Adults over 50 with a low vitamin B12 blood result had more fractures in a large 2026 health-record study. That finding makes B12 status worth understanding, especially if you have a reason to be deficient. It does not show that taking extra B12 prevents broken bones. In randomized trials, B12 given with folic acid did not lower fracture rates for older participants overall. The useful question is whether you have an actual deficiency to identify and address, alongside an assessment of your wider fracture risk.
The 2026 Yang study (https://www.nature.com/articles/s41598-026-74010-8) compared people who had already undergone B12 testing. Its result is a clue, not a prescription. Reading it beside intervention studies makes the difference clear: a low level can travel with greater risk even when raising a level through supplements has not been shown to remove that risk.
What the new study measured
Researchers used electronic records from many health organizations and focused on adults at least 50 years old who had a B12 test. They defined deficiency as a blood concentration of 199 picograms per milliliter or lower. Their reference group had results from 300 to 900. They matched 115,735 people with deficiency to the same number with normal-range results using recorded characteristics that might affect the comparison. The 2026 paper's abstract (https://www.nature.com/articles/s41598-026-74010-8) reports these eligibility rules and numbers.
The team did not assign anyone to have low B12. It looked backward at test results and forward through later records for fractures. Its one-year landmark required participants to remain free of the studied outcomes for the first year after testing before follow-up began. This design reduces the chance that a fracture already underway led to a B12 test and then appeared to follow the test. It does not erase differences in diet, mobility, illness, or medical care that the records may have missed.
The main result was a fracture hazard ratio of 1.33 for the deficient group, with a 95 percent statistical interval from 1.29 to 1.38. In plain language, the rate of recorded fractures over follow-up was about one third higher among people with low B12 while they were alive and under observation. That is a relative comparison, not a claim that one third of deficient adults will fracture. The paper also reports a similar result for years one through five after testing, a hazard ratio of 1.37. Its broader follow-up did not meet one of the statistical model's assumptions, so that narrower analysis is an important check.
Researchers also looked at falls and osteoporosis-related diagnoses. Their abstract describes associations in the same direction. A fall and a bone-density diagnosis are different outcomes from a fracture. The study did not measure whether B12 changed a person's bone directly, and a health-record association cannot establish a biological pathway. It can identify a pattern large enough to deserve attention and a stronger test.
Why a large matched study still has limits
Matching helps compare people who look similar in the information available to researchers. It cannot match what was not recorded well. A person may have been tested for B12 because of numbness, anemia, poor diet, or another health problem. Some of those problems might also increase falls or signal frailty. People with a normal test may differ in ways that no computer model can fully balance. The 2026 Yang study (https://www.nature.com/articles/s41598-026-74010-8) itself says the association is not a causal or predictive fracture marker.
The result applies most directly to patients selected for B12 testing, not all adults over 50. A clinician's decision to order a test can select people with symptoms or risk factors. The study used a single measured level to define exposure, while B12 status and treatment could change later. Its abstract does not demonstrate that people who corrected a deficiency had fewer fractures. It also does not turn the chosen 199 cutoff into a universal threshold for bone safety.
This matters because a fracture can happen through several routes. Bones may be fragile, a person may fall, or a forceful injury may break a healthy bone. B12 deficiency can cause nerve problems and loss of balance, so a link with fractures could involve falling as well as bone condition. The new study's fracture outcome includes fractures in general. Without knowing each injury's circumstances, one should not translate its main number into a measure of osteoporosis alone.
An older Framingham study offers context. Researchers followed 1,002 adults with an average age of 75 and compared blood B vitamins, bone loss, and later hip fractures. Low B12 was associated with more hip fractures in one analysis, but that association weakened and was no longer statistically clear after accounting for bone density, vitamin D, and homocysteine. The 2008 McLean study (https://pubmed.ncbi.nlm.nih.gov/18364381/) illustrates how connected factors can shift an apparent B12 signal. Its smaller size and different population do not settle the new result either.
The test that matters more than a supplement slogan
Vitamin B12 helps the body make blood cells and maintain nerves. A deficiency matters even if the fracture question remains unanswered. It can develop when intake is low or absorption is impaired. Older adults, people who eat no animal foods, those with certain stomach or intestinal conditions, and some people using metformin or acid-suppressing medicines deserve particular attention to their status. The NIH Office of Dietary Supplements explains these risks in its vitamin B12 fact sheet.
Serum B12 is usually the first blood test, but a number near a laboratory cutoff may not tell the whole story. Laboratories use somewhat different thresholds. The NIH fact sheet explains that methylmalonic acid, often shortened to MMA, may help clarify a borderline result, although kidney problems can also raise MMA. Homocysteine can rise with low B12 but is less specific because folate and kidney function affect it too. A reader should not interpret a single result as a personal fracture forecast or start a high-dose regimen solely because it looks near the study's cutoff.
Signs that merit timely medical attention include unexplained numbness, tingling, trouble with balance, unusual tiredness, or anemia. Those symptoms have several possible causes, and some effects of a prolonged B12 deficiency may become hard to reverse. The important practical step is evaluation of the symptoms and the cause of a confirmed low level. Someone taking metformin or a long-term heartburn medicine should discuss testing with the prescribing clinician rather than stopping medicine without a plan.
The related topic of heartburn medicines and hip fractures also shows how easy it is to mistake an association for a proven mechanism. Some acid-suppressing medicines can affect B12 absorption from food, but that does not prove B12 explains every fracture association seen among their users.
What people can get from food
The usual adult recommended intake is 2.4 micrograms of B12 a day, according to the NIH vitamin B12 guidance. Fish, meat, dairy foods, and eggs contain B12. A serving of salmon or a cup of milk can contribute a meaningful amount. Plant foods do not naturally provide reliable B12, but some cereals and nutritional yeasts are fortified. Check the package rather than assuming every brand is fortified.
For someone who avoids animal foods, a dependable fortified-food or supplement plan can prevent dietary deficiency. For someone whose stomach cannot absorb B12 normally, eating more B12-rich food may not correct the problem. Those are reasons to take a confirmed low result seriously. They are separate from a claim that extra B12 above adequate intake builds stronger bones. Food first can be sensible when intake is the issue, while the treatment for an established deficiency depends on its cause and severity.
The new study compared blood levels, not menus. It cannot tell us whether salmon, eggs, fortified cereal, or a supplement changes fracture risk. A higher blood value is also not automatically a better target. The practical aim is adequate B12 and investigation of deficiency, rather than chasing an especially high number for bone protection.
What happened when researchers gave B12?
The B-PROOF randomized trial (https://pubmed.ncbi.nlm.nih.gov/25411293/) assigned 2,919 adults aged 65 or older who had elevated homocysteine to a daily tablet containing 500 micrograms of B12 plus 400 micrograms of folic acid, or a placebo, for about two years. Both groups also received 600 international units of vitamin D. Researchers counted new osteoporotic fractures, a more direct outcome than a blood marker. Because assignment was random and participants and investigators were blinded, this design can test the effect of that supplement combination more convincingly than a record study can.
The trial did not establish an overall fracture-prevention benefit. It also did not test B12 alone or specifically recruit people with a confirmed B12 deficiency. The researchers reported an exploratory signal in compliant people over 80, but subgroup findings can arise by chance and need a separate test. A published erratum is linked to the original trial; its detailed correction was not available during this review, so the article does not rely on the original paper's exact fracture counts or effect estimate.
A later B-PROOF follow-up (https://pubmed.ncbi.nlm.nih.gov/32800386/) again found no clear overall fracture benefit. Its reach was limited because fewer than half of the original trial participants took part in the later follow-up. A separate analysis of two other randomized trials (https://pubmed.ncbi.nlm.nih.gov/28574605/) also found no significant hip-fracture reduction with folic acid plus B12. Those trials were designed primarily for cardiovascular questions, so hip fractures were a secondary analysis. The results still argue against turning the new observational association into a blanket supplement recommendation.
Some B-complex products add large amounts of vitamin B6. In the two-trial analysis, the high-dose B6 groups showed a higher hip-fracture risk during extended follow-up. That does not prove every B6-containing product harms bones, but it is another reason to avoid assuming that more B vitamins must mean fewer fractures. A well-known older positive study of folate and B12 after stroke was later retracted (https://jamanetwork.com/journals/jama/fullarticle/200453). It should not be used as evidence for fracture prevention.
A sensible way to use the finding
Check B12 status when your diet, medicines, symptoms, or medical history give a reason to suspect deficiency. If a test is low, work out why and correct the deficiency with appropriate guidance. That protects against established problems caused by too little B12. Whether correction also reduces fractures has not been established by the studies discussed here.
Fracture risk needs its own assessment. The 2025 US Preventive Services Task Force recommendation advises osteoporosis screening for women 65 and older and for younger postmenopausal women at increased risk. A previous low-impact fracture, repeated falls, and other personal risks also warrant a conversation about bone health. A B12 result is one piece of a larger picture, not a replacement for that assessment.
The new cohort study gives a reason to investigate an overlooked nutrient in someone who may be low. The randomized evidence sets a boundary: taking B12 and folic acid has not been shown to prevent fractures in older adults overall. Adequate nutrition and proper care for a deficiency are worthwhile for their established benefits. The claim that extra B12 is a general bone-protection treatment still needs a convincing trial.
