The claim that fat holds more stem cells than bone marrow gets repeated constantly, almost always with no citation attached. This article puts the published figures for adipose tissue stem cell yield vs bone marrow in one place, each with its study and year, alongside the strongest counterargument in the literature.
TLDR: Published research puts stem and progenitor cells at up to 3 percent of the uncultured stromal vascular fraction from fat, 2,500-fold more than in bone marrow. One study measured 404,000 plus or minus 206,000 cells per milliliter of lipoaspirate, another found colony forming units at 1 in 32 fresh SVF cells, and marrow averaged 612 progenitors per cubic centimeter before concentration. A same-patient comparison, though, found no significant per-cell differences. Banking adipose tissue is not a treatment and does not guarantee eligibility, access, or clinical benefit.
Important Disclaimer: Save My Fat is a tissue preservation connector service, not a medical practice. It does not provide FDA-approved treatments or cures, and it does not guarantee eligibility, access, or clinical benefit. The counts here come from laboratory research and describe starting material, not patient results. Adipose tissue, the stromal vascular fraction, and adipose-derived cells are investigational and are not FDA approved. This article is offered for educational purposes only, and readers should consult their own licensed healthcare professionals before any decision about banking, treatment, or research participation.
Ask ten people in this industry how many stem cells per gram of fat a donor can expect, and you will get ten confident answers and no sources. The published record is narrower than the slogan, and in places less flattering to fat.
Five papers carry most of the weight. Two counted cells in fat, one counted progenitors in marrow, one compared isolation across sources, and one took both tissues from the same patients. That last study decides how the others read.
What researchers are actually counting
Yield sounds like one measurement. In the literature it is at least three. The first is total nucleated cells recovered from a volume of tissue. The second is the share of those cells that behave like stem or progenitor cells, usually estimated with a colony forming unit fibroblast assay. The third is progenitors per unit of volume, which depends on the first two.
So a paper reporting 404,000 cells per milliliter and one reporting 612 progenitors per cubic centimeter are not in conflict. They count different objects, in different donors, with different assays. Careful adipose derived stem cell yield research lines up the units first.
Adipose tissue stem cell yield vs bone marrow: the published numbers
The table collects the primary figures with the paper behind each one. Every entry is quoted from its source rather than recalculated, and where a study did not measure a tissue, the cell says so. Filling those blanks by inference is how yield myths begin.
| Measure | Adipose tissue | Bone marrow | Source |
|---|---|---|---|
| Stem and progenitor share of uncultured SVF | up to 3 percent of whole cells, called 2,500-fold more than marrow | 0.001 to 0.01 percent of about 6 million nucleated cells per mL | Baer and Geiger, 2012 |
| Cells per gram of raw tissue | 0.5 to 2.0 million cells, 1 to 10 percent stem cells, so 5,000 to 200,000 stem cells | not reported | Baer and Geiger, 2012 |
| Nucleated cells per mL of lipoaspirate | 404,000 plus or minus 206,000, in 18 donors | not measured | Aust, 2004 |
| Colony forming units in fresh SVF | 1 in 32 cells, about 3.1 percent | not measured | Mitchell, 2006 |
| Progenitors per cubic centimeter of marrow aspirate | not measured | 612 plus or minus 134 before concentration, 2,579 plus or minus 1,121 after | Hernigou, 2005 |
| Isolation success across donors | 100 percent | 100 percent | Kern, 2006 |
One pattern holds. Fat is the richer starting material per gram, and marrow needs concentration to reach a few thousand progenitors per cubic centimeter. That describes tissue, not patients.
Why these figures cannot simply be multiplied
It is tempting to multiply Aust’s 404,000 cells per milliliter by Mitchell’s 1 in 32 colony frequency and produce a tidy stem cell count for fat. That number would be arithmetic, not evidence, since the two studies used different donors, different processing, and different assays.
The same caution applies across tissues. Aust counted every nucleated cell recovered; Hernigou counted only progenitors that formed colonies, so putting the two together sets a total against a subset. Every stem cell concentration comparison between tissue sources meets this problem. Hernigou’s own numbers still stand: bone marrow mesenchymal stem cell yield averaged 612 progenitors per cubic centimeter and reached 2,579 after concentration.
The same-patient comparison that complicates the story
Most yield articles leave this part out. In 2003, De Ugarte and colleagues took adipose tissue and bone marrow from the same patients, processed both, and compared the results. Their finding, in their words, was that no significant differences were observed for yield of adherent stromal cells, growth kinetics, cell senescence, multi-lineage differentiation capacity, and gene transduction efficiency (De Ugarte and colleagues, Cells Tissues Organs, 2003).
That does not erase the frequency data. It reframes it. The adipose advantage is accessible tissue volume and stem cell frequency per gram of raw tissue, not per-cell superiority. Once both populations are in a dish, the same-patient data does not show one beating the other, and any page claiming otherwise is going past the evidence.
The literature is not unanimous. Kern and colleagues found colony frequency highest in adipose tissue, isolation succeeding in 100 percent of bone marrow and adipose samples against 63 percent for cord blood, and bone marrow MSCs showing the shortest culture period and lowest proliferation capacity of the sources tested. Each tissue wins some comparisons and loses others.
What a higher frequency actually changes
Frequency per gram matters most at the front of the process, before anything is cultured. A tissue with more progenitors per gram, accessible in larger volume, gives a laboratory more room to work with. That is a logistics argument, not biological superiority. Providers weighing the two routes can review the bone marrow comparison and the closer look at adipose versus marrow cells.
Technique sits upstream of every number a laboratory reports, which is why the harvest procedure steps are standardized rather than improvised. Donor variation is real too. Aust found a significant negative correlation between yield and body mass index, and none with age.
What yield does not tell you
A cell count describes starting material. It does not predict what happens in a patient, and the trial record shows why.
The completed MILES trial, NCT03818737, at Emory University compared corticosteroid, bone marrow concentrate, adipose SVF, and umbilical cord tissue MSCs in 475 participants with knee osteoarthritis, with results posted to the registry. The ADIPOA2 Phase 2b trial of culture expanded adipose MSCs, published in Annals of the Rheumatic Diseases in 2025, reported no significant improvement in pain and function against placebo.
Adipose-derived cells and SVF remain investigational, and no adipose-derived product is FDA approved. Readers can review the ClinicalTrials.gov listing of recruiting adipose-derived MSC studies and take what they find to a licensed physician. The evidence is preliminary and more studies are needed.
Frequently Asked Questions
How many mesenchymal stem cells are found per gram of adipose tissue versus bone marrow?
Baer and Geiger reported 0.5 to 2.0 million cells per gram of adipose tissue, with stem cells at 1 to 10 percent, or about 5,000 to 200,000 per gram. For marrow they cited 6 million nucleated cells per mL, with stem cells at 0.001 to 0.01 percent.
Why does adipose tissue yield more cells than bone marrow?
The published difference is frequency in raw tissue. Baer and Geiger put stem and progenitor cells at up to 3 percent of uncultured SVF, 2,500-fold more than in bone marrow, and fat can be accessed in larger volume. The advantage is the tissue, not a better cell.
Does a higher yield mean better clinical outcomes?
No. Nothing in the yield literature measures patient outcomes. The ADIPOA2 Phase 2b trial of adipose MSCs in knee osteoarthritis found no significant improvement in pain and function against placebo. Adipose-derived cells are investigational and not FDA approved, and outcomes cannot be predicted.
What published research supports these yield figures?
Baer and Geiger, Stem Cells International, 2012, for the 3 percent SVF share and the 2,500-fold comparison; Aust, Cytotherapy, 2004, for 404,000 plus or minus 206,000 cells per mL; Mitchell, Stem Cells, 2006, for the 1 in 32 colony frequency; Hernigou, Journal of Bone and Joint Surgery, 2005, for marrow progenitors; and Kern, Stem Cells, 2006, for isolation success.
Does yield affect how much tissue needs to be collected for banking?
It is one input among several. Reported yields vary widely between donors, and processing method changes the count independently of the tissue. A physician and the receiving laboratory decide what is collected, and banking adipose tissue for potential future use does not guarantee eligibility or access.
Do age and body mass index change how many cells are recovered?
Aust and colleagues reported a significant negative correlation between cell yield per milliliter of lipoaspirate and body mass index, and no significant correlation with donor age. That is one study in 18 donors, so treat it as a signal rather than a rule. Donor variation is wide.
Key Takeaways
The frequency gap in raw tissue is real and documented. Baer and Geiger put stem and progenitor cells at up to 3 percent of uncultured SVF, 2,500-fold more than in bone marrow. Aust measured 404,000 plus or minus 206,000 cells per milliliter of lipoaspirate, Mitchell found colony forming units at 1 in 32 SVF cells, and Hernigou measured 612 progenitors per cubic centimeter of marrow before concentration and 2,579 after.
Against that, De Ugarte’s same-patient comparison found no significant differences in adherent stromal cell yield, growth kinetics, senescence, multilineage differentiation, or transduction efficiency, and Kern found bone marrow MSCs had the shortest culture period and lowest proliferation capacity. The conclusion is narrow. Adipose tissue is the richer and more accessible starting material per gram, the cells are not demonstrably superior, and none of these numbers predicts a clinical result.
Save My Fat operates as a tissue preservation service, not a medical practice or treatment provider. Stem cell and regenerative medicine regulations vary by state, including specific informed-consent and disclosure requirements in Florida, Utah, and Nevada governing tissue and stem cell services. Banking adipose tissue does not connect patients to any treatment pathway, and any future use depends on FDA regulatory status, physician guidance, and the availability of approved or investigational pathways at that time.
Patients weighing adipose tissue banking for potential future use can review current fees, and providers with questions can contact the team.
Save My Fat partners with L2 Bio for laboratory processing and storage.
This article is for educational purposes only and does not constitute medical or legal advice. Legal and medical review including neurology and neurosurgery input is required before publication. Please consult your neurologist or neurosurgeon before making any decisions about banking, treatment, or research participation.
About the author: Oscar Tellez is the founder and CEO of Save My Fat. He holds a Bachelor of Science in Exercise Science and Health Promotion from Florida Atlantic University. He has spent more than a decade in the regenerative medicine industry across product distribution, laboratory and vendor relationships, and provider training. He is not a licensed clinician, and this article is educational, not medical advice.
Related guide: the stem cell science glossary.
