Conclusions about Post-COVID Symptoms in Diabetics are Unreliable Unless any Bias Associated with the Study Design has been Removed

We were interested to read the article by Altynbekova et al. on an online survey on the nature and frequency of self-reported post-SARS-CoV-2 infection (SC2I) symptoms and the impact of SARS-CoV-2 vaccination (SC2V) on these symptoms in 212 type 2 diabetes patients from Kazakhstan [1] .More than half of the included patients reported persistent symptoms of SC2I, the most common of which were fatigue, exertional dyspnea, respiratory distress, headache, and insomnia that persisted >3 months after SC2I onset [1] .The study concluded that diabetics suffer SC2I symptoms longer than controls and that SC2V reduces the risk of hospitalization and the need for glucose correction in the post-COVID period [1] .The study is noteworthy, but several points should be discussed.The first point relates to the design of the study, which was an online survey of diabetics collected through an online invitation via social media to participate in the survey.Online surveys have several disadvantages.It is not easy to check whether the answers given are true, it cannot be ensured that the respondent is really the patient or another person who has access to the patient's email account, and it cannot be easily checked whether the patient really understood the questions correctly and was physically and mentally able to answer them sufficiently.The second point is that patients who do not have access to social media were automatically excluded from the study.Reasons for lack of access could be that they do not have a computer, are unable to use a computer, or are unable to complete the task.Therefore, we should know whether the country has universal internet access and whether there is an unequal distribution between rural and urban areas, between richer and poorer patients, between men and women, and between educated and uneducated patients.The third point is that it was not stated whether the control group consisted of healthy individuals or patients with diseases other than diabetes [1] .If it was a disease control group, we should know the current medication of the patients in the control group.The fourth point is that the title of the study is misleading because not all diabetic patients from Kazakhstan were included in the study, but only a small sample of 212 patients [1] .Since the total number of diabetics in Kazakhstan is estimated to be 380000, a sample size of 212 of these patients is not representative, so one cannot claim to be talking about the entire national cohort of diabetics.The fifth point is that the symptoms were not further investigated for the underlying disorder [1] .Since almost a quarter of the patients suffered from limb weakness, a quarter from postural tachycardia syndrome (POTS), a quarter from myalgias, 15% from paresthesias and ten percent from gait disturbances, it would have been imperative to screen the included patients for Guillain-Barre syndrome (GBS), which is a common complication of SC2I [2] , large fiber neuropathy and small fiber neuropathy, which may also represent PNS involvement in SC2I [3] .Since almost half of the patients reported headache, CNS imaging would have been essential to rule out sinus vein thrombosis, reversible cerebral vasoconstriction syndrome, subarachnoid hemorrhage, intracerebral hemorrhage, and infectious or immunologic encephalitis/meningitis [4] .In summary, it can be said that this interesting study has limitations that relativize the results and their interpretation.Removing these limitations could strengthen the conclusions and reinforce the message of the study.All unanswered questions need to be clarified before readers can uncritically accept the study's conclusions.Conclusions on post-acute COVID symptoms in diabetics may remain unreliable until all biases associated with the study design have been addressed.