CARTPT Genetics: Appetite Suppression, Energy Balance

Short answer: CARTPT is a real appetite gene with a thin human genetics record. As of September 2026, no CARTPT variant has an established effect on body weight in people. The gene has never produced a genome-wide significant association with BMI or obesity, ClinVar lists no pathogenic CARTPT variant, and the one mutation with a convincing weight effect was found in a single family and will not appear in a consumer DNA file. If you are looking through your raw data for something actionable about appetite, CARTPT is not it.

That conclusion is worth more to you than a long article would be. Below is the evidence behind it, the specific studies, and what those studies do and do not support — including the one variant you are most likely to actually find in your file.

Why CARTPT looks like it should matter

CARTPT encodes cocaine- and amphetamine-regulated transcript (CART), a neuropeptide that suppresses feeding. The biology is genuinely well established. CART is expressed in the hypothalamus, co-localised with POMC neurons in the arcuate nucleus, and injecting the peptide reduces food intake in rodents. A 2021 review in Peptides covers twenty-five years of this work, including the long-unresolved question of which receptor CART actually signals through (Singh et al., 2021).

Two things about that literature matter for anyone reading their own DNA.

First, most of it is animal work. The candidate receptor GPR160, for example, was shown to be required for CART's appetite-suppressing action in rat brainstem (Haddock et al., 2021). That is a finding about rats.

Second, even the animal work is more modest than the gene's reputation suggests. Mice engineered to lack CART became obese only when fed a high-fat diet, not on regular chow. The authors of that study concluded directly that CART "may not be a major anorectic signal compared with proopiomelanocortin or leptin" (Asnicar et al., 2001). The gene was never positioned by its own researchers as a master switch for human body weight.

What the human genetics actually shows

The one family: Leu34Phe

There is a single genuinely striking human finding. Researchers sequenced the CART gene in 130 unrelated obese Italian children and adolescents and found one 10-year-old boy carrying a heterozygous G729C change, substituting phenylalanine for leucine at codon 34. The mutation cosegregated with severe obesity across three generations of his family and was absent from controls. Resting metabolic rate was lower than predicted in the boy (−14%) and in his mother (−16%) (del Giudice et al., 2001).

The mechanism was later worked out: Leu34Phe proCART is missorted inside the cell, poorly processed, and secreted through the wrong pathway, leaving carriers with severely reduced bioactive CART (Yanik et al., 2006). Adolescents in the same family also scored higher on anxiety and depression measures than controls, though that was a very small group (Miraglia del Giudice et al., 2006).

This is real science and it is why CARTPT appears on clinical obesity gene panels. But read the numbers again: one proband out of 130 screened, in one family. A private mutation like this is not on consumer genotyping arrays, and if you had it, you would likely have had severe obesity since early childhood and a clinical genetics workup rather than a mail-order kit.

The common variants: a long record of non-replication

Everything else in CARTPT has been studied as common polymorphism, and the results do not line up. Laid out chronologically, the pattern is hard to miss:

Note the direct contradiction: A1475G showed nothing in the Danish and UK cohorts but was overtransmitted to overweight children in the Italian trios. The other two positive signals — the Japanese promoter variant and the French rs7379701 — were each reported once, in one population, and the French authors themselves wrote that confirming the role of the CART gene "will require investigation and replication in further populations." That replication did not follow.

rs2239670: the variant you are most likely to look up

If any CARTPT variant appears in your raw data, it is probably rs2239670. Here is what it is, checked against dbSNP on 2 September 2026:

• Location: chromosome 5, position 71,719,676 (GRCh38)
• Type: intron variant — it does not change the CART protein
• Alleles: A/G, with A the minor allele
• Frequency: roughly 10–20% in European-ancestry reference panels (ALSPAC A=0.121, TWINSUK A=0.129, GoNL A=0.138, Estonian A=0.100), and notably higher in East Asian panels (Japanese TOMMO A=0.289, Korea1K A=0.337, Vietnamese A=0.264)
• dbSNP clinical significance: benign

One point deserves emphasis, because the internet gets it wrong constantly: rs2239670 is intronic, not a 3'UTR variant, and it is not the same thing as the ΔA1457 and A1475G polymorphisms from the older papers. Pages that describe rs2239670 as sitting in the 3' untranslated region and altering mRNA stability are describing something that is not this variant.

The one study that tested rs2239670 directly against obesity as its primary question — 300 people in Malaysia, genotypes 67.7% GG, 30.0% GA, 2.3% AA — found no difference between obese and non-obese subjects in genotype or allele distribution, and no difference in any anthropometric measurement, before or after adjusting for age, gender and ethnicity (Lisa et al., 2012).

Whether rs2239670 is in your file at all depends on your chip and its version. Vendors do not publish per-chip marker manifests, so nobody can tell you in advance — search your raw data file for the string rs2239670 and you will have your answer in a second.

The gene–diet interaction papers, read carefully

Search rs2239670 and you will find several papers reporting that it interacts with diet to affect BMI, fat mass, glucose, insulin and blood pressure. Four of them are real and are cited below. Before you act on them, note what they have in common:

• Mahmoudi-Nezhad 2020 — interactions with Healthy Eating Index and Diet Quality Index-International
• Mahmoudinezhad 2021 — interactions with DASH and Mediterranean diet scores
• Khodarahmi 2022a — mediation by glycemic index and glycemic load
• Khodarahmi 2022b — interactions with dietary antioxidant capacity

All four are drawn from the same cross-sectional sample of 288 obese Iranian adults aged 20–50. They are not four independent replications; they are one cohort analysed four times against different dietary indices. Each paper tests many diet scores against many metabolic outcomes, and the resulting P-values cluster just under 0.05 — the direct effect on metabolic syndrome in one of them was P=0.043. The second paper reports plainly that there were no significant differences in general characteristics or biochemical parameters across genotypes at all, apart from one insulin sensitivity index in women.

The authors are appropriately careful, closing with lines like "our findings need to be confirmed with large prospective studies." That confirmation has not arrived. Cross-sectional gene–diet interaction findings from a single cohort are hypothesis-generating, not a basis for changing what you eat.

Three checks you can repeat yourself

You do not have to take any of this on trust. Three public databases settle the question, and all three were queried for this article on 2 September 2026.

1. The GWAS Catalog has no obesity signal for CARTPT. Across every variant mapped to CARTPT in the NHGRI-EBI GWAS Catalog, there are 54 reported associations. Not one is for body mass index, obesity, body weight, waist circumference, adiposity or appetite. The only body-size trait in the entire list is a single association with body height. For comparison, running the identical query on FTO — the gene this article's topic cluster is built around — returns body mass index dozens of times over, alongside obesity and body weight, from roughly six times as many catalogued variants (251 against CARTPT's 44).

2. ClinVar has no pathogenic CARTPT variant. CARTPT has 38 ClinVar records, six of which carry a pathogenic or likely pathogenic classification. Every one of those six is a large chromosome 5 copy-number gain or loss spanning between 42 and 1,456 genes — whole stretches of a chromosome in which CARTPT simply happens to be included. There is no CARTPT-specific pathogenic variant. MC4R, by contrast, has 372 records and 196 pathogenic or likely pathogenic classifications.

3. CARTPT is on clinical obesity panels but does not deliver diagnoses. Two recent sequencing series make this concrete. A two-centre Italian study tested 284 children and adolescents on a 15-gene monogenic obesity panel that included CARTPT; seven probands received a molecular diagnosis, and all of them carried MC4R or NTRK2 variants (Morandi et al., 2024). A Turkish study sequenced 41 obesity-related genes including CARTPT in 116 patients; its pathogenic and likely pathogenic calls landed on UCP3, ADRB2, MC4R, POMC and NR0B2 (Anlas et al., 2025). CARTPT is screened routinely and comes back empty.

What CARTPT does associate with

The gene is not inert in human genetics — the associations are simply about other things. Among the CARTPT-mapped entries in the GWAS Catalog, the recurring traits are gut microbiome measurements (often jointly with breastfeeding duration), systolic and diastolic blood pressure, pulse pressure, and bone tissue density. The bone signal has independent support: a CART polymorphism was linked to bone remodeling markers in postmenopausal women (Guérardel et al., 2006).

None of these are strong enough or replicated enough to act on either. They are mentioned because they are what the data actually contains, which is a different thing from what the gene is marketed as.

What to do with your DNA file instead

If appetite regulation and body weight are what you care about, spend your attention where the evidence density is higher:

• MC4R is the single most informative gene here, and unlike CARTPT it has a large body of pathogenic variants with real clinical consequence — it is where both sequencing series above actually found their diagnoses.
• FTO carries the best-replicated common-variant association with BMI in the general population, which is why it anchors this topic cluster.
• Treat single-gene appetite claims sceptically in general. Body weight is highly polygenic. A single intronic SNP with a benign classification and no GWAS signal is not going to tell you how to eat.

And when a page tells you that a specific CARTPT genotype raises your BMI by a specific percentage, check whether it names a PMID. The numbers in circulation for this gene largely do not trace back to anything.

Ask My DNA reads your raw data and tells you when the honest answer is "this variant doesn't tell you much"

Frequently asked questions

Does the CARTPT gene affect my appetite? CART peptide is genuinely involved in appetite regulation — that part is solid biology. But no common CARTPT variant has been shown to affect appetite or body weight in humans in a way that has replicated across populations. Knowing your genotype at rs2239670 does not tell you anything reliable about your hunger.

I found rs2239670 in my raw data. What does my genotype mean? Practically speaking, nothing actionable. It is an intronic variant classified as benign in dbSNP, absent from the GWAS Catalog, and the one study that tested it head-on against obesity found no association. Roughly 10–20% of European-ancestry chromosomes and around 30% of East Asian chromosomes carry the A allele, so it is common rather than rare.

What about the studies showing rs2239670 interacts with diet? Those four papers all come from the same 288-person cross-sectional Iranian cohort, testing many dietary indices against many outcomes. Their authors explicitly call for confirmation in large prospective studies, which has not happened. Treat them as a hypothesis, not as dietary advice.

Is CARTPT tested in medical genetics? Yes — it appears on multi-gene monogenic obesity panels. In the two recent published series covering 400 patients between them, it did not produce a single diagnosis; those came from MC4R, NTRK2, UCP3, ADRB2, POMC and NR0B2.

What was the Leu34Phe mutation? A missense change found in one Italian boy and traced through three generations of his family, where it cosegregated with severe obesity and reduced resting metabolic rate. It works by preventing CART from being processed and secreted correctly. It is essentially a single-family finding and is not present on consumer genotyping arrays.

Why do so many websites give exact percentages for CARTPT variants? Because much of the content written about minor genes is generated rather than researched, and invented precision reads as authority. If a claim about a specific percentage change in BMI does not carry a PubMed identifier you can open, assume it does not exist. Every number in this article traces either to a linked study or to a named public database you can query yourself.

Educational Content Disclaimer

This article provides educational information about genetic variants and is not intended as medical advice. Always consult qualified healthcare providers for personalized medical guidance. Genetic information should be interpreted alongside medical history and professional assessment.