Introduction Limited joint mobility (LJM) has been linked to deficient glycaemic control but is an understudied area of type 2 diabetes research. We set out to investigate the correlation between glycated haemoglobin (HbA1c) and the quantification of LJM of finger joints and non-invasive anthropometrics.
Methods Blood samples were taken from 170 participants at diabetes awareness drives in Trinidad. These participants were aged 59.61 ± 15.46, with a body mass index (BMI) of 29.73 ± 7.65 and HbA1c levels of 8.42 ± 2.22. There were 110 women and 60 men. Finger joint angles were recorded using a goniometer.
Results The K-Nearest Neighbour machine learning model was tested via 10-fold cross validation to differentiate good from poor glycaemic control (HbA1c ≤ 6.5%) using non-invasive features. There is some correlation between LJM and HbA1c. Our model scored a mean accuracy of 74.71% ± 1.81 (p=0.01) classifying the full dataset, 82.14% ± 2.20 (p=0.01) and 72.76% ± 1.41 (p=0.059) on the male/female subsets, respectively.
Discussion The time since diagnosis, age and BMI were important features linked to glucose control. Our results support the notion that the first signs of LJM in the fingers occur in the first and fifth fingers as these particular angles were ranked highly in the list of most important features.
Conclusion Our results show that LJM has some role to play in monitoring HbA1c, although not as important as more conventional anthropometrics. Our results support the idea that there should be a separate test for each sex.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
SR and AD contributed equally.
Contributors AD and BC created the hypotheses to be tested and collected the data with the help of the Diabetes Association of Trinidad and Tobago. SR performed the literature review, data analysis, and applied machine learning to generate the results. SR and AD wrote the paper. AJ and BC aided in analysing the results and the overall supervision of the work.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval We received ethics approval from the committee located in the Faculty of Medical Sciences at The University of the West Indies at Saint Augustine.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the article or uploaded as supplemental information. All deidentified participant data used are available in supplemental Table 1.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.