visceral fat loss is sustained long-term

There were 2 articles by the same Israeli research group that assessed the effects of lifestyle-induced visceral fat loss being associated with durable (5 and 10 years later) improvement in cardiometabolic health despite full weight gain: see visceral fat loss cardiovasc health Circ2026in dropbox, or DOI: 10.1161/CIRCULATIONAHA.125.079009)

    -- this blog will be followed by a subsequent blog noting improved brain architecture as well as cognitive performance through visceral fat loss

    -- given the potential profundity of their conclusions, i have included lots of data and graphs to demonstrate how they reached them

 

Details:

-- these researchers invited participants from the Israeli CENTRAL study (Effect of Weight-Loss Diet Strategies and Exercise on Dynamics of Body Fat Depots and Metabolic Rate) with 278 participants having a 10 year follow-up, and the DIRECT-PLUS study (Effects of Green-Mediterranean Diet via the Gut-Fat-Brain Axis) conducted from 2012 through 2014, with 294 participants and had 5 years of follow-up, both were randomized controlled trials where participants achieved considerable reductions in abdominal and ectopic fat, the latter includes intrahepatic fat and intrapancreatic fat. The researchers assessed post- intervention cardiometabolic profiles associated with these changes 5 and 10 years later as related to the composition of fat stores

    -- measures: MRI measurement of visceral adipose tissue (VAT), deep subcutaneous adipose tissue (SAT), superficial SAT, intrahepatic fat (IHF), and intra-pancreatic fat (IPF), along with measurements 10 years in the CENTRAL study and 5 years in the DIRECT-PLUS study after completion of these trials

    -- the inclusion criteria for these studies included individuals with abdominal obesity (waist circumference >102 centimeters for men and >88 centimeters for women) or dyslipidemia (serum triglycerides >150 mg/dL and HDL cholesterol <40 mg/dL for men and <50 mg/dL for women); all of the diets were designed for moderate, long-term weight loss with restricted consumption of trans fats and refined carbohydrates along with increased intake of vegetables

        -- since lunch was the primary meal for people in this region, the foods were prepared by the workplace cafeteria. For the CENTRAL study, the focus was a low-fat, Mediterranean, or low carbohydrate diets; for the DIRECT-PLUS trial the participants were randomized to one of 3 groups: healthy dietary guidelines only, Mediterranean diet, or green Mediterranean diet, all associated with physical activity

     

-- 381 eligible participants were involved, and 366 of them (96%) were available for follow-up

-- there were 5 MRI-assessed abdominal adipose tissue depots and ectopic fat depots (as per above: VAT, deep SAT, superficial VAT, IHF and IPF). these were measured at baseline, 18 months (the end of the studies), and 5 and 10 years postintervention (the interval depending on the study done)

 

-- baseline characteristics of the patients:

    -- mean age 50, 91% male

    -- the baseline and 5-and 10-year follow-ups were reported in the results section below, including the changes in the many comorbidities and other targets assessed

-- Main outcomes:

    -- the relationships between the various MRI-measured adipose tissue [visceral adipose tissue (VAT), deep subcutaneous adipose tissue (SAT), superficial SAT, intrahepatic fat (IHF), and intra-pancreatic fat (IPF)] at 18 months and again at 5 and 10 years after completion of the 2 trials above, as well as the many cardiometabolic variables noted below over these time periods

  

Results:

-- changes in fat distribution, information from 325 of the 381 participants in cm²:

        -- visceral adipose tissue (VAT)

            -- baseline: 148.6 cm²

            -- end of RCT (18 months later): 109.0 cm², with adjusted change during RCT -39.53, P<0.001

            -- after 5 and 10 years: 124.8 cm², with adjusted change from baseline of -22.05 cm²

        -- deep superficial adipose tissue (deep SAT):

            -- baseline: 242.1 cm²

            -- end of RCT: 185.1 cm² ,with adjusted change during RCT -60.65, P<0.001

            -- after 5 and 10 years: 211.5 cm², with adjusted change from baseline of - 32.44 cm², P<1.001

      

         -- superficial adipose tissue (SAT):

            -- baseline: 132.2 cm²

            -- end of RCT: 106.1 cm², with adjusted change during RCT -28.78, P<0.001

            -- after 5 and 10 years: 117.6 cm², with adjusted change from baseline of - 15.81 cm², P<0.001

 

-- Changes in fat proportion, in %:

        -- visceral adipose tissue (VAT):

            -- baseline: 28.9%

            -- end of RCT (18 months later): 27.5%, with adjusted change during RCT of -1.33, P<0.001

            -- after 5 and 10 years: 27.8%, with adjusted change from baseline of - 0.97%, P=0.002

        -- deep superficial adipose tissue (deep SAT):

            -- baseline: 46.1%

            -- end of RCT: 45.9% with adjusted change from baseline of -0.13, P=0.449

            -- after 5 and 10 years: 46.1%, with adjusted change from baseline of -0.15, P=0.617

        -- superficial adipose tissue (SAT):

            -- baseline: 25.1%

            -- end of RCT: 26.6%, with adjusted change during RCT of 1.44, P< 0.001

            -- after 5 and 10 years: 26.1%, with adjusted change from baseline of 1.08, P<0.001

        -- intrahepatic fat (IHF):

            -- baseline: 10.2%

            -- end of RCT: 6.4%, with adjusted change during RCT -3.99, P<0.001

            -- after 5 and 10 years: 9.6% with adjusted change from baseline of -0.71, P=0.100 (non-significant)

        -- intra-pancreatic fat (IPF):

            -- baseline: 14.6%

            -- end of RCT: 14.1%, with adjusted change during RCT -0.42, P=0.071

            -- after 5 and 10 years: 15.6%, with adjusted change from baseline of 1.14, P=0.002 (significant increase!!)

 

-- Changes in weight, in kilograms:

        -- baseline 91.7 kg

        -- end of RCT: 88.5 kg, with adjusted change during RCT -3.41, P<0.001

        -- after 5 and 10 years: 91.4 kg, with adjusted change from baseline of -0.27 kg, P=0.609 (no significant change from baseline)

 

-- Changes in BMI:

        -- baseline: 30.5

        -- end of RCT: 29.5, with adjusted change during RCT -1.14, P<0.001

        -- after 5 and 10 years: 30.6, with adjusted change from baseline of -0.03, P=0.848(no significant change from baseline)

 

-- Changes in waist circumference, in centimeters:

        -- baseline: 107.6 cm

        -- end of RCT: 102.4 cm, with adjusted change during RCT -5.23, P<0.001

        -- after 5 and 10 years: 103.5, with adjusted change from baseline of -4.13cm, P<0.001 (significant decrease)

 

-- Changes in systolic blood pressure, in mmHg:

        -- baseline: 127.0

        -- end of RCT: 127.0, with adjusted change during RCT -0.02, P=0.980

        -- after 5 and 10 years: 131.4, with adjusted change from baseline of 4.66, P<0.001 

 

-- Changes in diastolic blood pressure, in mmHg:

        -- baseline: 80.3

        -- end of RCT: 79.7, with adjusted change during RCT -0.47, P=0.282

        -- after 5 and 10 years: 80.0, with adjusted change from baseline of -0.866

 

---- Changes in fasting glucose, in mg/dL:

        -- baseline: 106.5

        -- end of RCT: 105.3, with adjusted change during RCT -0.99, P=0.246

        -- after 5 and 10 years: 100.1, with adjusted change from baseline of -5.83, P<0.001

 

-- Changes in hemoglobin A1c, in percent

        -- baseline: 5.5

        -- end of RCT: 5.5, with adjusted change during RCT -0.04, P=0.032

        -- after 5 and 10 years: 5.7, with adjusted change from baseline of 0.15,P<0.001

 

-- Changes in triglycerides, in mg/dL:

        -- baseline: 159.1

        -- end of RCT: 142.7, with adjusted change during RCT -15.94 P<0.001

        -- after 5 and 10 years: 151.2, with adjusted change from baseline of -8.15, P=0.055

 

-- Changes in HDL, in mg/dL:

        -- baseline: 44.2

        -- end of RCT: 47.8, with adjusted change during RCT 3.62, P<0.001

        -- after 5 and 10 years: 45.0, with adjusted change from baseline of 0.88, P=0.055

 

-- Changes in triglycerides/HDL ratio:

        -- baseline: 4.1

        -- end of RCT: 3.5, with adjusted change during RCT -0.62, P<0.001

        -- after 5 and 10 years: 3.8, with adjusted change from baseline of -0.26, P=0.078

 

-- Changes in METS-IR (Metabolic Score for Insulin Resistance):

        -- baseline: 48.2

        -- end of RCT: 45.4, with adjusted change during RCT -3.06, P<0.001

        -- after 5 and 10 years: 47.6, with adjusted change from baseline to follow-up of -0.84, P=0.022

-- Changes in hsCRP (high sensitivity CRP) in mg/L:

        -- baseline: 3.6

        -- end of RCT: 2.9, with adjusted change during RCT of -0.74, P<0.001

        -- after 5 and 10 years: 3.4, with adjusted change from baseline to follow-up of -0.29, P=0.278

-- Changes in Metabolic Syndrome score:

        -- baseline: 2.9

        -- end of RCT: 2.5, with adjusted change during RCT of -0.39, P <0.001

        -- after 5 and 10 years: 2.9, with adjusted change from baseline to follow-up of -0.02, P=0.762

-- Changes in % with prediabetes:

        -- baseline: 48.3%

        -- end of RCT: 51.4%

        -- after 5 and 10 years: 59.9%

-- Changes in % with diabetes:

        -- baseline: 10.2%

        -- end of RCT: 7.9%

        -- after 5 and 10 years: 21.8%

-- Changes in MET/hr of physical activity:

        -- baseline: 34.9

        -- end of RCT: 47.3, with adjusted change during RCT of 13.21, P <0.001

        -- after 5 and 10 years: 28.8, with adjusted change from baseline to follow-up of -5.54, P=0.017

-- Changes in composite risk score, which included 6 variables: waist circumference, average blood pressure ([systolic+diastolic]/2), fasting glucose, insulin, triglycerides, and inverted HDL-C (higher risk scores indicate greater risk):

        -- baseline: 0.0

        -- end of RCT:-0.3, with adjusted change during RCT of -0.28, P <0.001

        -- after 5 and 10 years: -0.1, with adjusted change from baseline to follow-up of -0.15, P<0.001

Here is a pictorial representation of the changes in the above over time:

Summary of the results in the above, perhaps not so user-friendly tables and picture, adjusting for age and sex:

    -- there were significant reductions in many anthropometric measures and MRI-assessed outcomes at the 18-month end of the studies and over 5- and 10-year follow-up

    -- the waist circumference and all but one abdominal adipose tissue groups were only partially regained but remained significantly reduced as compared to the baseline values. even in the 5- and 10- year follow-ups

        -- body weight and intrahepatic fat returned to baseline and intrapancreatic fat exceeded the baseline (ie, reestablishing the baseline body weight did not have full reestablishment of the visceral and deep subcutaneous abdominal body fat)

    -- those individuals who maintained VAT loss and those who regained it were similar at baseline, though those who maintained the VAT loss had lower weight and waist circumference and more favorable cardiometabolic markers

    -- Each 10% intervention-induced loss of visceral adipose tissue, superficial SAT, and intrapancreatic fat were associated with long-term postintervention improvements in Metabolic Score for Insulin Resistance, their composite risk score, and Metabolic Syndrome Severity Score (adjusted to weight change, Mediterranean diet adherence, and physical activity scores) at follow-up, and further measurements; all with P<0.05).

    -- a 1% waist circumference loss was associated with lower METS-IR (Metabolic Score for Insulin Resistance) and their composite risk score (both with P<0.05)

    -- per their supplementary materials, the highly statistically-significant differences in the groups maintaining vs regaining VAT were: follow-up BMI (30.1 vs 32.06), follow-up waist circumference (101.64 vs 108.39cm), follow-up VAT (109.21 vs 138.14 cm²), follow-up METS-IR (Metabolic Score for Insulin Resistance), and their follow-up composite risk score (-0.22 vs 0.06)

-- Diabetes risk decreased 28% for each 10% change in visceral fat, HR 0.72 (0.54-0.94), P=0.02

     -- there was a dose-response, with diabetes risk being:

         --5% VAT loss: 27% decrease, HR 0.72-0.96)

         -- 10% VAT loss: 30% decrease, R 0.70 (0.53-0.94)

         -- 15% VAT loss: 40% decrease, R 0.60 (0.38-0.94)

         -- 20% VAT loss: 48% decrease, R 0.52 (0.27-0.99)

    -- diabetes risk was not associated with weight loss, HR 1.06 (0.93-1.21), P=0.41

-- physical activity played a significant role in outcomes:

    -- the original two studies found that physical activity along with the different diets were associated with improved cardiometabolic measures, even in the setting of weight regain

ie, only the 10% loss of VAT reached statistical significance, though the 10% loss of Deep SAT was quite close to being statistically significant

Commentary:

-- visceral adipose tissue (VAT) is one of the strongest independent determinants of cardiometabolic disease, being associated with type II diabetes, metabolic syndrome, and all-cause mortality (see visceral fat consequences LancetDiabEndo2020 in dropbox, or Lancet Diabetes Endocrinol 2020; 8: 616–27)

    -- this is the specific intra-abdominal fat inside the peritoneal cavity, mostly in the omentum and mesentery around the intestines, that is metabolically active, and is associated with chronic inflammation, elaborating several circulating inflammatory markers, including IL-6, hsCRP,  TNF-a, MCP-1 (CCL2) and IL-1B

    -- this VAT accounts for about 10-20% of the total body fat in men and about 5-10% in women on average

    -- VAT has been known for many decades to be associated with hyperglycemia, dyslipidemia, hypertension, metabolic syndrome, and chronic inflammation; and chronic inflammation is associated with a large percentage of the chronic diseases with aging that are so prevalent now in our society: https://pubmed.ncbi.nlm.nih.gov/39078396/, https://gmodestmedblogs.blogspot.com/2026/02/racial-disparities-stress-and-mortality.html )

-- in contrast, the superficial fat tissue is associated with increased adiponectin concentrations (associated with increased insulin sensitivity, lower blood sugar, decreased inflammation, break-down of fatty acids), lower inflammatory infiltration, and more smaller blood vessels: see adipose abdominal tissue superficial inc adiponectin in dropbox, or doi: 10.1002/oby.20417 ). This likely results in decreased diabetes risk

    -- and gluteal (buttocks) fat has been found to have different effects on health:

        -- superficial gluteal fat maintains metabolic health in some studies, enhancing insulin sensitivity, likely related to its ability to store triglycerides

        -- deep gluteal subcutaneous adipose tissue (Deep SAT) is associated with local expression of inflammatory genes (CCL2, TNF-α, IL-18, macrophage markers CD14, CD68) and lower adiponectin expression

-- the figure below represents the multiple bad effects of chronic inflammation, from chronic inflammation diseases JImmunResearch2016 in dropbox or doi.org/10.1155/2016/8426874

 

-- of note, some lean people can also have increased visceral fat, leading to the perhaps unexpected increased risk of type II diabetes and cardiovascular disease, as found in overweight people

-- the 18-month CENTRAL and 18-month DIRECT-PLUS randomized controlled trials found that visceral adipose tissue (VAT), deep subcutaneous adipose tissue (deep SAT), and intrahepatic fat (IHF) reductions were associated with very significant improvements in cardiometabolic markers independent of overall weight loss; superficial subcutaneous adipose tissue was found to be metabolically neutral or favorable; intrahepatic fat was associated with modest reduction in cardiometabolic markers in the CENTRAL trial

-- Mediterranean diet is associated with polyphenol enrichment and carbohydrate reduction as well as increased physical activity and was helpful with decreases in VAT, deep SAT, and IHF; studies on Mediterranean diets confirm that weight or BMI are not determinant and that these specific fat depots are more important than weight or BMI.

    -- waist circumference and some other measures are significantly more predictive of visceral obesity: https://gmodestmedblogs.blogspot.com/2024/09/body-roundness-index-is-better.html)

This study found a few significant things:

-- Despite complete weight regain, there were partially preserved fat depots (specifically visceral adipose tissue, deep SAT, and superficial SAT) in their long-term, post-intervention follow-up

-- there was also only partial regain of waist circumference (not body weight, which was regained fully) long term, reinforcing the importance of measuring waist circumference over body weight or BMI as physiological markers of important future clinical events such as diabetes and other problems associated with chronic inflammation, as depicted above

    -- for studies supporting waist circumference and other similar markers of clinically-important measures of overweight/obesity over BMI, see https://gmodestmedblogs.blogspot.com/2024/09/body-roundness-index-is-better.html

        --ie, visceral fat loss rather than weight loss emerged as a key target for durable cardiometabolic health.

-- Each 10% intervention-induced loss of visceral adipose tissue, superficial SAT, and intrapancreatic fat were associated with long-term postintervention improvements in Metabolic Score for Insulin Resistance (their composite risk score), and the Metabolic Syndrome Severity Score (adjusted to weight change, Mediterranean diet adherence, and physical activity scores) at the 18-month follow-up, and in the further measurements 5 and 10 years later); all with P<0.05

    -- the long-term effects of diet and physical activity lifestyle interventions reinforces their benefit in improving cardiometabolic measures despite weight regain in the long-term

    -- and even a 5% decrease in VAT was associated with a significant decrease in diabetes risk

-- Diabetes risk decreased 28% for each 10% change in visceral fat, independent of baseline A1c levels

    -- and this risk tracked with the percent change in the VAT, in a dose-response manner: the more decrease in visceral fat, the less risk for diabetes

-- other fat depot changes were also long-term on the average, including deep and superficial adipose tissue. these were associated with cardiometabolic improvement, but to a lesser degree than VAT

-- in contrast to the above, intrapancreatic and intrahepatic fat losses had a different trajectory, with the long-term intrahepatic fat returning to the pre-intervention levels on average, and the intrapancreatic fat significantly exceeding the pre-intervention fat levels

    -- there is some evidence in studies that the intrapancreatic fat loss is associated with improved cardiometabolic indices over time, specifically metabolic syndrome, impaired glucose metabolism and diabetes. we need longer studies to interpret the true importance of intrapancreatic fat.

        -- and the concern: excessive intrapancreatic fat deposition (found in about 20% of people), in a 2025 systematic review of 23 studies, was found to confer a 3-fold increased risk of pancreatic cancer (in a dose-response way), a 4-fold increased risk of acute pancreatitis, and 21% increawed risk of diabetes: see pancreatic cancer inc if excess intrapancreatic fat AmJGastro2025 in dropbox or 1. doi.org/10.14309/ajg.0000000000003799 

        -- and studies have found that there has been an increase in intrapancreatic fat over time. is this an important reason for the increase in pancreatic cancer incidence now?? (https://gmodestmedblogs.blogspot.com/2021/12/pancreatic-cancer-increasing-incidence.html)

-- this all brings up the current elephant in the room: GLP-1 receptor agonists lead to really dramatic weight loss, which for most people far exceeds the achievable weight loss of diet and exercise. do they decrease VAT???

    -- several studies have been done finding the GLP-1's used for several indications (diabetes, weight loss, MAFLD) have found profound effects on multiple fat depots (VAT, SAT, total adipose tissue, epicardial fat, intrahepatic fat), with some finding disproportionately large VAT fat loss: https://dom-pubs.onlinelibrary.wiley.com/doi/10.1111/dom.16012 and others finding similar VAT reductions: https://pubmed.ncbi.nlm.nih.gov/35797355/ 

  -- the role of epicardial fat is complex, with some benefits including protecting the adjacent myocardium with a thermogenic function, but also some harm through elaborating pro-inflammatory and profibrotic cytokines and being associated with increased heart failure risk:  https://pubmed.ncbi.nlm.nih.gov/35296869/ and https://pubmed.ncbi.nlm.nih.gov/34045020/

    -- of note, there is concern about muscle loss with GLP1's. however, a quite impressive, well-conducted study found that GLP-1's decrease free fat mass loss with likely benefit (not harm) to skeletal muscle: https://gmodestmedblogs.blogspot.com/2024/11/glp-1-agonists-improve-skeletal-muscle.html

--a somewhat perplexing issue is that though there were pretty substantial sustainable decreased diabetes risk with increasing VAT loss even at the 5% level, the level of hsCRP, a sensitive but not independent measure of systemic inflammation, decreased from 3.6 to 2.9 at 18 weeks but then rose to 3.4 at the 5 and 10 year mark

    --there are an array of other inflammatory markers associated with high levels of systemic inflammation but no elevated hsCRP.

    --It has been shown that increases in hsCRP begin 6-8 hours after an inflammatory trigger, peaks at 36-50 hours and has a half-life of 19 hours, and returns to baseline in 4-10 days. Studies have shown that there can be initial increases in hsCRP that normalize despite ongoing inflammation: https://onlinelibrary.wiley.com/doi/10.1111/imm.13739

 

Limitations:

-- the CENTRAL and DIRECT-PLUS trials included in this analysis involved middle-aged participants and 91% were men, limiting the generalizability to women and a wider range of individual ages

-- these studies did not include important fat depots, including perivascular and epicardial fat, limiting conclusions about the likely highly significant associated  cardiovascular pathology, though the GLP-1 study cited above did find profound effects on epicardial fat

-- the researchers combined 2 studies with different interventions (particularly prescribed diets), different selection criteria, and different lengths of follow-up (5 vs 10 years). they did incorporate time-to-event data which helps but this does not eliminate potential biases in their results. and their use of multivariable adjustment did not alter the results by controlling for baseline adiposity, weight change, intervention group and some lifestyle factors (overall diet, activity level, and medications taken), but there may well be unaccounted confounders (comorbidities, medications, psychosocial issues) that were not taken into account and may have affected their results

-- this study also took place in Israel and there may well be substantial differences in the array of risk factors in different areas, in particular those related to diet, exercise, social stressors, psychological and psychiatric risk factors including depression, etc that would limit generalizability

-- in an 18-month study, the baseline values noted above may have changed: did some individuals stop or start smoking? increase or decrease their exercise? change their eating habits?  these types of changes could affect the results.

-- this very impressive study found continued benefit to VAT loss up to 10 years after the study stopped. it would be useful to see the results if this study continued longer (especially an issue since the intrapancreatic fat increased and led to bad outcomes). also, it would be useful to have granular data on the specifics of the long-term changes: did those in the studies who largely preserved their VAT loss continue after the study to adhere to the lifestyle changes done the study in the future or revert to their prior diets/physical activity? were there specific components of the 18-month study that they incorporated into their future lifestyle that preserved their VAT loss. and were there specific types of exercise or dietary changes that were particularly helpful in preserving the VAT loss despite weight gain? 

So, what do all of these results, graphs, etc mean?

--it seems that weight loss by diet/exercise is associated with visceral fat loss that is sustainable even up to 10 years after the studies above in many people

    -- GLP-1 receptor agonists seem to be particularly useful in preferentially decreasing VAT, though we do not have long-term studies to confirm this

-- this change in visceral fat is associated with significant cardiometabolic improvements, including the risk of diabetes

-- the changes in visceral fat and also in cardiometabolic outcomes were independent of weight gain

    --which reinforces the very well-known fact that BMI has a far inferior correlation to important cardiometabolic events than waist circumference or other measures that assess "body roundness": https://gmodestmedblogs.blogspot.com/2024/09/body-roundness-index-is-better.html

-- as a perhaps tangential comment: though this was an impressive study, it is disturbing that the authors did not comment on the issue of the potential positive role of GLP-1 receptor agonists, especially in light of prior concerns about their being associated with pancreatitis and pancreatic cancer. pancreatitis has mostly debunked though there may be a rare occurrence: see GLP-1 review NEJM2026 in dropbox, or DOI: 10.1056/NEJMra2500106), and pancreatic cancer has been largely negated in a recent systematic review/metanalysis (doi: 10.7326/ANNALS-25-02237). And the editors of this prominent journal (Circulation) did not mention this to the authors of the article...

    -- and, perhaps more importantly, there was an increase in intrapancreatic fat found in the study's long-term results, which has a real potential increased risk for pancreatitis and pancreatic cancer.....

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

to get access to the dropbox: https://www.dropbox.com/scl/fo/pcf4vymtpo77f49wd2ljr/ANyJqB40CuBd-DMEr67XDoI?rlkey=5dsalmuhf8ekd0svsgvn93xm1&dl=0

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them