Personalised Approach to the Management of Older People with Type 2 Diabetes Mellitus—A Comprehensive Narrative Review

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a review paper whose actual aim is rather unclear. In short, the paper lacks significant scientific value and does not present new information on research into type 2 diabetes in older adults. The majority of information stated in the paper is well known from available research and guidelines. None of the references is written in accordance with the rules for MDPI journals.

Author Response

Thank you for your comments on the manuscript. 

1. This is a review paper whose actual aim is rather unclear. In short, the paper lacks significant scientific value and does not present new information on research into type 2 diabetes in older adults. The majority of information stated in the paper is well known from available research and guidelines.

1. Although there is no new significant evidence in literature for the management of type 2 diabetes in older people, this manuscript provides a comprehensive update of the current evidence.  What the review adds to the existing literature is the new approach of management highlighting the heterogeneous nature of older people.  For frailty, it address the concept of metabolic phenotypes and the choice of hypoglycaemic therapy.  It also explores the new development in diabetes management-related technologies.  The review proposes a personalised approach based on biological-functional foundation rather than chronological age.  

To make this point clear, we have added a box (highlighted) to clarify what this review adds to the current literature.  We have also made changes throughout the manuscript to improve readability and reduce redundancies. 

2. None of the references is written in accordance with the rules for MDPI journals.

2. We will review the references format with the editorial board to comply with the MDPI journals. 

 

Reviewer 2 Report

Comments and Suggestions for Authors

Review of the manuscript entitled „Personalised approach to the management of older people with type 2 diabetes mellitus - A comprehensive review” (Manuscript ID: jpm-4175155)

 

This manuscript presents a broad narrative review addressing the management of type 2 diabetes in older adults, structured around heterogeneity of ageing, frailty phenotypes, cardiometabolic risk, hypoglycaemia vulnerability, and emerging technologies (CGM, mHealth, AI). The topic is highly relevant to the scope of Journal of Personalized Medicine, particularly given the increasing prevalence of multimorbidity and functional heterogeneity in ageing populations. The Authors aim to move beyond age-based management and propose a phenotype-driven framework integrating functional status, frailty subtype (sarcopenic obese vs anorexic malnourished), ASCVD risk, and dependency level into therapeutic decision-making. The manuscript is conceptually valuable and clinically important. However, its current structure is predominantly descriptive rather than critically synthetic. Methodological transparency, evidence stratification, redundancy reduction, and clearer personalization logic require strengthening to meet the standards of a SCOPUS-indexed narrative review in JPM.

Major Strengths

1. Clinically Relevant and Timely Topic: The intersection of ageing, frailty, and diabetes management is of high practical relevance. The manuscript appropriately recognises that chronological age alone is an insufficient determinant of therapeutic strategy.
2. Phenotype-Oriented Conceptual Framework: The distinction between sarcopenic obese (SO) and anorexic malnourished (AM) frailty phenotypes is conceptually interesting and clinically intuitive. This axis offers a biologically plausible foundation for differential therapeutic intensity.
3. Integration of Technology: The sections on CGM, mHealth, and AI appropriately reflect contemporary shifts in diabetes care. The inclusion of CGM metrics tailored to functional subgroups (Table 2) is particularly aligned with personalised medicine.
4. Emphasis on Hypoglycaemia Risk: The manuscript correctly frames hypoglycaemia as a dominant clinical threat in older adults and links it to frailty, cognitive decline, and institutionalisation. This is clinically grounded and appropriately prioritised.

Critical points

1. Even for a narrative review, minimal transparency is expected in JPM to avoid the appearance of selective synthesis. Without this clarification, the manuscript risks being perceived as expert opinion rather than structured scholarship. The Authors should explicitly clarify the databases searched (e.g., PubMed, Scopus, Cochrane), time frame covered, ehether guideline documents were systematically included, approximate number of screened and included sources, whether the review is narrative or semi-systematic?
2. Throughout the review, mechanistic reasoning, observational associations, and guideline-based recommendations are frequently presented in close proximity without explicit hierarchical separation. Could the Authors clarify whether specific statements are supported by:

• preclinical or mechanistic data,
• observational cohort analyses,
• randomised controlled trials (RCTs),
• formal guideline recommendations, or
• hypothesis-generating evidence?

3. In sections where RCT-level evidence is limited or absent, would more consistent use of cautious, evidence-aligned language (e.g., “may,” “suggests,” “is associated with”) better reflect the current state of the literature?
4. Several mechanistic and clinical themes are revisited across sections with partially overlapping language, particularly the frailty–insulin resistance relationship, hypoglycaemia vulnerability, ASCVD risk framing, functional decline associations, and CGM-related benefits. Would the Authors consider consolidating these into more focused mechanistic blocks and relying on cross-referencing rather than repetition?
5. Although the manuscript emphasises a “Personalised Approach,” the proposed stratification remains largely categorical (fit vs frail; independent vs dependent; SO vs AM). Could the Authors further operationalise this framework by introducing clearer decision algorithms, defining quantifiable thresholds (e.g., CFS score, BMI cut-offs), incorporating life expectancy estimation, and explicitly linking functional metrics to drug class selection?
6. Table 2 proposes CGM targets across independence levels. Are these targets evidence-based or expert consensus? How do they align with international CGM consensus (e.g., ATTD), and why is TIR ≥70% retained even in dependent patients? Is CV <36% validated in this specific population?
7. The AI section presents numerical performance claims (e.g., hospitalisation reduction, diagnostic accuracy, medication error reduction) with limited discussion of study heterogeneity, methodological quality, external validity, implementation barriers, algorithmic bias, and ethical considerations. The current tone appears aspirational. Given the scope of JPM, this section should be reframed as emerging and exploratory rather than clinically mature.
8. The recommendation of metformin + SGLT-2 inhibitor + GLP-1RA as first-line triple therapy in fit older adults with high ASCVD risk requires cautious positioning. Could the Authors clarify the consistency with current guideline sequencing, specific evidence in older populations, considerations regarding cost, tolerability, and polypharmacy.
9. Abbreviation density should be moderated, drug class formatting should be standardised.
10. Table 1 would benefit from explicit evidence-tier annotation.

General recommendation:

The manuscript addresses a highly relevant clinical domain and proposes a useful phenotype-oriented perspective. Its principal strengths lie in clinical breadth and recognition of heterogeneity. However, before being suitable for publication in Journal of Personalized Medicine, the manuscript requires clear methodological transparency, evidence stratification, reduction of redundancy, stronger personalization operationalisation, more cautious AI and technology framing, and more precise therapeutic positioning. With structural tightening and enhanced methodological discipline, the manuscript has the potential to become a valuable contribution to geriatric diabetology within a personalised medicine framework.

Author Response

Many thanks for your comments and suggestions to improve this manuscript.

Major Strengths

Many thanks for these comments.  We found it very useful to highlight the purpose and what this review adds to the literature.  We have therefore used it to create a new box: "What is new" to strengthen the manuscript, thank you. 

Critical points

1. Even for a narrative review, minimal transparency is expected in JPM to avoid the appearance of selective synthesis. Without this clarification, the manuscript risks being perceived as expert opinion rather than structured scholarship. The Authors should explicitly clarify the databases searched (e.g., PubMed, Scopus, Cochrane), time frame covered, ehether guideline documents were systematically included, approximate number of screened and included sources, whether the review is narrative or semi-systematic?

We have added a detailed Methods section (highlighted).

2. Throughout the review, mechanistic reasoning, observational associations, and guideline-based recommendations are frequently presented in close proximity without explicit hierarchical separation. Could the Authors clarify whether specific statements are supported by:

• preclinical or mechanistic data,
• observational cohort analyses,
• randomised controlled trials (RCTs),
• formal guideline recommendations, or
• hypothesis-generating evidence?

Thank you for this observation.  We have reviewed the whole manuscript taking into account these constructive comments. We feel that all the points listed above are directly applicable as older people, especially frail ones, are not commonly recruited to clinical trials. 

3. In sections where RCT-level evidence is limited or absent, would more consistent use of cautious, evidence-aligned language (e.g., “may,” “suggests,” “is associated with”) better reflect the current state of the literature?

Yes, we have reviewed the use of English to reflect the strength of evidence. We agree that this aspect of this review needs tightening and to this end, we have attempted to do this.

4. Several mechanistic and clinical themes are revisited across sections with partially overlapping language, particularly the frailty–insulin resistance relationship, hypoglycaemia vulnerability, ASCVD risk framing, functional decline associations, and CGM-related benefits. Would the Authors consider consolidating these into more focused mechanistic blocks and relying on cross-referencing rather than repetition?

This is a valid point and we have reviewed the whole manuscript to cut redundancies and repetitions. 

5. Although the manuscript emphasises a “Personalised Approach,” the proposed stratification remains largely categorical (fit vs frail; independent vs dependent; SO vs AM). Could the Authors further operationalise this framework by introducing clearer decision algorithms, defining quantifiable thresholds (e.g., CFS score, BMI cut-offs), incorporating life expectancy estimation, and explicitly linking functional metrics to drug class selection?

Thank you. Yes, it reads a bit categorical. We have reviewed the manuscript and tried our best to refine this personalised approach as possible. This will be a limitation, and our suggestions are based on our opinion due to the absence of clear evidence for each cut points of metrics such as CFS or BMI. Further research in this area should enable us to better define these exact metrics. We have suggested cut off points of BMI to define body weight. Still somehow categorical but we think is still practically useful in clinical practice. 

6. Table 2 proposes CGM targets across independence levels. Are these targets evidence-based or expert consensus? How do they align with international CGM consensus (e.g., ATTD), and why is TIR ≥70% retained even in dependent patients? Is CV <36% validated in this specific population?

Thank you for raising this point.  In fact we are following the ATTD recommendations of 70% TIR and <36% variability.  We have kept these time percent similar, to be in line with ATTD recommendations but changed the target ranges to suit older people categories.  The ATTD targets are tight and suit more younger people. 

The Table is our suggestion, rather than evidence based.  We feel that fixing the percent of time at 70% and variation at <36% (as ATTD recommends) and changing targets will avoid confusion in clinical practice.  

7. The AI section presents numerical performance claims (e.g., hospitalisation reduction, diagnostic accuracy, medication error reduction) with limited discussion of study heterogeneity, methodological quality, external validity, implementation barriers, algorithmic bias, and ethical considerations. The current tone appears aspirational. Given the scope of JPM, this section should be reframed as emerging and exploratory rather than clinically mature.

Thank you. Your comments are fully justified. We have therefore reviewed this section accordingly in the light of these comments.  

8. The recommendation of metformin + SGLT-2 inhibitor + GLP-1RA as first-line triple therapy in fit older adults with high ASCVD risk requires cautious positioning. Could the Authors clarify the consistency with current guideline sequencing, specific evidence in older populations, considerations regarding cost, tolerability, and polypharmacy.

Thank you. We have reviewed this section. We have referred to the guideline recommendations for triple therapy in people with high ASCVD risk according to the ADA.  We have also addressed the important points you have raised regarding tolerability, polypharmacy, etc. 

Our point was to highlight that in the SO frail population, although frail, they still benefit from triple therapy (if tolerated) due to their high ASCVD risk as compared with the AM frail. In other words, regarding the choice of therapy, we feel that frailty should not be seen as a single category of patient. 

9. Abbreviation density should be moderated, drug class formatting should be standardised.

Agreed.

10. Table 1 would benefit from explicit evidence-tier annotation.

Agreed and actioned. We have expanded in the text and added new references to support the data in the table. We have kept the table with no change for ease or reading as a summary of the text.  

General recommendation:

The manuscript addresses a highly relevant clinical domain and proposes a useful phenotype-oriented perspective. Its principal strengths lie in clinical breadth and recognition of heterogeneity. However, before being suitable for publication in Journal of Personalized Medicine, the manuscript requires clear methodological transparency, evidence stratification, reduction of redundancy, stronger personalization operationalisation, more cautious AI and technology framing, and more precise therapeutic positioning. With structural tightening and enhanced methodological discipline, the manuscript has the potential to become a valuable contribution to geriatric diabetology within a personalised medicine framework.

Thank you for your excellent suggestions. We have tried our best to address all these points and feel that the manuscript has improved considerably as a result - thank you.  

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The authors highlighted what this manuscript adds to the current knowledge in this field. Also, the manuscript has been significantly improved. Table 1 is a useful summary of medications and risks in the elderly. Although much of the information in this manuscript is well known, it integrates the essential information for treating diabetes in the elderly and can be accepted for publication once the authors have written the references in accordance with MDPI guidelines (at present, all references are written in the Vancouver style). 

Author Response

Thank you for your comments. We are pleased that you feel that the manuscript is now acceptable for publication. References will be adjusted to the JPM guidelines in due course. 

 

Reviewer 2 Report

Comments and Suggestions for Authors

Review of the manuscript entitled „Personalised approach to the management of older people with type 2 diabetes mellitus - A comprehensive review” (Manuscript ID: jpm-4175155) – revised version

 

The revised manuscript has clearly improved in structure and readability, and the addition of a dedicated „Methods” section and refinement of language are important steps forward. The topic is clinically highly relevant, and the phenotype-oriented perspective remains a valuable conceptual contribution.

However, despite these improvements, several key methodological and conceptual issues remain insufficiently resolved. Addressing these points is necessary before the manuscript may be considered for acceptance:

 

Points requiring revision:

1. Methodological transparency remains incomplete. Although a Methods section has been added, the review still lacks quantitative elements of transparency (e.g., approximate number of screened and included studies, selection logic, and clearer definition of the review type). This limits reproducibility and risks the perception of selective synthesis.
2. Lack of explicit evidence hierarchy. Mechanistic insights, observational data, RCT evidence, and guideline-based recommendations are still presented without consistent differentiation. A clear stratification of evidence levels throughout the manuscript is required to ensure scientific rigour.
3. Overstatement of therapeutic recommendations. Several statements (particularly regarding early or “first-line” triple therapy) remain insufficiently cautious relative to the available evidence in older populations. More consistent use of evidence-aligned language (“may,” “can be considered,” “based on extrapolated data”) is necessary.
4. Personalisation framework remains insufficiently operationalised. While conceptually strong, the proposed phenotype-based approach is still largely descriptive. Integration of more explicit clinical decision elements (e.g., functional scores, life expectancy considerations, or algorithmic guidance) would substantially improve applicability.
5. CGM targets lack adequate evidentiary justification. The proposed CGM thresholds appear to be largely expert-driven without clear validation in the studied population. This section requires either stronger evidence support or explicit framing as consensus-based guidance.
6. Tables lack evidence-level annotation. Table 1 (and related summaries) would benefit from explicit indication of evidence strength to align with the narrative discussion and improve interpretability.
7. AI and digital health section requires further critical balance. Although improved, the discussion remains somewhat optimistic. Additional consideration of methodological limitations, external validity, and implementation barriers would strengthen this section.
8. Supplementary material: The submitted supplementary file primarily reflects textual revisions rather than providing additional methodological or evidentiary support, and therefore does not substantially strengthen the scientific robustness of the manuscript. To enhance transparency and clinical utility, the supplementary material should be expanded to include a structured overview of the literature selection process (e.g., simplified PRISMA-style flow), an explicit evidence hierarchy framework distinguishing between mechanistic, observational, RCT and guideline-based data, a clear justification of the proposed CGM targets in relation to existing consensus (e.g., ATTD) versus expert interpretation, and a schematic decision algorithm operationalising the personalised approach. In addition, a brief critical appraisal of emerging areas such as AI and digital health would further improve balance. These additions would meaningfully increase both the methodological credibility and practical applicability of the review.

 

Summary recommendation:

The manuscript is clinically meaningful and conceptually valuable, but requires further structural tightening and clearer evidence-based positioning before it can be considered suitable for publication. With these revisions, it has strong potential to become a relevant contribution to the field of personalised diabetology in older populations.

Author Response

Thank you for your further comments.  We spent a considerable amount of time attempting to respond to your previous comments.  They were indeed constructive on the whole, but you may consider the lack of significant clinical trial evidence available and our genuine attempt to present findings often based on our real world clinical experience in this field. Nevertheless, we appreciate your significant interest in our work and your comments to allow us to enhance it.  We have done our best to address them as possible.

1. Methodological transparency remains incomplete. Although a Methods section has been added, the review still lacks quantitative elements of transparency (e.g., approximate number of screened and included studies, selection logic, and clearer definition of the review type). This limits reproducibility and risks the perception of selective synthesis.

We understand and acknowledge to an extent your comment, but we have now made it clear in the Title and in the Methods section that the manuscript represents a narrative review only. It represents our attempt to summarise, interpret from the existing literature as well as identify gaps. It was not an attempt to list the number of studies. We feel we have provided a comprehensive search, outlined key databases and key words, and a search strategy, and have provided a balanced perspective to our findings. 

As such, we believe that we have provided an appropriate methods section, and we hope that this is acceptable.   

2. Lack of explicit evidence hierarchy. Mechanistic insights, observational data, RCT evidence, and guideline-based recommendations are still presented without consistent differentiation. A clear stratification of evidence levels throughout the manuscript is required to ensure scientific rigour.

We have addressed this point previously with explicitly mentioning the types of each study we refer to across the manuscript. We appreciate the level of evidence varies according to type of study. Although we did not formally critically appraise each study (as in a systematic review), we have stated that, this is a narrative review. We have shown what type and design each study has and the level of evidence stated. The manuscript has wide references base (160 ref), which will be outside the scope of such broad review to address each study in detail.

3. Overstatement of therapeutic recommendations. Several statements (particularly regarding early or “first-line” triple therapy) remain insufficiently cautious relative to the available evidence in older populations. More consistent use of evidence-aligned language (“may,” “can be considered,” “based on extrapolated data”) is necessary.

Thank you. We have addressed this point and similar points and soften/modified our language as possible to address your concerns. We added this explicitly in the text (highlighted). 

4. Personalisation framework remains insufficiently operationalised. While conceptually strong, the proposed phenotype-based approach is still largely descriptive. Integration of more explicit clinical decision elements (e.g., functional scores, life expectancy considerations, or algorithmic guidance) would substantially improve applicability.

Yes, this is a valid point mentioned previously. We have discussed this between us several times. We need more clinical trial evidence to suggest a specific algorithm based on specific metrics in literature. We have added the BMI as a possible simple metric but the message we give to clinical practice at this stage remains clinical guidance on categorisation of patients with certain phenotypic criteria ( as you previously mentioned, this is more categorical than individual/personal), due to lack of sufficient evidence at this stage. We have previously published our descriptive data in this area – please see Metabolic Characteristics of Frail Older People with Diabetes Mellitus – A systematic search for phenotypes. Abdelhafiz AH, et al, Mertabolites 2023; May 29; 13(6): 705

5. CGM targets lack adequate evidentiary justification. The proposed CGM thresholds appear to be largely expert-driven without clear validation in the studied population. This section requires either stronger evidence support or explicit framing as consensus-based guidance.

Yes, it is our suggestion based on expert-opinion and the available evidence. We have mentioned this explicitly in text (highlighted) and in the footnote of the table. Also, the current recommendation of CGM in younger people are based on consensus and expert opinion. We wish to point out that no agreed and consensus - validated data on CGM targets in older people exist in the literature; two of the authors (AJS and AH) are members of an international consensus group on CGM metrics in older people with diabetes who will be presenting their recommendations in the fall.

6. Tables lack evidence-level annotation. Table 1 (and related summaries) would benefit from explicit indication of evidence strength to align with the narrative discussion and improve interpretability.

Table 1 is a summary of the current evidence from the studies. We have left it less crowded for ease of reading and reference. We have expanded on studies supporting the information in the table in the text. The evidence from these studies is limited due to lack of data and we have placed all the available information in text. Again, we feel, it is inappropriate to expand on critically appraising these studies as it is beyond the scope of our manuscript to expand only on one aspect of such a narrative/broad review. We have highlighted the limitations of clinical trials in this area in text (highlighted). 

We believe the table is useful addition as no similar tables in literature exist addressing effects of hypoglycaemic therapy on outcomes relevant to older people such as dementia, depression and frailty. 

7. AI and digital health section requires further critical balance. Although improved, the discussion remains somewhat optimistic. Additional consideration of methodological limitations, external validity, and implementation barriers would strengthen this section.

We have reviewed this again and modified the discussion and softened language as much as possible without altering conclusions.  

8. Supplementary material

We have made it clear that our manuscript is a narrative review. As explained and stated above, we believe the current methods section is appropriate for a broad/comprehensive  review of this type. 

Summary recommendation

Thank you.