Reduced Tissue Oxygenation and Altered Valsalva Hemodynamics in Young Adults with Type 1 Diabetes

Cardiovascular autonomic neuropathy (CAN) is a common microvascular neuropathic condition among individuals with type 1 diabetes (T1D). According to epidemiological studies, the prevalence of CAN ranges from 36 to 44 %, and they are commonly associated with reduced quality of life^{1, 2, 3}. Alarmingly, CAN has been observed in adolescents with T1D, with up to 32% showing early signs of CAN^{4, 5}. This highlights the early onset and clinical relevance of this complication.

The prevailing understanding of the development of diabetic neuropathy is that it is length dependent, affecting e.g. sensorimotor nerves in the extremities first, or the relatively long vagus nerve⁶. Affection of the vagus nerve manifests as parasympathetic withdrawal, which is considered a potentially reversible stage of autonomic dysfunction^{7, 8}. The gold standard for diagnosing CAN is the Ewing's test battery, also known as cardiovascular autonomic reflex tests (CARTs)⁹, where especially the expiration-to-inspiration (E:I) ratio is primarily parasympathetically driven¹⁰. However, its utility in detecting parasympathetic withdrawal may be limited by methodological inconsistencies¹¹. These include variations in body position (supine vs. seated), the relative duration of expiration to inspiration to ranging from 1:1 to 3:1, and how the E:I ratio is calculated, either by averaging the longest and shortest RR-intervals from each breathing cycle or by selecting the single longest and shortest RR-intervals across all cycles. Together, these methodological variations reduce reliability and increase the risk of misclassification, particularly in early, subtle, or borderline cases.

Although the standardized CARTs evaluate both RR-interval and blood pressure responses, clinical applicable practice increasingly emphasizes solely on RR-interval-based measures, potentially overlooking vascular abnormalities. The endothelium plays a critical role in the regulation of vascular tone, and studies have reported endothelial dysfunction even in uncomplicated T1D¹². This can contribute to altered vascular reactivity, which may alter hemodynamic responses during autonomic challenges, particularly those requiring rapid vasomotor adjustments, such as the Valsalva manoeuvre. Altered microvasculature with localized tissue hypoxia, might represent one of the earliest pathophysiological steps in the development of diabetic neuropathy¹³, and could serve as an early indicator of autonomic dysfunction, potentially preceding detectable changes in traditional nerve-based measures such as cardiac adaptability assessed as RR-interval responses.

We hypothesize that the tissue metabolism is altered in T1D leading to increased use of oxygen compared to controls. Thus, the purpose of this study was to assess cardiovascular autonomic function (ECG/RR-interval and blood pressure responses) and tissue metabolism of transcutaneous oxygen and carbon dioxide, as well as their associations, in younger adults with T1D compared with healthy controls.

The study was approved by the local Ethics Committee in the North Denmark Region (N-20230033) and informed consent was collected. It was registered with the Regional Data Security Authority (ID no. F2023-097).

Individuals with T1D without autonomic or peripheral neuropathy exhibited reduced tissue oxygenation despite normal systemic oxygen saturation. During the Valsalva’s manoeuvre, the individuals with T1D showed reduced systolic blood pressure response in phase I and early phase II, as well as exaggerated response in both systolic and diastolic blood pressure in late phase II. The exaggerated systolic blood pressure in late phase II responses may reflect a hyperadrenergic state with enhanced splanchnic autotransfusion. These findings point to possible sympatho-vagal imbalance, despite normal results on standard autonomic function tests.

Chronic hyperglycemia sets off a cascade of events leading to pathological alterations in neurons, glia cells and microvasculature, ultimately leading to nerve dysfunction, and progressive tissue hypoxia is likely to exacerbate this. Thereby compromising tissue perfusion, promoting oxidative stress and inflammation, over time leading to ischemia of the surrounding tissue. Structural vascular changes, including basement membrane thickening and altered capillary density, support the presence of diabetes-induced nerve ischemia¹⁹. In the present study, individuals with T1D exhibited reduced tcpO₂ despite normal systemic oxygen saturation, suggesting altered tissue oxygen metabolism, low tissue perfusion, or a combination. These findings imply that, in T1D, tissue requires higher oxygen availability to achieve the same functional or metabolic effect as observed in controls, or alternatively, reduced tissue perfusion. The borderline elevation in tcpCO₂ may indicate increased local CO₂ production, potentially reflecting a shift toward greater lipid oxidation compared with controls. These findings are consistent with the results of Breuer 1988²⁰. The presence of reduced tissue oxygenation with normal CARTs ratio aligns with the presumption that microvascular changes may develop early in the disease course, often within the first five years of T1D^{6, 21}. Since reduced tissue oxygenation has been demonstrated to be reversible with intensified glycaemic control in type 2 diabetes, identification of these early changes could be valuable in the attempt to prevent long-term damage²². Preclinical models have been able to reverse neuropathy with an angiogenic growth factor¹³.

The integrity of the central parasympathetic regulation was investigated with carotid massage and deep breathing. Carotid massage activates baroreceptors within the carotid sinus, which relay a ‘false’ signal to the nucleus tractus solitarius (NTS) in the brainstem via the glossopharyngeal nerve (Cranial nerve IX). This stimulation triggers enhanced parasympathetic output, leading to a reduction in blood pressure. The study observed comparable responses in both controls and individuals with T1D, indicating intact baroreflex function and central parasympathetic regulation across both groups. To our knowledge, this is the first formal investigation of carotid massage response in this context, and no previous comparative studies are available.

Physiologically, deep breathing exaggerates respiratory sinus arrhythmia through vagal modulation of heart rate. Lung inflation activates pulmonary stretch receptors that inhibit vagal outflow via brainstem nuclei, increasing heart rate during inspiration, whereas vagal tone increases during expiration, resulting in a cyclical heart rate pattern. Notably, we observed a similarly pronounced increase in CVT in both groups, indicating preserved central parasympathetic function, likely attributable to the participants’ relatively young age and diabetes duration. When examining the effect on peripheral microcirculation, there were no differences in overall gaseous influx and efflux, suggesting that circulation in terms of peripheral gaseous exchange remained intact. However, the T1D group continued to exhibit the microvascular abnormalities observed at baseline during deep breathing, specifically a lower maximum tcpO₂ and higher minimum tcpCO₂ which is likely to be a metabolic effect. The failure to normalize these gaseous levels during parasympathetic stimulation suggests a reduced compensatory capacity in individuals with T1D.

The integrity of the sympathetic nervous system was assessed by evaluating the participants' ability to maintain orthostasis and tolerate physical exhaustion with a 3-minute sympathetic excitation during the handgrip test. Upon standing, blood pressure initially drops and is normally corrected by sympathetic vasoconstrictor activation. Orthostatic hypotension, common in diabetes, reflects sympathetic failure^{4, 23}. In this study, the individuals with T1D responded similarly to the controls, indicating preserved sympathetic tone, underpinning the young cohort. However, the individuals with T1D exhibited consistently lower tcpO₂ both before and at the end of the handgrip test, alongside higher tcpCO₂, compared to controls. These findings suggest that although oxygen delivery improved, the underlying reduction in tissue oxygenation in T1D persisted. This demonstrates a reduced capacity to normalize tissue oxygenation under stress. But it can also be due to a change in metabolic substrates to those with low respiratory quotient, for example fats and proteins.

The Valsalva’s manoeuvre has five distinct phases of blood pressure changes with characteristic changes in the hemodynamics. This study found abnormal hemodynamics in T1D compared to the controls, despite comparable Valsalva RR-ratio responses. Notably, some of the controls had atypical hemodynamic responses, which are not uncommon and reflect normal physiological variability. The manoeuvre begins with increased intrathoracic pressure from forced exhalation, resulting in a transient rise in blood pressure (phase I), and the observed reduced systolic rise in blood pressure suggests that arterial elasticity is impaired, leading to diminished Windkessel function and thus less effective transmission of intrathoracic pressure. Arterial stiffness is a common feature of diabetes^{24, 25}. The increase in intrathoracic pressure in this phase then reduces venous return to the heart leading to a drop in blood pressure during early phase II. The blunted drop in systolic blood pressure in T1D likely reflects impaired blood redistribution to the peripheral circulation due to higher baseline vascular resistance, although a contribution from reduced stroke volume cannot be ruled out, it is considered less likely given the young age of the participants²⁶. As a result, less blood is sequestered outside the thoracic compartment during the strain, maintaining venous return and blunting the typical blood pressure fall. In normal physiology, this blood pressure drop is counteracted in late phase II by venous volume baroreflex activation and increased sympathetic tone to stabilise blood pressure. However, we observed elevated systolic and diastolic response in late phase II in the individuals with T1D. This pattern suggests exaggerated splanchnic auto-transfusion²⁷. The alpha-1 adrenoreceptors mediate vasoconstriction in both the splanchnic and peripheral vasculature, increasing venous return and systemic vascular resistance, which together contribute to the heightened blood pressure during this phase. The hyperadrenergic state is a known phenomenon in postural tachycardia syndrome and long-COVID^{28, 29}.

Diastolic blood pressure is related to changes in systemic vascular resistance mainly in the Windkessel vessels and its response during early phase II depends not only on intrathoracic pressure-induced reductions in preload, but also on the ability of blood to redistribute to the peripheral circulation to reduce the cardiac afterload. In our study, the diastolic blood pressure drop in early phase II was inversely correlated with baseline tcpO₂ indicating that low tissue oxygenation is associated with a blunted diastolic response. As tcpO₂ reflects the balance between tissue perfusion and oxygen consumption, lower tcpO₂ may indicate both increased metabolic demand and higher vascular tone. The blunted response may reflect a combined effect of increased vascular tone and reduced Windkessel vessel conductance, limiting peripheral blood pooling and thereby maintaining the central venous return. While our findings in the individuals with T1D reveal clear phase-specific abnormalities, a previous study in type 2 diabetes reported no such hemodynamic changes³⁰. However, that study involved a considerably older cohort, in whom age-related changes in cardiovascular and autonomic function would be expected.

We used the traditional RR-based ratios to categorize individuals into normal and abnormal CAN status, and noticeably, nearly all, regardless of diabetes status, greatly outperformed the age-adjusted reference values. These reference values are designed to detect overt dysfunction, and as such, offer limited sensitivity for identifying early or subclinical changes. Rather than relying solely on dichotomized CAN classification, examining individual-level changes in RR-based ratios may complement a more nuanced insight into the autonomic regulation that can reveal subtle reductions in autonomic adaptability that are not captured by binary thresholds. However, such individualized assessments are not yet suitable for routine clinical application.

Moreover, when comparing the mean E:I ratio to the max E:I ratio, the mean is 14% lower. This suggest that habituation occur during the one-minute breathing cycle. Notably, previous studies have varied in their approach: some report the mean E:I ratio, others calculate the ratio using the mean of the three longest and three shortest RR-intervals, and some use the single maximum RR-interval divided by the minimum to capture peak vagal modulation^{31, 32, 33, 34}. The degree of habituation differs between individuals. Therefore, given an average habituation of 14% and age-based staging thresholds that vary by only 2.7-4% per decade¹⁸, relying solely on the mean E:I ratio may underestimate autonomic function in individuals who exhibit strong habituation, potentially misclassifying several individuals. Thus, it is questionable to diagnose CAN based exclusively on the mean ratio, as the individual may simply habituate well to the test. Habituation may decrease with age, but since individual capacity cannot be predicted, it may be simpler, easier to translate between settings and more accurate to use the maximum E:I ratio when assessing parasympathetic function.

A major advantage of this study is the use of comprehensive and advanced methodology than standard clinical assessments, combining high-frequency ECG (5000 Hz) with beat-to-beat blood pressure monitoring with transcutaneous gas profiling. This multimodal approach captured subtle physiological changes that would be missed by traditional test equipment. Unlike traditional assessments of endothelial function, such as flow-mediated dilation or peripheral arterial tonometry, which primarily reflect large-vessel physiology, transcutaneous measurements capture capillary-level function and the tissue consequences of early microvascular changes dysfunction. There is a wide range of reported tcpO₂ values in the literature, and our measurements fall within the lower end of this range for both controls and individuals with T1D^{35, 36}. This may reflect methodological differences or potential device-related variation; therefore, we refrain from interpreting these findings as evidence of tissue hypoxia. Moreover, a relatively high prevalence of mild retinopathy in the cohort may limit generalizability and introduces a potential bias toward early-stage microvascular involvement. The relatively high mean HbA1c of this cohort with T1D may be related to the widespread use of continuous glucose monitoring systems and insulin pumps, allowing for more flexible glycaemic management rather than strict dietary restriction. Although circulating catecholamine measurements were not included in the present study, future studies incorporating plasma catecholamines could confirm the proposed hyperadrenergic response during the Valsalva manoeuvre. The controls were not selected based on normal autonomic testing, and some variation within the controls is to be expected. The limited sample size of this study affects generalisability, and thus, findings should be validated in larger cohorts. Longitudinal studies are also needed to investigate whether reduced tissue oxygenation follows the progression of autonomic dysfunction in T1D.

In conclusion, with the methodological limitations mentioned above, we show that individuals with T1D without autonomic or neuropathic neuropathy exhibited reduced tissue oxygenation despite normal systemic oxygen saturation, which may suggest altered tissue oxygen metabolism or delivery. Although conventional autonomic function tests showed preserved functions, altered hemodynamic responses during phase I and II of the Valsalva’s manoeuvre may reflect possible changes in the arterial stiffness, vascular resistance and sympatho-vagal balance in the splanchnic region, consistent with a hyperadrenergic state.

MBP: project administration, conceptualization, investigation, formal analysis, visualization, writing – original draft. PJ: methodology, conceptualization, software, writing – review & editing. MeS: investigation, writing – review & editing. MuS: conceptualization, validation, writing – review & editing. CB: conceptualization, resources, writing – review & editing.

This study was supported by a grant from A.P. Møller Mærsk Fonden.

Date and code are available upon reasonable request

1. 1.Braffett B H, Gubitosi-Klug R A, Albers J W. (2020) Risk Factors for Diabetic Peripheral Neuropathy and Cardiovascular Autonomic Neuropathy. in the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study.Diabetes 69(5), 1000-1010.
   Google Scholar

1. 2.Kempler P, Tesfaye S, Chaturvedi N. (2002) Autonomic neuropathy is associated with increased cardiovascular risk factors:. , the EURODIAB IDDM Complications Study.Diabet Med 900-909.
   Google Scholar

1. 3.Morsby S K, Bitsch Poulsen M, Mark E B. (2025) . The Symptom Burden of Autonomic Neuropathy Is Associated with Decreased Quality of Life in 6961 People with Diabetes.Diabetology 6-11.
   Google Scholar

1. 4.Rasmussen V F, Thrysøe M, Nyengaard J R. (2023) Neuropathy in adolescents with type 1 diabetes: Confirmatory diagnostic tests, bedside tests, and risk factors.Diabetes Res ClinPract. 201-10.
   Google Scholar

1. 5.Jaiswal M, Divers J, Urbina E M. (2018) Cardiovascular autonomic neuropathy in adolescents and young adults with type 1 and type 2 diabetes: The SEARCH for Diabetes in Youth Cohort Study.PediatrDiabetes. 19(4), 680-689.
   Google Scholar

1. 6.Feldman E L, Callaghan B C, Pop-Busui R. (2019) Diabetic neuropathy.Nat Rev Dis Primers. 5-1.
   Google Scholar

1. 7.Andersen S T, Witte D R, Fleischer J. (2018) Risk factors for the presence and progression of cardiovascular autonomic neuropathy in type 2 diabetes: Addition-Denmark.Diabetes Care. 41(12), 2586-2594.
   Google Scholar

1. 8.Gæde P, Oellgaard J, Carstensen B. (2016) Years of life gained by multifactorial intervention in patients with type 2 diabetes mellitus and microalbuminuria: 21 years follow-up on the Steno-2 randomised trial.Diabetologia. 59(11), 2298-2307.
   Google Scholar

1. 9.Ewing D J, Campbell I W, Clarke B F. (1980) Assessment of cardiovascular effects in diabetic autonomic neuropathy and prognostic implications.Ann Intern Med. , 92(2 Pt 2, 308-311.
   Google Scholar

1. 10.Julu P O, Hondo R G. (1992) Effects of atropine on autonomic indices based on electrocardiographic R-R intervals in healthy volunteers.J Neurol Neurosurg Psychiatry. 55(1), 31-35.
   Google Scholar

1. 11.Rasmussen T K, Hansen J, Low P A. (2017) Autonomic function testing: Compliance and consequences.AutonNeurosci. 208-150.
   Google Scholar

1. 12.Lespagnol E, Dauchet L, Pawlak-Chaouch M. (2020) Early endothelial dysfunction in type 1 diabetes is accompanied by an impairment of vascular smooth muscle function: A meta-analysis.Front Endocrinol (Lausanne). 11, 525054-10.
   Google Scholar

1. 13.Schratzberger P, Walter D H, Rittig K. (2001) Reversal of experimental diabetic neuropathy by VEGF gene transfer.J Clin Invest. 107(9), 1083-1092.
   Google Scholar

1. 14.. (2021).Postural Tachycardia Syndrome Gall N, Kavi L, Lobo MD, eds 10-1007.
   Google Scholar

1. 15.Brinth L, Pors K, Mehlsn J, Sletten D M, Terkelsen A J et al. (2021) Translation and linguistic validation. of the Composite Autonomic Symptom Score COMPASS 31 in Danish.Dan Med 69-3.
   Google Scholar

1. 16.Sletten D M, Suarez G A, Low P A, Mandrekar J, Singer W. (2012) COMPASS 31: A refined and abbreviated composite autonomic symptom score.Mayo Clin Proc. 87(12), 1196-1201.
   Google Scholar

1. 17.POO Julu, Hansen S, Jamal G A. (1996) Is baroreflex responsiveness (BRR) a function of height during isometric muscle contraction in human volunteers.Journal of Physiology. 491.
   Google Scholar

1. 18.Spallone V, Ziegler D, Freeman R. (2011) Cardiovascular autonomic neuropathy in diabetes: clinical impact, assessment, diagnosis, and management.DiabetesMetabRes Rev. 27(7), 639-653.
   Google Scholar

1. 19.Nowicki M, Kosacka J, Serke H, Blüher M, Spanel‐Borowski K. (2012) Altered sciatic nerve fiber morphology and endoneural microvessels in mouse models relevant for obesity, peripheral diabetic polyneuropathy, and the metabolic syndrome.JNeurosciRes. 90(1), 122-131.
   Google Scholar

1. 20.Breuer H W, Breuer J, Berger M. (1988) Transcutaneous oxygen pressure measurements in type I diabetic patients for early detection of functional diabetic microangiopathy.EurJ Clin Invest. 18(5), 454-459.
   Google Scholar

1. 21.Cé G V, Rohde L E, Silva AMV Da, MKP Coutinho, De Castro AC et al. (2011) Endothelial Dysfunction Is Related to Poor Glycemic Control. in Adolescents with Type 1 Diabetes under 5 Years of Disease: Evidence of Metabolic Memory.J Clin EndocrinolMetab 96(5), 1493-1499.
   Google Scholar

1. 22.Iino K, Yoshinari M, Doi Y, Shinohara N, Iwase M et al. (1997) Reduced tissue oxygenation and its reversibility by glycemic control in diabetic patients. Diabetes Res Clin Pract. 34(3), 163-8.
   Google Scholar

1. 23.Zhou Y, Ke S J, Qiu X P, Liu L Bin. (2017) Prevalence, risk factors, and prognosis of orthostatic hypotension in diabetic patients: A systematic review and meta-analysis.Medicine. 96-36.
   Google Scholar

1. 24.Kimoto E, Shoji T, Shinohara K. (2003) . Preferential Stiffening of Central Over Peripheral Arteries in Type 2 Diabetes.Diabetes 52(2), 448-452.
   Google Scholar

1. 25.Love K M, Horton W B, Patrie J T, Jahn L A, Hartline L M et al. (2022) Predictors of arterial stiffness in adolescents and adults with type 1 diabetes: a cross-sectional study.BMJ Open Diabetes Res Care. 10-1.
   Google Scholar

1. 26.Romney J S, Lewanczuk R Z. (2001) Vascular Compliance Is Reduced in the Early Stages of Type 1 Diabetes.Diabetes Care. 24(12), 2102-2106.
   Google Scholar

1. 27.Sandroni P, Benarroch E E, Low P A. (1991) Pharmacological dissection of components of the Valsalva maneuver in adrenergic failure.J Appl Physiol. 71(4), 1563-1567.
   Google Scholar

1. 28.Ocher R A, Padilla E, Hsu J C, Taub P R. (2021) . Clinical and Laboratory Improvement in Hyperadrenergic Postural Orthostatic Tachycardia Syndrome (POTS) after COVID-19 Infection.Case RepCardiol 1-3.
   Google Scholar

1. 29.Okamoto L E, Urechie V, Rigo S. (2024) Hyperadrenergic Postural Tachycardia Syndrome: Clinical Biomarkers and Response to Guanfacine.Hypertension. 81(11), 2237-2247.
   Google Scholar

1. 30.Rasmussen T K, Finnerup N B, Singer W, Jensen T S, Hansen J et al. (2022) Preferential impairment of parasympathetic autonomic function in type 2 diabetes.AutonNeurosci. 243-10.
   Google Scholar

1. 31.Hansen C S, MMB Christensen, Vistisen D. (2024) Normative data on measures of cardiovascular autonomic neuropathy and the effect of pretest conditions in a large Danish non-diabetic CVD-free population from the Lolland-Falster Health Study.Clinical Autonomic Research. 35(1), 101-10.
   Google Scholar

1. 32.Spallone V, Bellavere F, Scionti L. (2011) Recommendations for the use of cardiovascular tests in diagnosing diabetic autonomic neuropathy.Nutrition,Metabolism and Cardiovascular Diseases. 21(1), 69-78.
   Google Scholar

1. 33.LRM Tannus, Sperandei S, Montenegro Júnior RM. (2013) Reproducibility of methods used for the assessment of autonomous nervous system’s function.AutonNeurosci. 177(2), 275-279.
   Google Scholar

1. 34.Mathias C J, Corazza I, Guaraldi P, Barletta G, Cortelli P. (2017) Autonomic Nervous System: Clinical Testing. In:Reference Module in Neuroscience and Biobehavioral Psychology. Elsevier; 911-928.
   Google Scholar

1. 35.Kmiec M M, Hou H, Lakshmi Kuppusamy M. (2019) Transcutaneous oxygen measurement in humans using a paramagnetic skin adhesive film.MagnResonMed. 81(2), 781-794.
   Google Scholar

1. 36.Blake D F, Young D A, Brown L H. (2018) Transcutaneous oximetry: variability in normal values for the upper and lower limb.Diving and Hyperbaric Medicine Journal. 48(1), 2-9.
   Google Scholar