Misclassifying Diabetes: A Hidden Risk for British South Asians

Diabetes comes in two main types: type 1 diabetes (T1D) and type 2 diabetes (T2D). These types are caused by different factors, have distinct treatment needs, and lead to different experiences for those who have them. Doctors sometimes find it hard to spot which type a patient has because the signs can be similar. The rise in obesity among young people is making it even harder to tell these types apart. Moreover, some cases of type 1 diabetes develop later in life, which can lead to confusion as well.

Misclassification happens when a patient is incorrectly labeled as having T1D or T2D. Studies show that up to 13% of diabetes cases in the UK may be incorrectly coded. Identifying the right type of diabetes is often complicated because the signs of the two types can overlap. For instance, if someone needs Insulin soon after being diagnosed, it usually suggests T1D, but this can be hard to tell without proper tests.

#The Challenge of Misclassification

Doctors face difficulties in distinguishing between T1D and T2D, especially in populations with a high prevalence of T2D, such as those of South Asian descent. These individuals may present with diabetes at a younger age and may not always fit the typical profile seen in other populations. As a result, T1D might be misclassified as T2D, leading to severe health issues like diabetic ketoacidosis or increased long-term complications.

Other methods to understand diabetes types include genome-wide studies, which have provided insights into the genetic causes of T1D and T2D. These studies have developed polygenic risk scores (PRS) that may help with diabetes classification. However, many studies have focused on people of European descent, which raises questions about whether the findings apply to other ancestry groups.

#The Study Goal

This study looks to use genetic tools to better classify diabetes types among British South Asians. We want to assess how many individuals in this group with unclear diabetes features are actually T1D cases. Additionally, we want to see if the genetic scores relate to common clinical features used to determine diabetes types, such as age at diagnosis, body mass index (BMI), and time to insulin treatment.

#Study Population and Methods

Our research involved a large population of British South Asians living in East London. These individuals had given consent for their health records and genetic data to be used in research. We focused on those with available health data, yielding a sample of over 38,000 participants.

To ensure accurate data, we filtered the records for several key groups:

• T1D Cases: Individuals diagnosed with T1D, usually needing insulin soon after diagnosis.
• T2D Cases: Individuals with T2D who had been living with the condition for over three years and never needed insulin.
• Ambiguous Cases: Individuals under 60 who were treated with insulin but did not clearly fit either T1D or T2D classifications.
• Non-Diabetic Controls: Individuals with no diabetes diagnosis or related history.

We also looked closely at clinical features like BMI, age at diagnosis, and time before starting insulin treatment.

#Cleaning the Data

Before diving into the analysis, we cleaned the data to remove any outliers. For example, we ensured that weights and heights used to calculate BMI were reasonable. We excluded any extreme values that did not fit within standard medical ranges.

#Findings on Misclassification

We defined four groups based on health records and filtered them based on strict criteria. The filtering resulted in a small number of confirmed T1D cases, which led us to look at T1D data from a cohort in India. This combination helped create a clearer picture.

Using the genetic scores, we estimated that about 4.5% of individuals in the ambiguous group were likely misclassified as T2D instead of T1D. This low percentage is important because it highlights how many people could be getting the wrong treatment or missing out on life-saving care.

#Clinical Characteristics

When we examined the clinical features of the groups, we saw some significant differences. The ambiguous group had varying ages at diagnosis compared to the confirmed T1D and T2D cases. They also had differing levels of a key substance called C-peptide, which indicates how much insulin the body is producing.

Interestingly, many people in the ambiguous group were incorrectly coded as having T2D. This hints that doctors may be misclassifying cases based on limited information. We found that common clinical measures used to classify diabetes types didn’t correlate well with the genetic scores, further suggesting misclassification is likely.

#Importance of Accurate Classification

Identifying the correct type of diabetes is crucial, especially in populations with unique health patterns. If T1D is classified as T2D, patients might miss out on essential treatments that could help them manage their condition better. Insulin is a life-saving treatment for T1D, while for T2D, other medications may suffice.

Furthermore, misclassifying diabetes type could lead to dangerous outcomes, such as increased risk of diabetic emergencies or complications that could be avoided with proper treatment. The goal is to improve care for these individuals by accurately classifying their diabetes type.

#Conclusion

This study highlights the difficulties in correctly classifying diabetes among British South Asians. The findings suggest that a significant number of individuals may be misclassified, leading to potentially dangerous consequences. By using genetic tools alongside health data, we can enhance the precision of diabetes diagnosis and ensure people receive the most appropriate care.

Future research should focus on refining how we classify diabetes, particularly in diverse populations. With better tools and approaches, healthcare providers can deliver more effective and tailored treatments, ultimately improving the lives of many people living with diabetes.