Does diabetes mellitus type 2 affect vestibular function? A systematic review and meta-analysis

• Published in: Diabetes epidemiology and management 2021-10, Vol.4, p.100035, Article 100035
• Main Authors: Bakkali, Samera El, Taeymans, Jan, Senior, Clara Osafo Sasu, Dirinck, Eveline, Vereeck, Luc, Vissers, Dirk
• Format: Article
• Language: English
• Subjects: Diabetes mellitus; Meta-analysis; Vestibular evoked myogenic potentials; Vestibular function; Video head impulse test
• ISSN: 2666-9706; 2666-9706
• DOI: 10.1016/j.deman.2021.100035

•Diabetes mellitus type 2 tends to result in a high prevalence of vestibular dysfunction as a complication.•A broad variance in prevalence of vestibular dysfunction in people with diabetes is seen depending on the used diagnostic tools and due to variable person and diabetes-related characteristics.•Significant differences in vestibular function between people with and without diabetes mellitus type 2 depend on the used objective vestibular function test. To explore the impact of diabetes mellitus type 2 (DMT2) on the vestibular system by differing in vestibular function (VF) results between people with and without DMT2. Relevant studies were identified through databases, including adults with DMT2 and controls. Only studies using objective VF tests were selected. For cervical vestibular evoked myogenic potential (cVEMP), ocular VEMP (oVEMP) and video Head Impulse Test (vHIT), results of both populations were compared using the random effects model. The prevalence of vestibular dysfunction (VD) was also considered. 13 and 8 studies for qualitative and quantitative analysis were included. Prevalence of VD ranged from 7.7% to 84% in DMT2 individuals and 0%–15% in controls. VHIT and most VEMP-parameters yielded non-significant differences, except for n-amplitudes during oVEMP testing (-2.44 µV; 95%CI:-4.15 to -0.74; p = 0.005), interamplitude values in both cVEMP (-1.18; 95%CI:-2.28 to -0.09; p = 0.034) and oVEMP tests (-0.74; 95%CI:-1.24 to -0.23; p = 0.004; (subgroup-analysis)), although high heterogeneity was present (I² > 75%). There is evidence of higher VD prevalence in people with DMT2 compared to controls, independently of the presence of symptoms. More standardized measuring methods is needed to draw firmer conclusions regarding detection of subclinical VD.