Glycemic and Insulinemic Responses of Fresh, Freeze-Dried, and Cooked Apples: As Single Food or Preload

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study investigates the impact of raw, cooking and freeze-drying on the postprandial glycaemic, insulinemic responses and preloading effect of apples, including viscosity and digest analysis. This work has a very relevant topic, it is well thought out and structured, but I have a few complaints:

The introduction should include a brief review of previous research on this topic and emphasize what is new in this approach. Line 76-77, Specify whether all tested samples (raw, cooked, freeze-dried apples) came from the same listed species and batch?

Line 77, Were the freeze-dried apples used directly from the manufacturer or were they rehydrated before consumption?

83-84, Please explain why only young healthy volunteers were included?

Line 106-113, How it was achieved for the analyzed meals containing 50 g of available carbohydrates, i.e. whether total energy, fiber and water content were matched or adjusted between treatments?

Line 199-200, State the unit in which buffering capacity was expressed in this work.

Line 275-277, Specify whether the glucose response in FA is statistically significantly lower compared to RA and CA?

Line 460-462, What non-textural properties mainly?

Line 482-484, Try to explain the reasons for this discrepancy?

The discussion needs more comparisons with the previously published findings in order to have a more complete understanding of the results obtained.

The authors could conclude the practical significance of this research, for example the development of food products with added freeze-dried apple with a low glycemic index, and indicate the direction of further research.

Author Response

1.The introduction should include a brief review of previous research on this topic and emphasize what is new in this approach. Line 76-77, Specify whether all tested samples (raw, cooked, freeze-dried apples) came from the same listed species and batch?

Thank your for this question. The raw material of apple samples came from the same orchard in Shandong province, China, the material base of Haoxiangni Co. Ltd. The cooked apple was prepared in laboratory, while the freeze-dried apple was produced by the commercial vacuum-drying facility in the company. The company did not give any fund but they agreed to provide the apple and freeze-dried apple samples on our request. We detailed the source of apple produce and the preparation of sample in line 80-87.

 

2.Line 77, Were the freeze-dried apples used directly from the manufacturer or were they rehydrated before consumption?

Thank your for this question. The FA was made by Haoxiangni co., ltd, in nitrogen filled package. The FA was consumed as dry food without rehydration. Water was provided with FA and the subjet drink water whenever they feel like some water to help swallowing. We detailed the ingestion of apple samples in line 130-136, and line 176-178.

 

3.83-84, Please explain why only young healthy volunteers were included?

Thank you for this question. The objective of this study is to investigate the glycemic response characteristics of foods, rather than to address therapeutic or dietary management in diabetic patients. Young, healthy individuals of consistent age and physical activity levels, living in the same environment, were selected as subjects because their physiological responses following food intake exhibit relatively smaller variations, thereby better highlighting the intrinsic properties of the foods. In contrast, daily fluctuations in the clinical condition of individuals with diabetes lead to unstable glycemic responses from day to day, which would introduce additional confounding factors when comparing the characteristics of different food samples.

 

4.Line 106-113, How it was achieved for the analyzed meals containing 50 g of available carbohydrates, i.e. whether total energy, fiber and water content were matched or adjusted between treatments?

Thank you for this question. The calculation of available carbohydrate (AC) was as follow equation: 1.05 × (Disaccharides) + Monosaccharides. The sugar moiety of apple sample was determined using ion chromatography (line 127-128). The available carbohydrate was matched among test arms based on the result of determination. However, the volumes of fluid was not precisely matched in three arms.

 

5.Line 199-200, State the unit in which buffering capacity was expressed in this work.

Thank you for this question. In this study, acid buffering capacity (acid BC) is defined as the titratable acidity divided by the total pH change (initial pH= − 1.5), with the unit expressed as mmol H⁺/pH unit.

 

6.Line 275-277, Specify whether the glucose response in FA is statistically significantly lower compared to RA and CA?

Thank you for your question. Compared with RA, FA showed significantly lower blood glucose responses at 15 and 30 minutes postprandial (P=0.0006, P=0.0462). Compared with CA, FA demonstrated significantly lower responses at 30 and 45 minutes postprandial (P=0.0337, P=0.0159). The statistical differences between these characteristics are detailed in Table 4 below.

7.Line 460-462, What non-textural properties mainly? 

The food matrix texture primarily impacts the rate at which carbohydrates and other components are released from cellular structures, as well as the kinetics of interactions among various food constituents. Non-textural factors affecting glycemic response include the quantity and type of carbohydrates (i.e., sugars, in the context of fruits), as well as co-existing components such as organic acids, polyphenols, and dietary fibers including pectin. These aspects will be further elaborated in the subsequent discussion.

 

8.Line 482-484, Try to explain the reasons for this discrepancy?

Compared to fresh apples, cooked apples exhibit reduced chewiness, softened texture, and accelerated sugar release from cells—factors that should elevate the glycemic response. However, several factors may have counteracted this effect.

(1) During cooking, organic acids in apples are also released from cells. Consuming organic acids helps lower the glycemic response. Table 9 shows that the pH of cooked apples drops to 3.98, slightly lower than fresh apples, with a slight increase in buffering capacity. In a previous study on jujubes, fresh jujubes had lower acidity¹ with a pH of 5.35.

(2) Apple skins are rich in pectin, and pressure cooking may also cause pectin to leach out from the skin. Pectin itself has the effect of delaying the postprandial rise in blood sugar. (Unfortunately, we did not measure changes in pectin content).

(3) Although polyphenol content decreased after cooking, the reduction was not significant. Apple peel is an excellent source of polyphenols, with certain compounds like geniposidic acid concentrated in the skin. Cooking may cause polyphenols from the peel to migrate into the flesh, potentially enhancing insulin sensitivity in subjects. Previous studies have demonstrated that pre-meal supplementation with apple peel extract rich in alkaloids and polyphenols significantly reduces postprandial blood glucose and insulin spikes².

(4)The sugar composition of apples and dates differs. Fuji apples contain a higher proportion of fructose, which inherently has a lower GI value. Even after dissolution, it does not elevate the GI value. Furthermore, research suggests that consuming small amounts of fructose may improve hepatic glucose metabolism³.

Therefore, in summary, the effects of cooking apples differ from those of cooking dates. Under the combined influence of various factors, cooked apples did not induce a higher glycemic response than fresh fruit.

 

9. The discussion needs more comparisons with the previously published findings in order to have a more complete understanding of the results obtained.

This is a valuable suggestion. We have consulted prior studies on date cooking and research on apple polyphenols, apples, and dried apples in blood glucose response to supplement the discussion.

 

10. The authors could conclude the practical significance of this research, for example the development of food products with added freeze-dried apple with a low glycemic index, and indicate the direction of further research.

Thank you for your insightful suggestions. We hope this study provides a foundation for future development of low-GI foods using freeze-dried apple ingredients, particularly in the realm of healthy snacks and confectionery. We have incorporated relevant content line by line into the conclusion section: “This study highlights the diversity of glycemic characteristics among various fruits following processing and cooking.” On one hand, findings suggest that cooked fruits retaining polyphenols and pectin should not be equated with apple juice, as they may still exhibit moderate to low glycemic responses. On the other hand, this study first demonstrates that freeze-dried apples may offer an ideal glycemic response, positioning them not only as a healthy snack but also as a convenient ingredient for developing low-glycemic-response health foods. Future research should explore additional freeze-dried fruit products to uncover their potential health benefits for glycemic management."

1.Wei J ,Liu A ,Fan Z , et al.Cooking Increased the Postprandial Glycaemic Response but Enhanced the Preload Effect of Air-Dried Jujube[J].Foods,2025,14(7):1142-1142.DOI:10.3390/FOODS14071142.

2.Venugopal G , Dash R , Agrawal S ,et al.A Novel Nutraceutical Supplement Lowers Postprandial Glucose and Insulin Levels upon a Carbohydrate-Rich Meal or Sucrose Drink Intake in Healthy Individuals—A Randomized, Placebo-Controlled, Crossover Feeding Study[J].Nutrients, 2024, 16(14):14.DOI:10.3390/nu16142237.

3.Ruixin Z , Zhihong F , Yang D ,et al.Postprandial Glycaemic Responses of Dried Fruit-Containing Meals in Healthy Adults: Results from a Randomised Trial[J].Nutrients, 2018, 10(6):694.DOI:10.3390/nu10060694.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The paper Freeze-drying lowered postprandial glycaemic and insulin response of apples: randomised controlled trial, phenolics retention and matrix texture analysis is interesting and worth attention.

However, I have some remarks

 

Remarks:

Was the same variety of fresh apples used to prepare the cooked apples? To properly compare the products, freeze-dried apples from the same variety should be prepared. What variety of freeze-dried apples was used?

 

Please add the number of repetitions to all types of measurements

What was the shape and size of the freeze-dried samples? Were they also cut for textural testing?

Line 446 The favorable glycemic profile of freeze-dried apple may be partially attributed to their superior phenolic retention.  I don’t agree with it. Many papers showed that in the Folin-Ciocalteau micro method, the high content of TPC was linked with better extraction of polyphenols from dried samples than from fresh. This does not mean a higher retention of polyphenolic compounds in the dried tissue, but rather their better extraction from this tissue.

Line 454 -459: How textural characteristics might contribute to the stable glycemic response of FA is unclear. It may be more likely that the effect is due to the amount of water content.

Conclusions are mostly statements and resemble an abstract.

Author Response

1.Was the same variety of fresh apples used to prepare the cooked apples? To properly compare the products, freeze-dried apples from the same variety should be prepared. What variety of freeze-dried apples was used?

Thank your for this question. The raw material of apple samples came from the same orchard in Sandon provinc, China, the material base of Haoxiangni Co. Ltd. The cooked apple was prepared in laboratory, while the freeze-dried apple was produced by the commercial vacuum-drying facility in the company. The company did not give any fund but they agreed to provide the apple and freeze-dried apple samples on our request. We detailed the source of apple produce and the preparation of sample in line 80-87.  

 

2.Please add the number of repetitions to all types of measurements

Thank your for this advice. We have revised the manuscript and added the repetition of measurements in the material and method part (line 207, 228, 239 and 261).

 

3.What was the shape and size of the freeze-dried samples? Were they also cut for textural testing?

Thank your for this question. We detailed the treatment on apple materials on line 85-93. The apple was diced right before the test session because the freeze-dried apple we received was 2-cm dices.

 

4.Line 446 The favorable glycemic profile of freeze-dried apple may be partially attributed to their superior phenolic retention.  I don’t agree with it. Many papers showed that in the Folin-Ciocalteau micro method, the high content of TPC was linked with better extraction of polyphenols from dried samples than from fresh. This does not mean a higher retention of polyphenolic compounds in the dried tissue, but rather their better extraction from this tissue。

Thank you for your comment. Indeed, after cooking or processing, the extraction rate of phenolic compounds may increase due to the disruption of cellular structures. Numerous studies have found that the content of phenolic substances may be significantly higher in processed or cooked vegetables and fruits compared to fresh ones. However, a similar phenomenon may occur within the human digestive system. Phenolic compounds in processed or cooked foods may be more readily extracted in the digestive tract rather than passing into the colon as undigested food residues. Additionally, it cannot be ruled out that during processing and cooking, some bound phenolic compounds on plant cell walls are released, thereby increasing the total amount of available phenolic compounds¹. Based on your feedback, we have revised the relevant content to clarify that the extraction rate of phenolic compounds may vary depending on the food matrix. Therefore, while the differences in phenolic content between samples correlate with glycemic response, this does not establish a causal relationship between the two.

5.Line 454 -459: How textural characteristics might contribute to the stable glycemic response of FA is unclear. It may be more likely that the effect is due to the amount of water content.

Thank you for your comment. Previous studies have indicated that a firm texture and high masticatory effort required for foods can delay the release of sugars from cellular structures, thereby contributing to a lower postprandial glycemic response. However, FA does not exhibit a particularly hard texture. In fact it is very soft. As you suggested, we can not rule out the possibility that, the reduced moisture content, and the amount of liquid intake at meal or after meal, may affect the postprandial glycemic response by modifying the rate of sugar release and gastric emptying.

 

We provided 400g drinking water with FA in the test meal (we described the protocol in line   124 to 130). However, we did not ask the subject to drink all the water. They are free to drink any amount of water as they feel comfortable to swallow the dried apple meal. In fact, most of them left some water when they finish the dried apple. It is true that the fluid intake of cooked apple meal is the highest, followed by the fresh apple meal, and the freeze-dried apple meal is the lowest.

 

We surmised that the decrease in water content could prolong the time required for sugars the dehydrated cells to fully dissolve, in spite of the drinking water provided with FA. On the other hand, consumption of low-moisture foods might lead to slower gastric emptying due to lowered intragastric fluid volume. It has been reported that co-ingestion of water with starchy foods can accelerate gastric emptying[2], thereby elevating the glycemic response. This hypothesis warrants further validation in future studies.

 

6.Conclusions are mostly statements and resemble an abstract.

Thank your for your suggestion. We have modified the statement in the part of conclusion .

 

[1]G. Gómez-Mascaraque, Laura, Dhital S ,López-Rubio, Amparo,et al.Dietary polyphenols bind to potato cells and cellular components[J].Journal of Functional Foods, 2017, 37:283-292.DOI:10.1016/j.jff.2017.07.062.

[2] Matsunaga T , Adachi T , Yasuda K .The effect of co-ingesting rice and liquid on glycemic response and gastric emptying in healthy subjects[J].Physiology & Behavior, 2025, 292(000).DOI:10.1016/j.physbeh.2025.114837.

Reviewer 3 Report

Comments and Suggestions for Authors

Line 2–4, please, shorten and clarify the title to reflect design, endpoints, and key mechanisms without overclaiming.

Line 12, please, change “was investigated” to “investigated” (active voice).

Line 12–30, please, fix grammar/wording (“insulinemic,” “led”), soften claims, and ensure the Abstract precisely matches the Results.

Line 19–21, I recommend, resolve the Abstract contradiction about insulin excursions (align with Figure 4/Results).

Line 24–30, please, replace “lead” with “led” and remove subjective phrasing (e.g., “remarkable advantage”).

Line 63–73, please consider adding a sharper gap analysis (what’s unknown about freeze‑dried vs raw/cooked fruit on PGR/PIR; prior human data; hypothesized mechanisms).

Line 75–81, it would be beneficial to report FA product specs (slice thickness, residual moisture, a_w, porosity, rehydration ratio/time, additives like SO₂).

Line 80, please, add units/conditions (“115 °C”, pressure, come‑up/cool‑down handling, cut size) to make cooking reproducible.

Line 94–101, please consider reporting the power‑calc effect size, assumed SD, and the specific endpoint; clarify if both sub‑studies were powered.

Line 109–118, please, state the 250 g water co‑ingestion instructions (timing; together vs sequential), as this affects rehydration and gastric emptying.

Line 112–118, I recommend, detailing the sugar analysis used to compute AC (instrument, standards, QC/replicates, uncertainty).

Line 119–124, please, describe randomization (generation, allocation concealment) and any blinding of assay analysts.

Line 136–141, I recommend, justifying capillary sampling for insulin (matrix validation, handling time, freeze–thaw cycles) and reporting intra/inter‑assay CVs.

Line 152–163, please, harmonize treatment nomenclature across text, tables, and figures (RA+R, CA+R, FA+R, W+R).

Line 158–166, please, fix figure numbering and cross‑references in the preload protocol schematic.

Line 167–178, it would be beneficial to add extraction yield/QC (matrix spikes) for TPC/TFC and control peel:flesh ratio and homogenization across treatments.

Line 179–191, please consider reporting specimen conditioning, orientation, and whether FA texture was tested dry or rehydrated; align with in‑mouth conditions since FA was co‑ingested with water.

Line 192–200, please, provide titration curves and electrode calibration/ionic‑strength control for buffering capacity; define acid/alkali BC consistently.

Line 202–206, I recommend, justifying the 20 s, 13 000 rpm blending in the oral phase (potential over‑homogenization vs physiological mastication).

Line 214–218, please, avoid freeze–thaw prior to rheology or validate its impact (freeze–thaw can alter pectin/starch networks).

Line 219–225, it would be beneficial to justify shear‑rate ranges, report temperature control, evaporation mitigation, gap/normal‑force settings, and replicate acceptance criteria.

Line 235–236, I recommend, removing “postprandial HOMA‑IR AUC” as an insulin‑sensitivity proxy; use validated dynamic indices (e.g., Matsuda) or describe your metric as exploratory.

Line 237–255, please consider defining the mixed‑model structure (random intercepts/slopes), reporting diagnostics, and applying multiplicity control; avoid liberal LSD across repeated measures (prefer Tukey/Holm/BH; sphericity corrections as needed).

Line 273–281, please, replace “variety‑method‑time‑effect” with standard “treatment×time interaction.”

Line 285–293, it would be beneficial to explain NAUCglu (clinical meaning) and how dips below baseline were handled (baseline drift, meter bias).

Line 307–312, please consider reframing conclusions about “insulin sensitivity” to descriptive PIR features unless a validated index is used.

Line 322–337, I recommend, continuing the above reframing and clearly labelling any composite indices as exploratory.

Line 351–354, please, correct Table 6 caption/labels (currently “insulinemic” while columns are glycemic metrics).

Line 351–381, please, harmonize captions/labels for Tables 6–7 (glycemic vs insulinemic) to match contents.

Line 421–426, please consider tempering causal language that in‑vitro viscosity “explains” PGR/PIR without direct mediational evidence.

Line 428–433, please, standardize statistical notation in figures/captions (e.g., “type×phase interaction”) and clarify what letters/symbols denote.

Line 435–470, I recommend, presenting viscosity, TPC/TFC, and other factors as candidate mediators rather than confirmed mechanisms.

Line 451–457, please, soften phrasing from “contribute substantially” to “may contribute,” since mechanisms weren’t measured.

Line 463–470, please consider qualifying statements about enzyme inhibition/transport modulation as hypotheses to be tested.

Line 471–479, I recommend, analytically separating matrix‑softening (cooking) from sugar‑profile and acidity effects (e.g., acidifying RA to CA pH; matched fructose:glucose boluses).

Line 481–484, it would be beneficial to continue the above sensitivity analyses to disentangle mechanisms.

Line 499–507, please, tighten preload conclusions: all apple preloads attenuated peaks vs water; between‑apple differences were modest and time‑window dependent.

Line 548–549, please, fix abbreviation typos (“frezz‑dried” → “freeze‑dried”) and ensure all abbreviations appear in text before the list.

Author Response

1.Line 2–4, please, shorten and clarify the title to reflect design, endpoints, and key mechanisms without overclaiming.

Thanks for your advice. We change the title to“Glycemic and insulinemic Responses of fresh, freeze-dried and cooked apples: as single food or preload.”

 

2.Line 12, please, change “was investigated” to “investigated” (active voice).

Thank you for pointing out the error. The correction has been made in the text (in line 11 ).

 

3.Line 12–30, please, fix grammar/wording (“insulinemic,” “led”), soften claims, and ensure the Abstract precisely matches the Results.

Thank you for pointing out the error. The correction has been made in the text (in line 17, 19).

 

4.Line 19–21, I recommend, resolve the Abstract contradiction about insulin excursions (align with Figure 4/Results).

Thank you for pointing out the error. The correction has been made in the text (in line 23).

 

5.Line 24–30, please, replace “lead” with “led” and remove subjective phrasing (e.g., “remarkable advantage”).

Thank you for pointing out the error. The correction has been made in the text (in line 23).

 

6.Line 63–73, please consider adding a sharper gap analysis (what’s unknown about freeze‑dried vs raw/cooked fruit on PGR/PIR; prior human data; hypothesized mechanisms). 

Thank for your advise. We revised the gap analysis part and make clear that the glycemic response, the insulinemic response, and the preload effect of freeze-dried fruits are yet to be investigated (line 55 to line 71).

 

7.Line 75–81, it would be beneficial to report FA product specs (slice thickness, residual moisture, a_w, porosity, rehydration ratio/time, additives like SO₂).

Thank you for your suggestion. The FA was 2 cm diced instead of sliced. The water content of FA was 2%。It was not rehydrated before ingestion. Instead, drinking water was provided to help swallowing the FA. No additives was added to FA. The FA was light yellow instead of white in color.

 

8.Line 80, please, add units/conditions (“115 °C”, pressure, come‑up/cool‑down handling, cut size) to make cooking reproducible.

Thank you for pointing out the error. The correction has been made in the text, in line 91.

 

9.Line 94–101, please consider reporting the power‑calc effect size, assumed SD, and the specific endpoint; clarify if both sub‑studies were powered.

Thank you for your suggestion. To supplement, a previous study on the postprandial glycemic response to apples, conducted with 10 subjects, yielded a power-calc effect size of 80% and SD is lower than 55.15, with the endpoint being iAUC. The relevant details have been incorporated into the data analysis section in line 110.

 

10.Line 109–118, please, state the 250 g water co‑ingestion instructions (timing; together vs sequential), as this affects rehydration and gastric emptying.

Thank you for your advice. We detailed the ingestion instructions in line 130-136

 

11.Line 112–118, I recommend, detailing the sugar analysis used to compute AC (instrument, standards, QC/replicates, uncertainty).

Thank you for your suggestion. The determination of sugars was conducted by the Feed Analysis and Testing Center at China Agricultural University. The specific method employed was non-standard ion chromatography using a Dionex CarboPac PA20 column (150 mm × 3 mm, 6 μm).

12.Line 119–124, please, describe randomization (generation, allocation concealment) and any blinding of assay analysts.

Thank you for your advice. The test order was randomized with http://www.example.com. The trial was open-labeled because the the remarkable difference in shape and texture among three test samples made it difficult to blind the test foods. We added this information in line 138-141.

 

13.Line 136–141, I recommend, justifying capillary sampling for insulin (matrix validation, handling time, freeze–thaw cycles) and reporting intra/inter‑assay CVs. 

Thank you for your suggestion. In the experiment, subjects were instructed to complete blood collection within 3 minutes to prevent hemolysis. Plasma used for insulin measurement was separated from serum after sampling and stored at -80°C until testing, remaining frozen and untouched until analysis. Assays were performed by trained personnel strictly following the kit protocol. Each sample included two replicates, with intra-assay and inter-assay coefficients of variation (CVs) of <10% and <15%, respectively. We added this information in line 163-168.

 

14.Line 152–163, please, harmonize treatment nomenclature across text, tables, and figures (RA+R, CA+R, FA+R, W+R).

Thank you for pointing out the error. The correction has been made in the text (in line 179).

 

15.Line 158–166, please, fix figure numbering and cross‑references in the preload protocol schematic.

Thank you for pointing out the error. The correction has been made in the text (in line 192-197).

 

16.Line 167–178, it would be beneficial to add extraction yield/QC (matrix spikes) for TPC/TFC and control peel:flesh ratio and homogenization across treatments.

Thank you for your suggestion. The extraction yield for TPC matrix spikes is 84.5 ± 1.5%, while that for TFC matrix spikes is 89.2 ± 3.2%. When sampling fresh apples, cut them into 2cm cubes, ensuring each cube has only one side with peel.We added these information in line 213  and line  215 .

 

17.Line 179–191, please consider reporting specimen conditioning, orientation, and whether FA texture was tested dry or rehydrated; align with in‑mouth conditions since FA was co‑ingested with water.

Thank you for your suggestion. FA is in a dry state during texture analysis. When consuming FA samples, drinking water is provided, but the samples are not soaked in water before consumption. Instead, water is used to aid chewing and swallowing.We added this information in line 130 to line 136.

 

18.Line 192–200, please, provide titration curves and electrode calibration/ionic‑strength control for buffering capacity; define acid/alkali BC consistently. 

Thank you for your suggestion. The titration curve is shown in the figure below.

19.Line 202–206, I recommend, justifying the 20 s, 13 000 rpm blending in the oral phase (potential over‑homogenization vs physiological mastication).

Thank you for your suggestion. This parameter is based on the previous pre-experiment setting, which may be different from human physiological chewing. Further experiments are needed to verify the existence of this difference.

 

20.Line 214–218, please, avoid freeze–thaw prior to rheology or validate its impact (freeze–thaw can alter pectin/starch networks).

Thank you for this advice. The freeze-dried sample was obtained from a major freeze-dry food manufacturer, stored and handled under room temperature during the study and the texture analysis. There was no freeze-thaw transition in the sample. Regarding the carbohydrate network in food matrix, since there is almost no starch in apple so what is important is pectin. The freeze-dried food was dehydrated by removing the sublimed water vapor from a frozen food. That means that the freeze-dried food lost almost all free water while kept its cell wall skeleton including the pectin component.

 

21.Line 219–225, it would be beneficial to justify shear‑rate ranges, report temperature control, evaporation mitigation, gap/normal‑force settings, and replicate acceptance criteria.

Thank you for your suggestion. Regarding the shear rate range and temperature control, please refer to the previous study (1). The researchers also noted that the rheological properties of the digestive fluid were not significantly affected by storage time, indicating that storing thawed digestive fluid in ice water does not significantly impact its rheology. And the texture analysis of all samples was performed under room temperature. It is not likely to cause evaporation mitigation during the texture tests.

1 Wu P , Dhital S , Williams B A ,et al.Rheological and microstructural properties of porcine gastric digesta and diets containing pectin or mango powder[J].Carbohydrate Polymers, 2016, 148:216-226.DOI:10.1016/j.carbpol.2016.04.037.

 

22.Line 235–236, I recommend, removing “postprandial HOMA‑IR AUC” as an insulin‑sensitivity proxy; use validated dynamic indices (e.g., Matsuda) or describe your metric as exploratory.

Thank you for your suggestion. We have added Matsuda Index in our research.

 

 

23.Line 237–255, please consider defining the mixed‑model structure (random intercepts/slopes), reporting diagnostics, and applying multiplicity control; avoid liberal LSD across repeated measures (prefer Tukey/Holm/BH; sphericity corrections as needed).

Thank you for your suggestion. The text has been updated to refer to Tukey's multiple comparisons.

 

24.Line 273–281, please, replace “variety‑method‑time‑effect” with standard “treatment×time interaction.”

Thank you for pointing out the error. The correction has been made in the text (in line 328).。

 

25.Line 285–293, it would be beneficial to explain NAUCglu (clinical meaning) and how dips below baseline were handled (baseline drift, meter bias).

Thank you for your professional advice. The definition of NAUCglu has been thoroughly explained in the text, with relevant literature (Reference 42) cited to clarify its clinical significance. Negative Area Under the Glucose Curve (NAUC_glu) refers to the area enclosed by the baseline (i.e., the blood glucose level at time zero) and the data points below it in the postprandial glucose curve¹. Research indicates that the negative area occurring after the glucose peak subsides—which falls below the preprandial blood glucose level—accurately predicts postprandial self-reported hunger and subsequent energy intake. We added this in line 269 to line 273.

1. Wyatt P, Berry S E, Finlayson G, O’Driscoll R, Hadjigeorgiou G, Drew D A, Khatib H A, Nguyen L H, Linenberg I, Chan Andrew T, Spector T D, Franks P W, Wolf J, Blundell J, Valdes A M. Postprandial glycaemic dips predict appetite and energy intake in healthy individuals. Nature Metabolism, 2022, 3(4), 523-529. doi.org/10.1038/s42255-021-00383-X.

 

26.Line 307–312, please consider reframing conclusions about “insulin sensitivity” to descriptive PIR features unless a validated index is used. 

Thank you for your suggestion. The modification has been made here to analyze insulin sensitivity using the Matsuda index.

 

27.Line 322–337, I recommend, continuing the above reframing and clearly labeling any composite indices as exploratory.

Thanks for the advice. We acknowledged that the HOMA-IR was a composite index and compared it with the Matusda index in the 3.3 part in the revised manuscript. The result of the two index are similar. In fact, the FA induced both a milder glycemic response and a lower insulinemic response. It did not curb the glycemic at the expense of elevated insulin recruitment.

 

28.Line 351–354, please, correct Table 6 caption/labels (currently “insulinemic” while columns are glycemic metrics).

Thank you for pointing out the error. The correction has been made in the text (in line 400).

 

29.Line 351–381, please, harmonize captions/labels for Tables 6–7 (glycemic vs insulinemic) to match contents.

Thank you for pointing out the error. The correction has been made in the text (in line 426).

 

30.Line 421–426, please consider tempering causal language that in‑vitro viscosity “explains” PGR/PIR without direct mediational evidence。

Thank you for your suggestion. The wording has been revised accordingly. While indicators such as in vitro viscosity may correlate with blood glucose test results, this does not necessarily imply a causal relationship between the two.

 

31.Line 428–433, please, standardize statistical notation in figures/captions (e.g., “type×phase interaction”) and clarify what letters/symbols denote.

 

Thank you for your suggestion.，Type effect: a, b, c, d, used for comparison between types (within the phase) based on general linear model with Tukey adjustment (P<0.05). Phase effect: 1,2, used for comparison between phases (within the type) based on general linear model with Tukey adjustment (P<0.05). Type*phase: examines the mixed effects of type and phase based on general linear model with Tukey adjustment.

 

32.Line 435–470, I recommend, presenting viscosity, TPC/TFC, and other factors as candidate mediators rather than confirmed mechanisms.

Thank you for your suggestions. We agree that viscosity, TPC/TFC, and texture are all potential influencing factors, but they may not necessarily represent the underlying mechanisms driving changes in glycemic response. We have made revisions based on our discussions.

33.Line 451–457, please, soften phrasing from “contribute substantially” to “may contribute,” since mechanisms weren’t measured.

Thank you for pointing out the error. The correction has been made in the text (in line 500).

 

34. 34.Line 463–470, please consider qualifying statements about enzyme inhibition/transport modulation as hypotheses to be tested.

Thank you for your suggestion. The text has been revised to clarify that whether the inhibitory effect on the enzyme correlates with the experimental results remains a hypothesis requiring further experimental validation.

35.Line 471–479, I recommend, analytically separating matrix‑softening (cooking) from sugar‑profile and acidity effects (e.g., acidifying RA to CA pH; matched fructose:glucose boluses).

Thank you for your suggestion. Your insight is valuable. The matrix softening effect can be separated from factors such as the sugar profile and acid effects. Future studies could be designed to investigate whether altering the sugar composition or modifying the sugar-to-acid ratio within the same matrix conditions can adjust the postprandial glycemic characteristics of fruit products.

36.Line 481–484, it would be beneficial to continue the above sensitivity analyses to disentangle mechanisms.

Thank you for your suggestions. While our current data is insufficient to fully assess the relevant mechanisms, we can revise the discussion section to clarify the direction for future research on these mechanisms as much as possible.

37.Line 499–507, please, tighten preload conclusions: all apple preloads attenuated peaks vs water; between‑apple differences were modest and time‑window dependent.

Thank you for your suggestions. We will revise the wording of the conclusion to incorporate the detailed descriptions you mentioned.

 

38.Line 548–549, please, fix abbreviation typos (“frezz‑dried” → “freeze‑dried”) and ensure all abbreviations appear in text before the list.

Thank you for pointing out the error. The correction has been made in the text.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The paper was considerably improved and It can be accepted.