 indicative of deficiency has not been established. Moreover, the serum B12 test has low sensitivity and specificity and may overestimate B12 defi­ciency when used alone (Mineva et al., 2021). For example, clinical cases of vitamin B12 defi­ciency have been reported even in persons with low-to-normal total serum vitamin B12 levels (Lindenbaum et al., 1988). The latter may be explained by the presence of a physio­logically inactive proportion of serum B12 being bound to the carrier protein hapto­corrin, as explained earlier. Many, but not all of these cases, have shown a positive response to treatment with vitamin B12, based on improvements in abnormal metabolite levels and clinical indices (Stabler et al., 1990).

Most available prevalence data on B12 defi­ciency based on total serum B12 concen­trations are focussed on older adults. In reports based on 48 cohorts, 10‑19% of adults more than 60 years of age had total serum B12 levels below 148 or 150pmol/l and said to be indicative of B12 defi­ciency. A high prevalence of low serum / plasma B12 has also been reported among pregnant women (27.5% in 11 cohorts) and children (12.5% in 14 cohorts)(Smith et al., 2018).

Factors affecting total serum vitamin B12

Age influences levels of total serum vitamin B12. Infants and children have higher serum B12 concen­trations compared with adults, irrespective of ethnicity, perhaps in part attributed to their higher intake of milk in early childhood and a more efficient hepatic storage of B12 (Abildgaard et al., 2022; Sobczynska-Malefora et al., 2023). During adulthood, serum B12 concen­trations tend to decline with age, especially in the elderly (Wahlin et al., 2002), probably arising from the gradual decrease in both gastric acidity and intrinsic factor production that occurs with aging (Asselt et al., 1998; Carmel, 1997). An age-related decrease in the intake of vitamin B12 may be an additional factor (Johnson et al., 2003). However, such a decline in serum B12 among elderly adults is not consistent (Vogiatzoglou et al., 2009; Mineva et al., 2019), possibly due to B12 supple­ment use in this age group. In the United States NHANES 2011 to 2014 survey, for example, serum vitamin B12 concen­trations were reportedly 13% higher in persons aged >70 y compared to those in the 20‑39y age group, a trend attributed mostly to B12 supple­ment usage (Mineva et al., 2019).

Gender may affect serum vitamin B12 concen­trations, although results have been inconsistent. In smaller, less well controlled studies levels for men and women have been similar, irrespective of age (de Carvalho et al., 1996; Wahlin et al., 2002). In studies that have controlled for factors known to affect total serum vitamin B12, concen­trations were higher in women than in men (Mineva et al., 2019; Metz et al., 1971; Fernandes-Costa et al., 1985), whereas in others no differences between the sexes were found, irrespective of ethnicity (Sobczynska-Malefora et al., 2023).

Pregnancy can lead to a 25‑30% fall in total serum vitamin B12 concen­trations between 20‑ and 30‑weeks of gestation. The decline is due to decreased synthesis of hapto­corrin and hemodilution. Holotrans­cobalamin (holoTC) concen­trations, however, remain unchanged during pregnancy so the decline is unlikely to reflect B12 depletion. Indeed, by 14 weeks postpartum, total serum B12 concen­trations are no longer low (Allen et al., 2018).

Lactation is accompanied by a 40% increase in serum B12 at 6 weeks post-partum, remaining at this elevated level while the mother is lactating. A corre­sponding moderate increase in the functional meta­bolic biomarkers — plasma methymalonic acid and homo­cysteine — suggests that some depletion of maternal intracellular vitamin B12 has occurred even though serum B12 concen­trations are high (Varsi et al., 2018).

Ethnicity affects serum/plasma B12 concen­trations. People of African origin have higher serum B12 concen­trations than other ethnic groups. For example, in the US NHANES 2011-2014 survey, non-hispanic black adults had higher B12 concen­trations than non-hispanic white adults, even after controlling for age group and B12 supple­ment use (Mineva et al., 2019). This ethnic difference is attributed in part to a higher concen­tration of B12 binding proteins, most notably the transport protein hapto­corrin. However, because B12 bound to hapto­corrin is not meta­bolically active, use of only total serum B12 may result in Black patients being incorrectly identified as having a high B12 status. Clearly, failure to apply ethnic-related serum B12 reference intervals will lead to under­diagnosis of B12 defi­ciency and any associated comorbidities (Sobczynska-Malefora et al., 2023).

Folate status, both deficiency and excess, might affect total serum vitamin B12 levels. Folate deficiency may result in moderately low total serum vitamin B12 concen­trations because both folate as a cosubstrate (methyl donor) and vitamin B12 (as a coenzyme) are required in the remeth­yl­ation meta­bolic pathway of homo­cysteine (see Figure 22b.2) (Klee, 2000).

The mechanism for the interaction between excess folate and B12 is uncertain and warrants further investigation. One proposed hypothesis, as yet untested, is that high-dose folic acid supple­ments cause depletion of the transport protein holotrans­cobalamin in serum, thus exacerbating B12 defi­ciency. Older adults with both low total serum B12 concen­trations and increased levels of serum folate appear to have a higher risk of anemia and dementia, based on a review of the clinical and epidemiological evidence. The high serum folate levels are presumed to arise from exposure to high intakes of folic acid, the synthetic form of folate used in fortified foods and supple­ments (Miller et al., 2024).

Bacterial over­growth of the small intestine may lead to low serum B12 concen­trations and vitamin B12 defi­ciency because the bacteria utilize vitamin B12 for their own meta­bolism so B12 cannot be absorbed by the body.

Fish tapeworm also leads to B12 mal­absorp­tion because the tapeworms consume B12 for their own needs, so serum B12 concen­trations fall.

Chronic infection with Helicobacter pylori, often accompanied by intestinal bacterial over­growth, may result in low serum B12 concen­trations. The latter arise from mal­absorp­tion from food-bound B12 induced by the hypochlorhydria associated with atrophic gastritis caused by the H pylori jnfection of the stomach (Sobczynska-Malefora et al., 2021).

Lack of gastric intrinsic factor caused by pernicious anemia or gastrectomy leads to low total serum vitamin B12 concen­trations. Intrinsic factor (IF), a glycoprotein synthesized by the gastric parietal cells, is required for the absorption of vitamin B12 in the small intestine. In some cases of pernicious anemia, parietal cell auto-antibodies lead to the destruction of IF‑producing parietal cells in the stomach resulting in B12 mal­absorp­tion, and thus deficiency. The incidence of pernicious anemia increases with age, and is especially common among persons of European or African descent (Stabler & Allen, 2004). Parietal cell autoantibodies are also found in other autoimmune conditions such as Graves disease, hypothyroidism, and Addison’s disease (Sobczynska-Malefora et al., 2021).

Other conditions that produce elevated hapto­corrin levels increase total serum B12. Examples of such conditions include myeloproliferative disorders (e.g., chronic myelogenous leukemia, polycy­themiavera), severe liver diseases, and alcoholism. In these conditions, although total serum vitamin B12 concen­trations may appear elevated or apparently normal, tissue vitamin B12 concen­trations will be low (Arendt & Nexo, 2012).

Some disease states are associated with low total serum vitamin B12 concen­trations because of an inability to absorb or digest protein-bound vitamin B12. Disease states associated with B12 mal­absorp­tion due to villous atrophy or mucosal impairment include ileal Crohn’s or Celiac disease, respectively. In disease states such as atrophic gastritis and pancreatic insufficiency, low serum B12 concen­trations arise because digestion of protein-bound B12 is compromised. Other intestinal diseases such as tropical sprue, intestinal lymphoma, amyloidosis, and short bowel syndrome have also been associated with low serum B12 levels and B12 depletion.

Surgery involving partial or complete removal of any part of the gastro­intestinal tract, including some bariatric surgery procedures, will reduce food-bound B12 absorption and lower serum B12 concen­trations.

Inborn errors of intracellular B12 meta­bolism can result in low total serum B12 levels through their effects on absorption, transport, or meta­bolism of B12. The genetic defects range from rare gene deletion and mutations to single nucleotide polymorphisms with more mild and subtle effects. In congenital pernicious anemia, low serum total B12 concen­trations arise from B12 mal­absorp­tion caused by mutations in a gene (GIF,CBLIF) and defects in the synthesis of intrinsic factor. The rare autosomal recessive disorder Immerslund-Grasbeck (IGS) disease is caused by mutations in either the gene Amnion Associated Transmembrane Protein (AMN) or Cubilin (CUBN) gene responsible for the synthesis of cubam receptors (Sobczynska-Malefora et al., 2021).

Genetically predetermined deficiencies of the two B12 transport proteins, congenital trans­cobalamin deficiency (gene locus 22q11.2-qter) or hapto­corrin deficiency (gene locus 11q11-q12), can be associated with total serum B12 values that may be normal, or falsely low, respectively (Amos et al., 1994). Individuals with defects of methylmalonyl CoA mutase or polymorphisms occurring in the genes that code for enzymes in the cobalamin adenosylation pathway, may also have low serum vitamin B12 levels (Kano et al., 1985). Eight genetic defects in B12 meta­bolism have been identified all of which result in a failure to utilize B12 by the target cells; for more details see Sobczynska-Malefora et al. (2021).

Restricted dietary intake of B12 is associated with veganism and to a lesser extent vegetar­ianism, dietary patterns that avoid all or certain animal-source foods, respectively. Vegans have low total serum B12 concen­trations unless crystal­line B12 from fortified foods or supple­ments are consumed. To date, how long it takes for vitamin B12 defi­ciency to develop when individuals adhere exclusively to a fully plant-based diet is unknown.

There have been several reports of low plasma vitamin B12 levels among Asian Indians in Great Britain, the United States, and the Indian Subcontinent, many of whom are likely to practice vegetar­ianism so intake of animal-source foods is low (Antony, 2003; Antony, 2001; Carmel et al., 2002). Several investigators have reported an inverse relationship between intake of animal-source foods and total serum vitamin B12 levels. In a study in Germany of omnivores and vegetar­ians, none of whom were taking vitamin supple­ments, those consuming omnivorous diets had higher serum B12 concen­trations than the lacto‑ or lacto‑ovo vegetar­ians both of whom avoided meat, poultry, and fish, who in turn had higher levels than the vegans (Herrmann et al., 2003). Similar trends in serum B12 concen­trations have been reported among British male omnivores, vegetarians, and vegans in the EPIC Oxford cohort study (Gilsing et al., 2010).

Dietary intakes of omnivores are not generally strongly related to serum vitamin B12 levels; low correlations in daily B12 intakes have been linked to the large size of liver vitamin B12 stores (2‑3mg) (Doets et al., 2013). A meta-analysis of randomized controlled trials (RCTs) and observational studies investigated the dose-response relationship between vitamin B12 intake and serum or plasma vitamin B12 concen­trations. Doubling vitamin B12 intake was estimated to increase the vitamin B12 concen­tration in serum or plasma by 11% (95% CI: 9.4%, 12.5%) (Dullemeijer et al., 2013). Several cross-sectional studies among healthy adults with normal absorption of B12 show that serum B12 concen­trations reach a maximum plateau and stabilize at B12 intakes of 4‑7µg/d (Bor et al., 2010). Mode of infant feeding affects total serum B12 levels; infant formula has a higher B12 content than human milk (Greibe et al., 2016).

Vitamin B12 supple­ment use increases serum B12 concen­trations. In the US NHANES 2011‑2014 survey, serum B12 concen­trations for adults were about 40% higher in crystal­line B12 supple­ment users than nonusers overall (i.e., for age groups 20‑39, 40‑59, 60‑69, and >70 years), with the greatest increase among older persons. In non‑ supple­ment users, however, serum B12 concen­trations were similar across all four age groups (Mineva et al., 2019).

Medications such as histamine 2 receptor antagonists and proton pump inhibitors (PPIs) (lansoprazole, omeprazole, and esomeprazole) that suppress gastric acid may lower total serum B12 levels by impairing the release of B12 from food (Damodharan et al., 2021; Mumtaz et al., 2022). Metformin, used as treatment for type 2 diabetes, induces B12 mal­absorp­tion and has also been associated with low total serum B12 concen­trations(Al‑Fawaeir & Al‑Odat, 2022). However, the low serum B12 levels may not reflect true B12 defi­ciency in view of the finding in animal studies that metformin increases liver accumulation of vitamin B12 (Greibe et al., 2013; Obeid et al., 2024).

Storage of blood samples prior to separation of the serum affects B12 concen­trations if separation is delayed. Such delays can lead to an initial rise in total serum B12 concen­trations, followed by a significant drop after 1 to 3 days (Allen et al., 2018). Hence to ensure accurate measure­ments, it is important to separate the blood samples and freeze the serum as soon as possible after collection.

Analytical method may affect the reported serum vitamin B12 concen­trations. Some of the nonfunctional analogs of vitamin B12, as well as the cobalamins, were measured in earlier radio-assays. This led to overestimates of actual serum B12 concen­trations. This problem has now been overcome by using purified intrinsic factor which does not react with vitamin B12 analogs as the binder in the immune-enzymatic assays (Klee, 2000). However, in patients with pernicious anaemia these assays should not be used because the intrinsic factor antibodies in the patient's serum interferes with the binding of B12 to the intrinsic factor used in the assay, generating spurious results.

For the analytical platform used it is essential to adopt reference ranges that are compatible with the chosen platform (Harrington, 2017). A standard reference material for serum B12 has been developed by the US National Institute for Standards and Technology (NIST) (SRM 3951). This SRM consists of three sera with differing vitamin B12 concen­trations: 100pg/mL; 200pg/mL; 450pg/mL.

Interpretive criteria

Interpretation of serum B12 concen­trations deemed as adequate or inadequate is challenging. The normal levels for serum vitamin B12 concen­trations in healthy persons are affected by many factors unrelated to vitamin B12 status, some of which are not considered when interpreting the results. These factors may include age, ethnicity, pregnancy, use of certain medications, individual genetic variation, the assay method and any measure­ment errors associated with its use. All these factors may affect the sensitivity and specificity of serum B12 as a biomarker of B12 status and result in false-positive or false-negative classification of vitamin B12 defi­ciency.

Traditionally, adequate concentrations of biomarkers such as serum B12 are defined by a reference interval which covers 95% of the results obtained in a healthy, non-diseased population. In practice, however, it is difficult to compile a “true” healthy reference and few exist for serum B12.

Two methods are used to define whether a biomarker is deemed “inadequate”, both of which have been employed for serum B12 concen­trations. In some studies, a serum B12 concen­tration is considered inadequate when it falls below a statistically defined lower reference limit, often based on the 2.5th percentile value from an “apparently healthy” reference distribution compiled within a country. Hence, values for serum B12 that represent the 2.5th percentile will vary, depending on the local reference distribution applied.

Alternatively, the diagnostic criteria used to identify inadequate B12 status have been based on a known relationship between the serum B12 biomarker and low body stores, functional impairment, or clinical signs of deficiency. In such cases, the term “cutoff” should be used. In practice, the term “cutoff” is often used when the value for inadequate B12 has been based on a statistically defined reference limit. Again, cutoffs for serum B12 designated in this way can also vary, depending on both the local conditions and adverse outcome applied. Frequently, the cutoffs for serum B12 indicative of inadequacy have been set based on the relationship between serum B12 and serum methyl­malonic acid. The latter is a functional biomarker considered to be relatively specific and sensitive biomarker of B12 status. For more discussion on the Evaluation of nutritional assessment indices, see Chapter 1: Introduction Section 1.6.

Clearly, the reference limit or cutoff applied to designate inadequate vitamin B12 status may vary, leading to differences in the literature on reports for the prevalence estimates for vitamin B12 defi­ciency. Reports on the various methods used to interpret serum B12 concen­trations as adequate or inadequate are outlined below.

The United Kingdom compiled the mean, median, and critical lower and upper percentiles for serum B12 concen­trations by sex and age for children, young people aged 4 to 18y, and people aged >65y based on data from the U.K. National Diet and Nutrition Surveys (Gregory et al., 1995; Gregory et al., 2000; Finch et al., 1998). More recently, Kerr et al. (2009) have proposed age-specific 5th and 95th percentile reference ranges for serum B12 based on data from a representative sample of British children aged 4 to 18y from the National Diet and Nutritional Survey. Serum B12 concen­trations decreased significantly with age, with values for those consuming fortified breakfast cereals being significantly higher than non-consumers, a trend that was independent of age and gender. Up to aged 14 years, the 5th and 95th percentiles of serum B12 concen­trations were 242‑749pmol/L for children aged 4‑10 years (n = 317) and 172‑641pmol/L for children aged 10‑14 years (n = 263). For children from aged 15‑18 years, sex specific 5th and 95th percentiles were compiled. These values were 139‑452pmol/L for boys (n = 113) and 108‑502pmol/L for girls (n = 132) (Kerr et al., 2009). Serum vitamin B12 assays in this survey were conducted on a semiauto­mated analyzer using microparticle enzyme immuno­assay (MEIA) technology (Kerr et al., 2009).

In the United States, 95% reference intervals for serum B12 concen­trations for four age groups have been reported based on data from apparently healthy adults aged 20 > 70 years from NHANES 2011 to 2014 (Mineva et al., 2019). These are shown in Table 22b.1.

Table 22b.1 Serum vitamin B12 medians, central 95% reference intervals, by age group in US persons ≥20y of age, NHANES 2011‑2014. Serum vitamin B12 was measured by Roche E‑170 immuno­assay. Data from Mineva et al. (2019).

Age group	Serum vitamin B12, Median (95% CI,pmol/L)	Serum vitamin B12, 5th‑97.5th percentile
All	378 (372, 384)	158‑1140
20‑39	373 (364, 381)	167‑840
40‑59	365 (357, 378)	160‑1150
60‑69	398 (375, 420)	152‑1530
≥70	416 (401, 428)	137‑1720

Pregnant and lactating women were excluded from the data set. It is of interest that age, race/Hispanic origin, and vitamin B12 supple­ment use were significantly associated with serum B12 concen­trations in this survey. Non-Hispanic white persons had lower serum B12 concen­trations than non-Hispanic black persons. In this study, a “true cutoff” of >300pmol/L was used to define adequate vitamin B12 status (i.e., repletion) based on earlier research using serum methyl­malonic acid as a functional biomarker (see Section 22b.2.4).

Abildgaard et al. (2022) established 95% age-adjusted reference intervals for plasma B12 for children, adults, and elderly individuals in a Danish population. Blood samples from healthy individuals were collected and analyzed and routine clinical plasma B12 and methyl­malonic acid results extracted to establish reference intervals. The 95% reference intervals for plasma B12 based on blood samples and routine patient data from birth to aged >65 years categorized into five age groups are shown in Table 22b.2.

Table 22b.2 Proposed 95% Reference Intervals (RI) for plasma B12 based on blood samples and patient data from a Danish population. Data fromAbildgaard et al., 2022.

Age group (y)	95% RI, B12 (pmol/L)	% of B12 patient samples below/above 95% RI
0 ‑ <1	180‑1400	3.3%/0.0%
1 ‑ <12	260‑1200	5.5%/2.1%
12 ‑ <18	200‑800	2.8%/3.4%
18 ‑ <65	200‑600	3.0%/12.1%
≥65	200‑600	2.2%/19.2%

Total plasma B12 was measured with an auto­mated competitive chemi­lum­inescent immuno­assay. The highest plasma B12 concen­trations were found in infants, with levels gradually decreasing with age. A reference interval of 200‑600pmol/L was established for all adults, with a lower plasma B12 reference limit of 200pmol/L, irrespective of age. The authors emphasize that when plasma B12 is used alone to screen for B12 defi­ciency in surveys which include children and adolescents, then age-dependent reference limits should be applied.

Ethnicity has also been shown to be a factor that should be considered when interpreting serum or plasma B12 concen­trations (O’Logbon et al. 2022). Dietary habits alone do not appear to explain these ethnic differences. Instead, higher concen­trations of B12 binding proteins, most notably hapto­corrin, coupled with higher B12‑binding capacities in persons of Black ethnic origin, may be responsible for the higher serum B12 levels observed in Black persons compared to their White counterparts.

In response to this concern, Sobczynska-Malefora et al. (2023) have compiled age‑ and ethnicity-related reference intervals for serum vitamin B12 for a UK population. Unlike the traditional approach whereby data are obtained from individuals deemed “healthy”, these investigators applied an indirect method designed to overcome the ethical and resourcing difficulties experienced when sampling apparently healthy individuals. They accessed large datasets of laboratory results from an ethnically diverse South-east London patient population attending primary care clinics. All laboratories used an auto­mated Chemiluminescent Microparticle Assay (CMIS) with microparticles coated with intrinsic factor for the serum B12 assay. Pregnant women were not excluded.

These laboratory data were categorized into five ethnic groups: Asian or British Asian (Asian), Black or British Black (Black), White, Mixed, or Other ethnic groups. Reference intervals were established for children for four age groups from birth to 13y. For children from age 14y and adults, reference intervals are presented for eight age groups and by two ethic groups (Black and Asian/White) as shown in Table 22b.3.

Table 22b.3 Age and ethnicity related reference intervals (RIs) with confidence intervals (CIs) (pmol/L) determined by indirect calculations using mined patient test results. Data from Sobczynska-Malefora et al., 2023. The values are shown by age for children up to the age of 13 and for subjects >13y ‑ 49y by age and ethnicity. Data for subjects >50y are given in Sobczynska-Malefora et al. (2023).

Age groups (y)	0‑1	2‑5	6‑9	10‑13
Data size (N)	105	474	492	661
95 % CI Lower Limit of RI	142–178	246–308	221–270	168–207
RIs (pmol/L)	159–1025	276–1102	245–798	187–643
95 % CI Upper Limit of RI	971–1081	1042–1165	750–848	593–697
Age groups (y)	14–17	18–29	30–39	40-49
Data size Black (N)	267	1038	1534	1983
95 % CI Lower Limit of RI Black	152–191	146–175	150–186	161–198
RIs Black (pmol/L)	171–639	160–632	167–730	178–807
% Upper Limit of RI Black	585–697	579–691	665–801	739–882
Data size Asian/White (N)	297	2611	3614	3390
95 % CI Lower Limit of RI Asian/White	135–165	119–148	126–161	127–160
RIs Asian/White (pmol/L)	149–456	133–458	143–506	142–501
95 % CI Upper Limit of RI Asian/White	420–494	416–505	460–555	458–548

In this study, the lower reference limits for serum B12 concen­trations chosen to designate B12 defi­ciency for Black and Asian White people aged >13 years were: 166pmol/L and 134pmol/L, respectively. The authors acknowledge that these reference limits require validation using functional B12 biomarkers such as serum methyl­malonic acid and homo­cysteine.

As expected, the Black population had higher serum B12 concen­trations than the White population. However, surprisingly there were no differences in serum B12 levels between Asian and White individuals for most age groups in these UK adults. This finding may be associated with the treatment practice in the UK of recom­mending B12 supple­ments for Asians; lower serum B12 concen­trations in Asians compared to other ethnicities have been reported elsewhere (Quay et al., 2015; Devi et al., 2018). Children in all ethnic groups also had higher serum B12 concen­trations compared with adults, perhaps in part attributed to their higher intake of milk in early childhood and a more efficient hepatic storage of B12. Clearly, use of these age‑ and ethnicity-related reference intervals will assist in preventing under reporting of B12 defi­ciency in both children and the Black population in the UK.

The BOND project has set cutoffs for serum B12 based on the relationship between serum B12 and serum methyl­malonic acid, the latter a relatively specific and sensitive functional biomarker of B12 status (Section 22b.2.4). Recognising the importance of establishing normal levels for serum B12 during infancy and early childhood, the BOND group provide tentative age-specific ranges for adequacy from birth up to age 24 months based on data from healthy breastfeeding Norwegian infants and children. (Mineva et al., 2019, Table 22b.4).

Table 22b.4 Cutoffs for biomarkers of B12. Values are ranges. Data from (Allen et al., 2018).

	Serum B12 (pmol/L)		Serum holotrans- cobalamin (pmol/L)		Serum methyl- malonic acid (nmol/L)
	Deficient	Adequate/ Normal	Deficient	Adequate/ Normal	Deficient	Adequate/ Normal
Newborn cord blood		120-690		33-240		170-500
6 mo		121-520		12-90		140-220
12 mo		165-580		19-100		120-830
24 mo		183-260		29-110		120-300
Adults	Severe <75 Deficient 75 - <150 Depleted 150-221	>221	<35-40	40-150 or 40-200	>376 or >271

For adults, a single cutoff for adequacy represented by the upper limit (i.e., >221pmol/L) of the range said to be indicative of B12 depletion (i.e., 150‑221pmol/L) is also presented. In addition, the BOND group also present two serum B12 cutoffs said to be indicative of severe (<75pmol/L) and very severe (75‑<150pmol/L) B12 defi­ciency.

Clearly, although serum B12 is the most frequently used biomarker of B12 status, the marked variation in the literature for the reference limits and cutoffs said to be indicative of inadequate B12 status, make the diagnosis of B12 defi­ciency and the provision of prevalence estimates based on serum B12 alone challenging. For example, depending on the assay, serum B12 cutoffs for deficiency for adults may range from 120 to 180pmol/L, with 148pmol/L based on 3‑SDs from the mean of an adult reference range, said to be the most common cutoff for frank deficiency (IOM, 2000; Yetley et al., 2011).

Moderately low serum B12 concen­trations between 148 or 150 and 221pmol/L or 260pmol/L are often considered indicative of depleted or subclinical B12 defi­ciency. Subclinical mild B12 defi­ciency is much more common and generally asymptomatic with no hemato­logical or neuro­logical manifestations (Obeid et al., 2024). Such levels are difficult to interpret because they can also occur in association with megalo­blastic anemia produced by folate deficiency and in iron deficiency (Amos et al., 1994; Layrisse et al., 1959). Even with very low serum B12 concen­trations (i.e., below 75pmol/L), clinically diagnosable symptoms of B12 defi­ciency are said to be apparent in only about 50% of people (Stabler et al., 1990).

Even with total serum B12 concen­trations considered to be within the normal range, neuro­logical or in some cases hemato­logical symptoms related to B12 defi­ciency have been reported. This discrepancy may arise in part because the majority of B12 in serum is bound to the transport protein hapto­corrin which is not physio­logically active and thus unavailable for B12-dependent enzymatic reactions in cells (Green et al., 2017). In addition, factors unrelated to vitamin B12 status, and noted earlier, may have contributed to the seemingly normal or sometimes high levels of serum B12.

In an effort to resolve the discrepancies in the serum B12 cutoffs indicative of inadequacy, Bailey et al. (2013) used statistical modeling to identify a single change point at which the relation between plasma methyl­malonic acid and serum B12 changes slope to differentiate between inadequate and adequate B12 status. However, Instead of a single change point, they reported three slopes resulting in two change points and three subgroups. The data used were from the earlier US NHANES 1999‑2004 surveys. The first group considered at high risk for severe B12 defi­ciency had serum B12 <126pmol/L and the highest plasma methyl­malonic acid (281nmol/L). The second group deemed likely to have adequate B12 status had a serum B12 >287pmol/L and a median methyl­malonic acid concen­tration of 120nmol/L, while the third group was classified as indeterminate and difficult to interpret with an intermediate serum B12 (126‑287pmol/L).

Clearly, more research is warranted to establish appropriate cut‑offs for interpreting serum B12 concen­trations. In response to the concerns noted above that reduce the sensitivity and specificity of serum B12 and confound the diagnosis of B12 defi­ciency, many investigators now recom­mend measuring one functional B12 biomarker (serum methyl­malonic acid or serum homo­cysteine) if serum B12 concen­trations appear inadequate to ensure a more accurate assessment of B12 status. These additional tests are more sensitive and specific biomarkers of functional B12 defi­ciency and both may be elevated in persons with serum vitamin B12 levels even in the low to normal range (100‑300pmol/L) (Nexø et al., 1994; Stabler et al., 1996). They are described here in Section 22b.2.4 for methyl­malonic acid, and in the chapter on folate (Chapter 22a. Section 22a.2.2) for homocysteine.

Measure­ment of serum vitamin B12

Initially serum B12 was measured by microbiological methods, using bacteria with relatively specific requirements for vitamin B12 for growth. Examples include both Lactobacillus delbrueckii subsp. lactis (ATCC 4797), also known as L. leichmannii, or Euglena gracilis (Herbert et al., 1984). The microbiological assay has several limitations. For example, bacteriostatic substances in the blood, including antibiotics or cancer chemotherapeutic agents, inhibit the growth of the microorganism and interfere with the assay, producing misleadingly low serum vitamin B12 concen­trations. The main disadvantage of the microbiological method, however, is its low specificity as it measures a variety of cobalamin analogs that are not necessarily biologically active (Sobczynska-Malefora et al., 2021).

Competitive-binding assays that use radioisotope dilution methods for detection are simpler and less time consuming than the microbiological assays. Many of the available competitive-binding assays methods have been auto­mated (Chan, 1996). and unlike the microbiological assays, they are not affected by antibiotics or cancer chemotherapeutic agents. The coefficient of variation among the six laboratories for serum vitamin B12 analyzed by auto­mated competitive-binding assays has been reported to vary from 4.4% to 10.0% (Klee, 2000). However, the method requires the addition of radioactive cyanocobalamin to compete with vitamin B12 in the serum for the binding sites on an added cobalamin-binding protein. Purified hog intrinsic factor is now often used as the binding protein. Alkaline conditions are used to disrupt the vitamin B12 from the binding proteins, after which the vitamin B12 is generally converted to cyanocobalamin by potassium cyanide, prior to measure­ment (Klee, 2000). However, the assay is falling into disuse with possible discontinuation of the kits (Allen et al., 2018).

More recently, clinical laboratories are using high throughput auto­mated competitive binding chemiluminescence assays (CBLA), using purified intrinsic factor as a reagent to measure total vitamin B12 after its release from endogenous binding proteins. For this method both samples and comparators should be protected from light during collection and separation. Caution should be observed when this method is used to interpret results in patients with pernicious anemia. In these patients some of the competitive binding chemiluminescence assays can be influenced by the presence of interfering anti‑IF antibodies, thereby providing spuriously elevated serum cobalamin concen­trations, as noted earlier.

Serum B12 can also be determined using B12‑antibodies and B12 enzyme-linked immuno­sorbent assay (ELISA) kits. Various kits are available. Although these could potentially provide an alternative to IF‑based assays, they have not been completely verified (Sobczynska-Malefora et al., 2021).

Both serum and plasma samples in which EDTA has been used as the anticoagulant can be used for all the B12 analytical methods. However, use of lithium heparin should be avoided, as this anticoagulant may produce gelatinous serum and elevated B12 concen­trations. Blood samples for analysis of serum B12 should be protected from light during collection, and whole blood separated as soon as possible as delays in centrifugation and separation can lead to an initial rise in serum B12 concen­trations (Allen et al, 2018).

22b.2.2 Serum holotrans­cobalamin

Serum holotrans­cobalamin (holoTC) is the physio­logically active metabolite of vitamin B12 that delivers the vitamin to all DNA synthesizing cells, as discussed earlier. Only 20‑30% of the total vitamin B12 in serum is bound to the trans­cobalamin protein. Holo‑TC accounts for 5‑20% of total trans­cobalamin which is made in the ileal entero­cytes from intra­cellularly synthesized trans­cobalamin and absorbed vitamin B12.

Serum holoTC, like serum B12, reflects a broad range of intakes and status (Yetley et al., 2011), but is more sensitive and specific than serum B12. For example, because holoTC has a short half-life, concen­trations quickly fall below normal after vitamin B12 absorption ceases. Hence, low holoTC concen­trations in serum are often considered the earliest indicator of negative vitamin B12 balance. With increasing intakes of vitamin B12, concen­trations of holoTC rise continuously at least until vitamin B12 saturates the transport proteins. The holoTC assay is unaffected by interference from high-titre intrinsic factor antibody levels. Moreover, serum holoTC is not subject to the 25‑30% fall in pregnancy that is bserved with serum B12. At postpartum, serum B12 rises substantially. Instead, serum holoTC increases during pregnancy, a trend that continues to 6 weeks post-partum. Thus, in pregnancy serum holoTC offers a diagnostic advantage over total serum B12 (Varsi et al., 2018). Serum holoTC is also a useful test for identifying patients who may suffer from suspected deficiencies in the two vitamin B12 binding proteins, trans­cobalamin and hapto­corrin.

Serum holoTC is a component of the CobaSorb test now used as a surrogate for the dual isotope Schilling test (Nexo & Hoffman-Lucke, 2011) (see Section 22b.2.7). In the past the Schilling test was the gold standard method to evaluate the functional capacity of the ileal IF‑B12 receptor and required the use of cobalt-labeled B12. However, the Schilling test is no longer available worldwide due to safety factors related to use of radioactive B12 and is now considered obsolete.

Factors affecting serum holotrans­cobalamin

Age, sex, and race and their effects on serum holoTC concen­trations appear limited. Only a few reports are available with inconsistent results (Nexo & Hoffman-Lucke, 2011). Hence, currently data are insufficient to take into account the slight differences in serum holoTC concen­trations by age, sex, and possibly race for adults.

Pregnancy does not affect serum holoTC, levels increasing from 18 weeks gestation to 6 weeks postpartum. This trend contrasts with the decline in serum B12 observed during pregnancy (Varsi et al., 2018). The latter results from a decrease in the synthesis of hapto­corrin and possibly hemodilution. Conse­quently, assay of serum holoTC rather than serum B12 is preferable in studies of pregnant women.

Inherited defects such as trans­cobalamin deficiency is associated with unmeasurable levels of holoTC in serum. This genetic defect is associated with severe B12 defi­ciency because lack of the trancobalamin protein prevents B12 binding and formation of holoTC. Early treatment, however, can lead to a good clinical outcome.

Presence of the rare minor allele re35838‑82 (p.R215W) in the trans­cobalamin gene is also associated with unmeasurable levels of holoTC in serum. In this defect, there is proportionately more B12 bound to serum hapto­corrin, the physio­logically inactive B12 transport protein. This genetic defect is more common in South Asians and those of African origin, and unlike transcobalmin deficiency, is associated with other B12 biomarkers in the normal range and no clinical symptoms.

Transcobalamin receptor (TCblR/CD320) polymorphisms may also impact on serum holoTC concen­trations. In such cases, proportionately more B12 is bound to holoTC leading to elevated serum holoTC levels. Therefore, in such rare conditions (present in 5% of older adults), holoTC might not be a marker of "true" intracellular B12 (Sobczynska-Malefora et al., 2021).

Disease conditions affecting serum holoTC levels include renal patients with impaired kidney function. With decreasing kidney function, levels of serum holoTC rise to ensure sufficient holoTC is delivered into cells. Therefore, in patients with renal insufficiency (assessed by measure­ment of serum creatinine), serum holoTC concen­trations may appear normal, and are no longer a useful biomarker to predict B12 status (Herrmann et al., 2005). Other conditions that might elevate serum holoTC concen­trations include liver diseases and possibly the devel­op­ment of autoantibodies against trans­cobalamin.

High folic acid intake has been associated with reduced serum holoTC concen­trations. This adverse effect on B12 status has only been implicated with synthetic folic acid and not with natural folate from food sources (i.e., methyl­tetra­hydrofolate) (Selhub et al., 2022). However, in countries with mandatory folic acid fortification without B12 fortification, a high-folate-low vitamin B12 interaction may be cause for concern (Sobczynska-Malefora et al., 2021).

Dietary B12 intakes affect serum holoTC concen­trations with, as expected, lower levels in vegans than omnivores due to their lower B12 intake. Serum holoTC is a better biomarker of recent B12 intake than serum B12 (or methymalonic acid or tHcy) because serum holoTC concen­trations increase much faster after ingestion of B12 from a meal (i.e., 6 hours) or crystal­line B12 from supple­ments.

Interpretive criteria

Available serum holoTC assay methods, unlike serum B12, give similar values according to a European comparison study (Morkbak et al., 2005). However, to date there is insufficient data to account for the slight differences in serum holoTC concen­trations by age, sex, and possibly race for adults. Furthermore, uncertainty exists about the appropriate serum holoTC cutoff to diagnose B12 defi­ciency. For adults, the cutoff for serum holoTC, based on the relation of holoTC to serum methyl­malonic acid, is <35‑40pmol/L. according to the BOND group. They provide no cutoff for deficiency for children. However, for early childhood, the BOND group set tentative normal reference ranges for four age groups from birth to 24 months of age; see Table 22b.4 (Allen et al., 2018).

Currently the BOND group have set a tentative range indicative of adequate/normal B12 status of 40‑150 or 40‑200pmol/L for healthy adults, irrespective of age and the assay used.

A summary of reference intervals for serum holoTC using the most common assays are shown in Table 22b.5.

Table 22b.5 Selected reference intervals for holotrans­cobalamin (holoTC) with the use of the most common assays for holoTC measure­ment. From Nexo and Hoffmann‑Lucke (2011)
¹ Radioimmuno­assay: precipitation of trans­cobalamin on antibodycoated beads and measure­ment of vitamin B12 with the use of an isotope dilution method.
² Microbiology: precipitation of trans­cobalamin by antibodycoated beads and measure­ment of trapped vitamin B12 with the use of a microbiological method. ELISA: measure­ment of holoTC with the use of ELISA after removal of apotrans­cobalamin with vitamin B12–coated beads.
³ ELISA, enzyme-linked immuno­sorbent assay.
⁴ Direct: measure­ment of holoTC by ELISA, with the use of an antibody that is specific for holoTC.

Method	Sample (µL)	Reference range (pmol/L)	n	Reference
^1 Radio-immuno­assay	400	24–160 37‑170	105 303	Ulleland et al. (2002) Loikas et al. (2003)
^2 Microbiology	<150	42–160	500	Refsum et al. (2006)
^3 ELISA	100	40‑150	137	Nexo et al. (2002)
^4 Direct	200	19–130 36‑220	292 276	Brady et al.(2008) Aarsetøy et al. (2008)

Based on these data, a reference interval of 40‑200pmol/L is often used as indicative of the normal range for healthy adults. Nevertheless, where possible, it is preferable to establish a local reference interval for interpreting results in population-based studies and in clinical practice.

Laboratories nowadays use a diagnostic strategy for B12 defi­ciency that involves several bio­markers with serum holoTC or B12 as the initial test, followed by serum methyl­malonic acid or homo­cysteine as the second line test. See Section 22b.8 on multiple biomarkers for more discussion.

Measure­ment of serum holo trans­cobalamin

Several methods are available for measuring serum holoTC. They include a radioimmuno­assay (RIA) method based on monoclonal antibodies against trans­cobalamin (Vu et al., 1993) and an enzyme-linked immuno­sorbent assay procedure that is easier to use and measures holoTC directly without sample pretreatment (Brady et al., 2008). The holoTC‑RIA method has now been replaced by an assay that uses holoTC-specific monoclonal antibodies and yields results comparable to those of the holoTC‑RIA. An auto­mated assay has been developed for this method using the Abbott AxSYM immuno­assay analzyer (Brady et al., 2008). Tests have confirmed that results based on the holoTC‑RIA and the holoTC-enzyme-linked immuno­sorbent assay yield similar values (Morkbak et al., 2005).

Blood sampling and storage conditions have limited effect on serum or plasma holoTC concen­trations. Therefore, no special precautions are necessary for drawing blood samples and concen­trations are stable in serum stored at −20°C to −70°C for at least 16 months. Both serum and plasma can be used for the assay. Fasting and non-fasting blood samples can be used because holoTC concen­trations are unaffected by the intake of a normal diet (Nexo & Hoffman-Lucke, 2011).

22b.2.3 Deoxyuridine suppression test

This sensitive in vitro test has been used to diagnose early vitamin B12 (or folate) deficiency even in the absence of morphological changes in the blood. It was developed by Killman (1964) and Metz et al. (1968). Abnormal deoxyuridine suppression is the biochemical expression of disordered DNA meta­bolism. Although either lymphocytes or whole blood can be used for this test to detect past vitamin B12 or folate status, generally bone marrow cells are preferred because they measure the acute status (Colman 1981). The test is rarely used today because it requires bone marrow cultures, uses a radioactive label, is difficult to control (Carmel et al., 1996; Chanarin & Metz, 1997), and is not specific to vitamin B12 defi­ciency.

22b.2.4 Serum methyl­malonic acid

Methylmalonic acid (MMA) accumulates in the serum or plasma when the supply of vitamin B12 is reduced; it does not rise in folate deficiency (Savage et al., 1994). Hence, the measure­ment of methyl­malonic acid in serum or plasma, is a sensitive and specific marker of tissue vitamin B12 defi­ciency, reflecting stores rather than intake. Concen­trations are elevated in vitamin B12 defi­ciency, even in the absence of clinical signs or symptoms or of morphological changes in the blood. For example, in a cross-sectional study of 2919 Dutch elderly people with elevated serum homo­cysteine (tHcy) levels (i.e., >12 mol/L), the association of total serum B12 levels with serum MMA (and tHcy) was explored with restricted cubic splines (Figure 22b.5).

Figure 22b.5 Associations between Serum vitamin B12 and Serum MMA, adjusted for age, sex and creatinine levels. Modified from van Wijngaarden et al., 2017

Of the participants, 50% were women and all were >65 years (mean age of 74.1 years) with a mean BMI of 27.1. Even with concen­trations of serum total B12 below about 330pmol/L, a rise in both serum MMA (and tHcy) levels was observed, with an even steeper increase when total serum B12 levels fell to below 220pmol/L. These findings are in line with other studies, which have observed inflections, or change points with serum total B12 concen­trations between 200 and 500pmol/L. In the Dutch study, serum MMA (and serum tHcy) also rose when serum HoloTC concen­trations fell to less than about 100pmol/L (van Wijngaarden et al., 2017). An additional finding was the significant correlations reported between vitamin B12 intake and total serum B12, methyl­malonic acid and holoTC.

In some cases, elevated serum MMA (and tHcy) concen­trations have been reduced by vitamin B12 therapy, thus confirming vitamin B12 defi­ciency (Lindenbaum et al., 1990; Joosten et al., 1993). For example, Bolann et al. (2000) noted that in patients (n = 196) with elevated MMA concen­trations in both serum and urine (>376 nmo/L; >0.38 >mol/24 h, respectively) MMA concen­trations in serum were reduced by more than 50% after vitamin B12 supple­mentation.

Table 22b.6 Plasma levels of cobalamin and metabolites in macrobiotic and healthy control infants. MMA, methyl­malonic acid; Hcy, homo­cysteine. Data from Schneede et al. (1994).

	Macrobiotics (n = 41)	Controls (n = 50)	Significance
MMA (µmol/L)	1.44 (0.17‑12.15)	0.18 (0.06‑0.51)	p < 0.0001
Hcy (µmol/L)	13.5 (6.8‑26.8)	7.59 (5.3‑11.0)	p < 0.0014
Cobalamin (pmol/L)	141 (59‑340)	399 (194‑821)	p < 0.0001
Cysteine (µmol/L)	164 (119‑226)	184 (145‑233)	p = 0.0001
Methionine (µmol/L)	6.0 (1.5‑23.8)	6.9 (1.1‑41.9)	p = 0.44

The validity of serum or plasma methyl­malonic acid concen­trations as a sensitive and specific test for the diagnosis of nutritional vitamin B12 defi­ciency during infancy has also been explored. The results of a study of 41 Dutch infants on macrobiotic diets and 50 healthy omnivorous controls (Table 22b.6) showed markedly higher plasma methyl­malonic acid concen­trations in the infants on macrobiotic diets compared to the controls (Schneede et al., 1994). Moreover, plasma methyl­malonic acid concen­trations were inversely related to plasma vitamin B12 levels. Logistic regression showed that methyl­malonic acid followed by total homo­cysteine and cobalamin in plasma, in that order, were the strongest predictors of vitamin B12 defi­ciency in the macrobiotic group.

Several other investigators have compared the sensitivity of serum or plasma methyl­malonic acid in diagnosing vitamin B12 (or folate deficiency) with that of serum or plasma total homo­cysteine (see Section 22a.2.2 for more details). Results of a study by Savage et al. (1994) (Table 22b.7) emphasize the high sensitivity of both serum methyl­malonic acid concen­trations and total homo­cysteine levels for the diagnosis of functional vitamin B12 defi­ciency.

Table 22b.7 Performance of serum metabolite assays in patients with clinically defined cobalamin or folate deficiency. Data from Savage et al. (1994).

Serum level	Prevalence in cobalamin deficient patients (%)	Prevalence in folate deficient patients (%)
Elevated MMA	98	12
Elevated Hcy	96	91
Elevated MMA, Hcy normal	4	2
Elevated Hcy, MMA normal	1	80
Hcy and MMA both normal	0.2	7

Serum methyl­malonic acid concen­trations are not only a sensitive biomarker but also a very specific biomarker of functional vitamin B12 defi­ciency. Only a few conditions confound their use as outlined below.

Factors affecting serum methymalonic acid

Age tends to increase serum MMA concen­trations, particularly in persons >70y. This trend was reported in the US NHANES 2013‑2014 survey of adults >20 years, with levels about 60% higher in those aged >70y compared to the youngest age group (i.e., 20‑39 y) (Mineva et al., 2019). It is of interest that this age-related trend in serum MMA levels was apparent even among persons in the NHANES 2013‑2014 survey with serum total B12 concen­trations indicative of replete B12 status (i.e., >300pmol/L) and with normal renal function, as shown in Table 22b.8.

Table 22b.8 Serum MMA (mmol/L), sample size, medians, and central 95% reference intervals, by age group in US persons ≥20y, vitamin B12 replete and with normal renal function. NHANES 2011‑2014. Data from Mineva et al., 2019.

Age group (y)	Sample size	Median (95% CI)	2.5th‑97.5th percentile
All	5481	130 (127, 133)	67.2‑281
20‑39	2178	120 (118, 124)	63.5‑254
40‑59	1988	133 (128, 138)	70.5‑293
60‑69	824	143 (134, 149)	72.4‑281
≥70	491	161 (151, 167)	84.3‑317

This age-related trend is said to reflect the increasingly inadequate absorption of vitamin B12 that occurs with aging (Stabler & Allen, 2004), which arises from the gradual decrease in both gastric acidity and production of intrinsic factor (Asselt et al., 1998).

Ethnicity affects serum MMA. In both the US NHANES 2011‑ and 2013‑2014 surveys, non-Hispanic white persons had higher serum MMA (about 25%) concen­trations than non-Hispanic black persons, even after adjustment for other covariates (Mineva et al., 2019).

Pregnancy affects serum/plasma MAA. In a longitudinal study of healthy women during pregnancy, after correction for hemodilution, plasma MMA concen­trations showed a moderate increase throughout pregnancy from 20 weeks gestation (Murphy et al., 2007). The increase was interpreted to indicate a depletion of maternal intracellular vitamin B12 stores during pregnancy. Other reports, however, propose that an increase in plasma or serum MMA during pregnancy is normal and does not reflect a low vitamin B12 status and caution that serum MAA may not be a reliable biomarker of vitamin B12 status among pregnant women (Bae et al., 2015).

Impaired renal function, an indicator of chronic kidney disease, also leads to higher serum MMA concen­trations even from the early stages of renal impairment, as shown in (Figure 22b.6).

Figure 22b.6 Serum MMA geometric mean concen­trations by age group and renal function in US persons ≥20y of age, NHANES 2011‑2014. Black bars represent normal renal function and gray bars represent impaired renal function (stages 1‑5 chronic kidney disease). Error bars represent 95% CIs. Data from Mineva et al. (2019).

Data include geometric mean concen­trations of serum MMA by age group and renal function in persons ≥20y from US NHANES 2011‑2014 (Mineva et al., 2019). These findings emphasize that renal function should always be assessed when measure­ments of serum MMA are used, especially in the elderly.

Hypovolemia (decrease in the volume of circulating blood) is associated with rising levels of serum MMA.

Hypothyroidism affects serum MMA, increasing concen­trations.

Small-bowel bacterial over­growth may elevate serum MMA concen­trations due to the production of high levels of propionic acid (the precursor of MMA) by bacteria. When this condition is suspected, treatment with antibiotics should result in a decline in serum methyl­malonic acid concen­trations (Snow, 1999).

Inherited meta­bolic defects that elevate MMA concen­trations in serum and urine are a group termed "isolated methyl­malonic acidemia". The latter can be caused by complete or partial deficiency of the enzyme methyl­malonyl‑CoA mutase, which requires vitamin B12 as a co-factor (5‑deoxy-adenosyl-cobalamin (Ado‑Cbl)) (Figure 22b.7).

Figure 22b.7 Conversion of methylmalonylcoenzyme A (CoA) to succinyl‑CoA via methylmalonyl CoA mutase (mutase).

Use of supple­ments of crystalline B12 reduce serum MMA concen­trations. In the US NHANUES 2011‑2014 survey lower serum MMA levels were reported in those participants reporting B12 supple­ment con­sump­tion in the past 30 days (yes or no), even after controlling for other covariates.

Fasting times may influence serum MMA concen­trations. In the US NHANES 2011‑2014 surveys, shorter fasting time was associated with about 10% higher serum MMA concen­trations (Mineva et al., 2019).

Interpretive criteria

Currently there is no consensus for the cutoff value for serum MMA acid concen­trations indicative of vitamin B12 defi­ciency. Two tentative “cutoffs” for deficiency in adults are presented in Table 22b.4 from the BOND group (Allen et al., 2018). These so called cutoffs are derived from serum MMA concen­trations compiled from US NHANES 2003‑2006 data in which the central 95% reference intervals (2.5th to 97.5th percentile) for serum MMA concen­trations were presented (Pfeiffer et al., 2013).

Bailey et al. (2013) argues that use of only one serum MAA cutoff point can result in those with a B12 status considered intermediate being misclassified into the sufficient and deficient groups. To resolve this uncertainty, these investigators examined the complex relation between serum B12 and MMA (adjusted age, glomerular filtration rate, and hours of fasting) based on data from 12,683 adult participants of the NHANES 1999‑2004 study. Based on statistical modeling, those participants with a serum B12 <126pmol/L had the highest serum MMA (ie., median 281nmol) whereas for those with a serum B12 >287pmol/L (who likely had adequate B12 status), serum MMA was low (i.e., 120nmol/L). They also identified an intermediate group (i.e., between those with severe B12 defi­ciency and those with optimal vitamin B12 status) who were defined by a serum MMA of 148nmol/L, although their vitamin B12 status was difficult to interpret.

A range of adequate/normal serum MAA concen­trations for infants and children up to 2y are also provided by the BOND group (See Table 22b.4). These values are from the same sample of well-nourished breastfed Norwegian infants used to set the age-specific adequate/normal values for serum B12. Data from the US NHANES 2011‑2014 survey have been used to establish 95% reference intervals for serum MMA concentrations by age for US adults >20y. Reference intervals for serum MMA based on the overall sample are available and also on a sub­pop­ulation with serum B12 values said to be indicative of B12 repletion (i.e., ≥300pmol/L) and normal renal function based on serum creatinine (See Table 22b.8). Pregnant and lacating women were excluded in both datasets.

Not surprisingly, these age-specific central 95% reference intervals established for healthy adults (i.e., with serum B12 >300pmol/L and normal renal function) are lower than the upper‑end MMA cutoffs in earlier reports whose participants were from "apparently healthy populations" but with an unknown clinical history. In these earlier reports, some older participants probably had impaired renal function that was not considered. Clearly, in studies designed to establish adequate/normal serum MAA adult values, it is advisable to evaluate renal function using serum creatinine, especially in older adults.

Note that if both serum methyl­malonic acid and homo­cysteine are analysed, a distinction can be made between vitamin B12 and folate deficiencies. Elevated levels of both metabolites are expected in vitamin B12 defi­ciency, whereas in folate deficiency, only increases in serum homo­cysteine concen­trations occur.

Measure­ment of serum methyl­malonic acid

The analytical procedures most frequently used for serum methyl­malonic acid involve liquid chrom­atography-tandem mass spec­troscopy (LC‑MS/MS). Advantages of these procedures include the small sample size needed, their high throughput, and good precision. A summary table of the characteristics of the differing LC‑MS/MS procedures available is presented in Jin et al. (2022). In some cases, the LC‑MS/MS procedure has been used to analyze MMA in serum, urine, or dried blood spots. The major drawshy;back of the LC‑MS/MS procedures is their cost, the availability and maintenance of the equipment, and expertise required (Allen et al., 2018).

A more cost-effective method requiring small volumes of reagents with high throughput has been developed for serum MAA using stable-isotope-dilution LC‑MS/MS. Application of this method will facilitate early identification of vitamin B12 deficiency among the elderly (Jin et al., 2022). Levels of MMA in serum are stable for several years, when samples are stored at −70°C.

22b.2.5 Urinary methyl­malonic acid excretion

Vitamin B12 serves as a cofactor for methylmalonyl CoA mutase (EC 5.4.99.2) as noted earlier.This enzyme acts in the conver­sion of methylmalonyl CoA to succinyl CoA (Figure 22b.7). Therefore, in vitamin B12 deficiency, methyl­malonic acid accumulates in the blood and is excreted by the kidneys in increased amounts in the urine, resulting in methyl­malonic aciduria (Norman & Morrison, 1993).

The biomarker methyl­malonic acid in urine is sensitive, detecting early functional vitamin B12 defi­ciency before the onset of macrocytic anemia, as noted for MMA in serum. Similarly, this test is also highly specific and is not affected by folate deficiency. Only rare congenital enzyme defects discussed earlier (i.e., methyl­malonic aciduria) interfere with the test.

In elderly patients with renal disease, methyl­malonic acid accumulates so levels in urine, like serum, are elevated leading to a decreased specificity for B12 defi­ciency. In such cases, MMA concen­trations in urine should be adjusted for creatinine excretion (Norman & Morrison, 1993) and the ratio of urinary MMA to creatinine (uMMA/C) applied to assess B12 status. Such an approach is recommended in view of the finding that uMMA/C is not dependent on renal function (Supakul et al., 2020). Normal levels for serum creatinine range from 0.6‑1.1mg/dL for women and 0.7‑1.3mg/dL for men(Allen et al., 2018).

Interpretive criteria and measure­ment of methyl­malonic acid in urine

The reduced specificity of plasma MMA during renal impairment has led to the devel­op­ment of the uMMA/C ratio in the diagnosis of B12 deficiency. A threshold for uMMA/C ratios of 1.45µmol/mmol has been established for adult patients with plasma B12 concen­trations in the low-normal range (i.e., 201‑350ng/L). This threshold has good diagnostic performance for B12 defi­ciency in patients with subnormal serum or plasma B12 concen­trations and renal impairment (Supakul et al., 2020).

Assay of urine for MMA avoids the need for blood collection, although excretion of urinary MMA increases after meals. Analysis of 24h urine samples avoids this problem, but collection of 24hr urine samples is hard to achieve. Instead, analyses of two casual urine samples to correct for variability in the MMA concen­tration in urine can be used.

Combined gas/ liquid - chromщatography mass-spec­trometry techniques (LC‑MS/MS or GC‑MS/MS) are used to measure MMA in urine, as discussed for serum MMA. They are sensitive and reliable (Norman & Morrison, 1993; Allen et al., 2018) but the procedure is technically difficult; the equipment is expensive, needs high maintenance and expertise. The test is best performed on a 24h urine sample.

Recently a method has been developed and validated in which MMA/creatinine can be assayed in fasting urine samples collected on filter paper. After eluting the dry urine samples from the filter paper with a solution containing internal standards followed by filtration, the MMA/creatinine ratios are measured by ultra-performance LC‑MS/MS. The dried urine samples are stable at 86 days at room temper­ature and can be used for screening B12 defi­ciency in the elderly (Boutin et al., 2022).

Methylmalonic acid in urine is stable at room temper­ature, provided concentrated hydro­chloric acid is added to the urine sample, and for several years, when urine specimens are frozen at −70°C.

22b.2.6 Red blood cell methyl­malonic acid

Unrecognized deficiency of vitamin B12 during early childhood can lead to long-term health outcomes in children including impaired cognitive function and devel­op­mental delay. Maternal B12 defi­ciency during pregnancy is the most common cause of vitamin B12 defi­ciency during early infancy, as both B12 fetal stores and vitamin B12 concen­trations in breast milk are low. During later infancy and early childhood inadequate intakes of B12 due to low intake of animal-source foods and possibly mal­absorp­tion, may play a role (Sobczynska-Malefora et al., 2023).

Methylmalonic acid is considered the most specific functional biomarker and a sensitive indicator for B12 defi­ciency, as noted earlier. In many countries, analysis of dried blood spots are used as a diagnostic tool for newborn screening. In many countries, however, even though dried blood spots are collected routinely from newborns, the blood spots are not routinely screened for B12 status. Therefore, develo­ping a sensitive assay of MMA in dried blood spots through a newborn screening programme has the potential to facilitate the timely diagnosis and treatment of neonatal B12 defi­ciency. Such a method in dried blood spots could also facilitate testing B12 status in large-scale surveys and in remote populations (Schroder et al., 2014). Details of the assay of MMA in dried bloods spots and their interpretation are given below.

Interpretive criteria and measure­ment of methyl­malonic acid in dried blood spots

A new highly sensitive procedure based on stable isotope dilution with liquid chrom­atography-tandem mass spec­trometry (LC‑MS/MS) has been used to establish a reference interval for MMA in dried blood spots for healthy term newborn infants. Details of the method are available in Schroder et al.(2014). The mean dried blood spot MMA concen­tration quantified from 160 newborns was 16.8pmol/L (95th CI:15.9‑17.6pmol/L per 8‑mmspot), with 95% reference intervals (2.5th to 97.5th percentile) for MMA of 9.89 to 29.3pmol/8‑mm blood spot (0.450 to 1.33µmol/L whole blood). Concen­trations of MMA above the upper limit of this reference interval indicate an increased risk of B12 defi­ciency. The establishment of this neonatal MMA reference interval in dried blood spots will assist in the timely diagnosis and treatment of neonatal B12 defi­ciency through newborn screening programmes. To date, comparable reference intervals for MMA in dried blood spots of adults have not been established.

Concen­trations of MMA in dried blood spots increase with storage at room temper­ature so care must be taken to store dried blood spots below −80°C prior to analysis. Nevertheless, because storage at room temper­ature for at least one week has no effect on the dried blood spot MMA concen­trations, they can be shipped at room temper­ature (Schroder et al., 2016).

22b.2.7 Measure­ment of vitamin B12 absorption

Once vitamin B12 defi­ciency has been diagnosed, it is often of interest to establish if mal­absorp­tion is the cause. Originally, the Schilling test was used for this purpose, but it is no longer used in view of the safety factors related to the use of radioactive B12 (⁵⁷Co‑cyanocobalamin) (Sobczynska-Malefora et al., 2021). Instead, alternative tests are now employed that measure either the change in serum holoTC (CobaSorb) concen­trations or the absorption of microbially produced cobalamin labeled with ¹⁴Carbon or ¹³Carbon. However, the tests employing both ¹⁴Carbon or ¹³Carbon labeled cobalamin require more validation to diagnose B12 mal­absorp­tion in clinical patients (Miller & Green, 2020).

CobaSorb test depends on the fact that newly absorbed B12 circulates in the plasma as holoTC. The test is used to clarify whether the B12 defi­ciency diagnosed in a person is caused by an inability to absorb the vitamin. It is important that the test is performed before participants are treated for B12 deficiency in order to detect an increase in serum holoTC. The test is based on the analysis of serum holoTC prior to and following an oral dose of unlabeled B12. For the test, first a blood sample is taken to measure plasma holoTC, after which a dose of 9µg of crystalline B12 in water is given orally at 6h intervals over a 24h period. The next day, a second blood sample is taken, so that the change in the amount of B12 bound to serum holoTC can be determined. A small or no increase in holoTC is indicative of mal­absorp­tion. The test measures relative absorption and does not provide a quantitative estimate of B12 bioavailability Brito et al., 2018). Both the limited use of holoTC assays and lack of interpretative knowledge in many clinical settings have prevented widespread use of the CobaSorb test.

¹⁴C-labeled B12 absorption test uses carbon‑14-labeled vitamin B12 (¹⁴C-B12) produced by growing a genetically modified strain of Salmonella enterica in a medium containing ¹⁴C dimethy­benz­imidazole. After oral ingestion of ¹⁴C‑B12, enrichment of ¹⁴C in blood, urine, and stool samples can be measured using accelerator mass spectrometry. This approach has been used to assess absorption and bioavail­ability of B12 from aqueous solutions as well as some endogenously enriched foods. Exposure to radioactivity is negligible because the enrichment levels of ¹⁴C‑B12 can be measured in very small amounts (microliters) of the biological samples (Carkeet et al., 2006).

¹³C-labeled B12 absorption test is a nonradioactive method for determining vitamin B12 absorption and bioavailability. ¹³C‑cyano­cobalamin is a safe stable isotope-labeled vitamin B12 also biosyn­thesized from Salmonella enterica (Devi et al., 2020). For this test, a baseline blood sample followed by serial blood samples between 5 and 7h post oral administration of the ¹³C‑cyano­cobalamin are required. After processing, enrichment of plasma concen­trations of ¹³C‑cyano­cobalamin is assessed at 5 and 7h post dose by ultra-HPLC-MS. The kinetics of the plasma appearance of ¹³C‑cyano­cobalamin in a 2‑compart­ment model provide an index of oral absorption and bioavailability of B12. Both the ¹⁴C and ¹³C labeled absorption tests show promise but have not yet been developed for clinical tests, requiring further testing and validation (Brito et al., 2018).

Alternative procedures that can assist in the diagnosis of mal­absorp­tion from classic pernicious anemia arising from an autoimmune disease or from long-term atrophy or atrophic gastritis, include a combination of anti-intrinsic factor antibodies, and gastrin and pepsinogen A and C in serum (Green et al., 2017). These tests can all be performed using serologic assays, and if used in combination they can identify most cases of mal­absorp­tion from pernicious anemia (Lindgren et al., 1998).

22b.2.8 Multiple biomarkers

Defects in the hematopoietic system do not generally occur until the final stage of B12 defi­ciency. At this stage, the hemato­logical defects are indistinguishable from those of folate deficiency and include megalo­blastic anemia, characterized by macro-ovalocytic erythrocytes, abnormal red cell indices, and a low hemoglobin concen­tration (see Chapter 17, Sections 17.2.7 and 17.4). However, identifying impaired vitamin B12 status in the earlier stages is more challenging, but can be enhanced by using one or more plasma biomarkers of B12 status. No single biochemical biomarker is suitable for diagnosing B12 defi­ciency in all persons, and the use of several biomarkers is preferred. For a useful summary table comparing the relative strengths and weaknesses of the B12 biomarkers, see Allen et al. (2018). Algorithms based either on a selection of biomarkers measured sequentially or, multiple biomarkers analyzed simultaneously and then combined into a single diagnostic indicator, are both approaches that are used to circumvent the inherent limitations of each individual biomarker.

Algorithms based on the selection of sequential assays are used by several laboratories to assess B12 status. For example, initially, serum total B12 or serum holoTC can be measured, and if the concen­tration is deemed low (i.e., serum B12<148pmol/L or serum holoTC<35pmol/L), then B12 defi­ciency is diagnosed.

Figure 22b.8 Biomarkers indicating B12 defi­ciency in Denmark (Nexo & Parkner, 2024). Total serum B12 and MMA are included in the figure. HoloTC may replace total B12 as a first line test, and homo­cysteine may be used instead of MMA, notably in folate replete populations. * For patients with an eGFR below the reference interval (60mL/min), an elevated serum MMA concentration may be caused by the impaired kidney function rather than B12 deficiency. B12 represents total plasma B12; MMA, methyl­malonic acid; eGFR, estimated glomerular filtration rate.

However, if the concen­tration of either serum B12 or holoB12 is in the intermediate range between inadequate and adequate B12 status (i.e., between 148 and 250pmol/L for serum total B12 or between 35 and 50pmol/L for serum holoTC), then a functional biomarker such as serum methymalonic acid or serum homo­cysteine is measured as the second-line test. If either serum methyl­malonic acid or homo­cysteine is considered elevated, then B12 defi­ciency is diagnosed.

As an example, algorithms employed in Denmark are presented in (Figure 22b.8). Slightly different procedures are used in The Netherlands. For both algorithms, cutoffs indicative of B12 defi­ciency are selected, generally from the lower limits of the local reference intervals.

Table 22b.9 presents commonly seen vitamin B12 biomarker patterns in selected clinical scenarios (Sobczynska-Malefora et al., 2021).

Table 22b.9 The commonly seen vitamin B12 marker patterns in selected clinical scenarios. Cbl: cobalamin; holoTC: holotrans­cobalamin; MMA: methyl­malonic acid; tHcy: total plasma homo­cysteine. Modified from Sobczynska-Malefora et al., 2021.

Serum holoTC	Serum total B12	Plasma MMA	Plasma tHcy	Possible diagnosis
within ref. range	within ref. range	elevated	elevated	Suboptimal B12 status
decreased	decreased	within ref. range or elevated	elevated	Mild B12 deficiency, on antibiotics
within ref. range	within ref. range	elevated	within ref. range	Bacterial over­growth, B12 replete
very very low	very very low	highly elevated	highly elevated	Pernicious anemia
within ref. range or decreased	decreased	within ref. range or elevated	within ref. range	Pregnancy, B12 replete
decreased	very low	elevated	within ref. range or elevated	Pregnancy, B12 deficiency
within ref. range	within ref. range	within ref. range	elevated to very highly elevated	Mild to severe folate deficiency, B12 replete
within ref. range or elevated	very low	within ref. range	within ref. range	Haptocorin deficioency
very very low	within ref. range or decreased or elevated	highly elevated	highly elevated	Transcobalamin deficiency
within ref. range	within ref. range	very highly elevated	very highly elevated	CblC, D. F, J, disorder
within ref. range	within ref. range	very highly elevated	within ref. range	CblA, B disorder
within ref. range	within ref. range	within ref. range	very highly elevated	CblE, G disorder
within ref. range or decreased	within ref. range or decreased	within ref. range or elevated	very highly elevated	Nitrous oxide abuse
highly elevated	highly elevated	within ref. range	within ref. range	On Vitamin B12 injections
very highly elevated	very highly elevated	within ref. range	within ref. range	Chronic myeloid leukemia

Concurrent measure­ment of several biomarkers is an alternative to sequential approaches. In the original mathematical model four biomarkers are combined: serum holoTC, serum B12, serum methyl­malonic acid (MMA), serum homo­cysteine (Hcy) to yield a combined indicator termed the cB12 or 4cB12. The combined four-component algorithm adjusts for age and folate status, in view of the likely increase in serum MAA with age in the elderly, and the increase in serum Hcy in folate deficiency. 4cB12 can be calculated:

4cB12=log10[(holoTC.B12)/MMA.Hcy)]-(age factor)

The model 4cB12 is currently considered the best biomarker stratification for B12 status and yields one of five diagnoses: elevated B12, adequate B12, decreased B12, possible B12 deficient, and probably B12 deficient. The advantage of 4cB12 is that it is independent of local reference values and can be adjusted to correct for folate status and age. However, cost is a major disadvantage of the 4cB12 model as well as lack of availability of all four tests in routine clinical practice.

Suggested cutoffs and the interpretation of values of the combined 4cB12 indicator for use by researchers and clinicians and for epidemiologic purposes are shown in Table 22b.10.

Table 22b.10 Suggested cutoffs and interpretation of values of the combined indicator 4cB12 From Fedosov et al. (2015). B12, vitamin B12; Hcy, homo­cysteine; holoTC, holotrans­cobalamin; MMA, methyl­malonic acid.

	For researchers and clinicians		For epidemiologic purposes
Classification	Biological interpretation	Guidelines	Class	Guidelines
Elevated B12 >1.5	The pathogenesis of high B12 is not fully understood	Consider possible causes of high concen­trations such as liver disease or current or recent supple­mentation or treatment	B12 adequacy > −0.5	No action
B12 adequacy −0.5 to 1.5	Expected to accomplish all B12 status–dependent functions	No action advised
Low B12 −1.5 to −0.5	Potential subclinical manifestations of B12 deficiency, i.e., absence of hemato­logic changes, but subclinical neuro­logic impairment	Consider recom­mending oral supple­ments	Transitional B12 status −0.5 to −2.5	Need fortification
Possible B12 deficiency < −2.5	Potential manifestations of B12 deficiency	Potentially prescribe oral supple- ments, assess again in 3‑6mo	Low B12 status < −2.5	Need supple­ment- ation, rechecking, monitoring of status over time
Probable B12 deficiency < −2.5	It is possible to observe clinical manifestations of B12 deficiency. Clinical outcomes are needed to confirm potential clinical deficiency	Consider immediate treatment with i.m.injections, determine cause, consider particularly possible pernicious anemia

So far these cutoffs for 4cB12 have been validated against hemoglobin concen­trations and cognitive status (based on the Mini Mental State Examination), although additional validation studies using other cognitive and neuro­logical outcomes are needed. As yet, the 4cB12 model has not been validated for use in pregnancy but has proven very useful in the research setting (Fedosov et al., 2015).

More recently, in view of the drawbacks of 4cB12, refined models have been developed based on "missing" biomarkers. These alternative cB12 equations, missing one (3cB12) or two (2cB12) biomarkers of B12 status, have been compared with the established four-parameter 4cB12. Revised cut-points and guidelines for using this approach can be found in Fedosov et al. (2015) and Campos et al. (2020). The missing biomarker methods need further validation, especially for use in pregnancy and early childhood.