Correlation between Active-B12 and Total B12

Active-B12 shows a reasonable correlation with the Total B12 assay.

In the chart below, the Total B12 result is plotted against the Active-B12 result of each patient. The important statistic is the r value (the correlation coefficient), in this case 0.74.

This tells us that there is a quite a high, positive relationship between Active-B12 and Total B12 (a perfect relationship would be 1). This relationship can vary depending on the types of patients the samples are taken from and the number of samples. In this case 468 is quite a high number so we can be reasonably confident of the relationship.

In various populations that have been investigated the r value has varied in the region 0.6 to 0.8 depending on the patients.

So why use Active-B12?

The first question that might be asked is “If Active-B12 and Total B12 are so highly related does it really matter which one is used?” This is a good question. To answer this it will help to look at the relationship in a bit more detail.

The chart below shows the same 468 patients as in the chart above. The red lines break down the patients by generally accepted Total B12 cut-offs into different categories:

Group A: Below 150pmol/L (200pg/ml). This level is most commonly used to signify vitamin B12 deficiency.

Group B: Between 151 and 300pmol/L. This covers a range that might be described as the Total B12 “grey-zone” where the true status of the patient is uncertain.

Group C: Above 301 pmol/L. At this level it would not be expected to find (much) true vitamin B12 deficiency.

We can see immediately that a large portion of the results are indeterminate by Total B12; in this case 314 patients out of 468, or 67% of results fall in the grey-zone.

To see how Active-B12 provides further utility, we then add another line, this time applying a 35pmol/L cut-off for Active-B12; below this level is generally indicative of Vitamin B12 deficiency and above this concentration, Vitamin B12 is probably replete.

This gives us 6 categories and allows us to see where Active-B12 can help to resolve indeterminate Total B12 results but more importantly, can show us where the Total B12result has been misleading.

It is evident that much of the agreement between Active-B12 and Total B12 is at the extremes (either very likely deficient or very likely not deficient). However, in the case of Group C where Total B12 levels are 301pmol/L and higher there are several abnormal Active B12 results. This small pecentage of patients will have been classified as ‘not deficient’ based on their high Total B12 results but in fact have very lowActive-B12 levels.

In the case of Group A where Total B12 levels are below 150pmol/L the majority ofActive-B12 levels are also low. We can see that there are some normal Active-B12samples in this group; it is most likely that these will come from samples where the low Total B12 level is actually due to low haptocorrin levels and therefore not true B12deficiency.

So what about the middle Group, 150-300pmol/L? There is a substantial proportion of patients (the true proportion will be between 30 and 50% depending on the population studied) in Group B with Total B12 levels above the 150pmol/L cut-off used to define deficiency but with low Active-B12 levels, indicating likely deficiency. If it is accepted that Active-B12 levels are a truer indication of B12 status then these patients would remain undetected and untreated under current scenarios for investigating B12deficiency that mainly rely on only the Total B12 test.

A resolving test…?

These results, and others like them, suggested that the initial clinical use of Active-B12could be as a resolving test for indeterminate Total B12 results, as in the algorithm below:

…or a screening test?

However as we can see, even below total serum B12 levels of 150pmol/L there are patients who will NOT be B12 deficient and therefore the ultimate use of the Active B12assay could be as a replacement for the total serum B12 test.

A recent publication (7) recommended that the Active-B12 assay should be the first-line test to check vitamin B12 status and showed that the Total B12 assay does not give any further information once the Active-B12 assay has been performed.

This improvement in diagnostic accuracy removes most of the need (and cost) for supplementary tests used with the Total B12 assay to help resolve indeterminate results. As we saw above, as much as 50% of the Total B12 testing results are indeterminate. With Active-B12, this falls to only 14%. This new algorithm is shown below.

We may also find that the accuracy of Active-B12 is high enough to perhaps use as a general screen in higher risk patients, the elderly being one obvious example.


Active-B12 in at-risk groups

As expected Active-B12 levels are low in patients with biochemical signs of vitamin B12deficiency (1). Low values have been reported in vegetarians (2,3) vegans (4) and populations with a low intake of vitamin B12 (5).

LV/LOV = lactovegetarians and lacto-ovo vegetarians

As expected, extremely low Active-B12 levels are found in untreated pernicious anaemia patients (6)

1. Obeid R et al., Clin Chem 2002 ;48 : 2064-5.
2. Herrmann W et al., Clin Chem Lab Med., 2003 ;41 :1478-88
3. Herrmann W et al., Am J Clin Nutr., 2003;78: 131-36.
4. Lloyd-Wright Z et al., Clin Chem., 2003; 49: 2076-8.
5. Miller JW et al., Clin Chem., 2006 ;52 : 278-85.
6. Brady J et al, Clin Chem 2008;54:567-73.
7. Valente E et al, Clin Chem 2011;57:6



Total B12 and holohaptocorrin levels can fall in normal pregnancy

Vitamin B12 deficiency, defined as low serum vitamin B12 concentration has been reported to occur in 10-28% of uncomplicated pregnancies (1). A more recent study (2) was undertaken to monitor vitamin B12 status throughout uncomplicated pregnancy in women with an adequate vitamin B12 intake. This showed that in the third trimester almost 35% of the participants had serum vitamin B12 concentrations less than 150pmol/L with no apparent consequences. The studied showed that the fall in total vitamin B12 levels was actually due to a fall in holohaptocorrin levels (see graph). The conclusion was that this fall reflects normal physiologic changes during pregnancy.

What does Active-B12 do during normal pregnancy

A later study was carried out in healthy pregnant Danish women (3) in which the investigators had the ability to measure Active B12 (holotranscobalamin).

In agreement with previous studies they found an almost 50% decline in vitamin B12 during late pregnancy (which reversed after birth) but that this was NOT reflected in the Active B12 (holotranscobalamin) levels.

The conclusions were that a) healthy Danish women do not seem to suffer from an inadequate supply of vitamin B12 and the drop in total B12 levels was due to a decline in holohaptocorrin and b) that Active B12 could be used as a reliable marker for vitamin B12 deficiency during pregnancy.


Why might this be important?

This finding may be important for two reasons:

1. The vitamin B12 status of the newborn depends on the vitamin B12 status of the mother so the baby may be born with inadequate vitamin B12 stores. This could then be compounded if the mother then breastfeeds. What might be observed is that the infant has a failure to thrive. Symptoms may be reversible upon vitamin B12 treatment but if treatment is delayed then permanent neurological damage may result. The earlier that true vitamin B12 deficiency can be diagnosed in the mother the earlier B12 deficiency in the baby can be prevented (4)

2. The link between low folate levels and neural tube defects has been established and women intending to become pregnant are advised to supplement with folic acid. There may also be a link between low vitamin B12 levels and risk of NTDs therefore it may be important to accurately know maternal vitamin B12 status (5). These studies tell us that total vitamin B12 levels are not reliable for this purpose.


1. Shojania AM Clin Perinatol 1984;11:433-59

2. Koebnick C et al Clin Chem 2002;48:928-33

3. Morkbak AL et al Haematologica 2007;92:1711-12

4. Honzik T et al Eur J Paediatric Neurol 2010; 14:488-95

5. Thompson MD et al Am J Clin Nutrition 2009;89:697S-701S.