Overall findings
We here performed an exploratory analysis of the possible associations between 818 metabolites and orthostatic blood pressure reactions in the general population. We found that eleven of the 818 metabolites were associated with systolic orthostatic blood pressure responses after adjusting for multiple testing and conventional cardiovascular risk factors. No metabolite was significantly associated with manifest OH after Bonferroni adjustment in this population.
Looking specifically at the relationships, 22 metabolites, half of them lipids, were associated with systolic orthostatic blood pressure responses when adjusted for age, gender and systolic blood pressure. Of these, eleven metabolites, likewise predominantly metabolites in lipid pathways, remained significantly associated when adjusting also for diabetes, BMI, smoking and anti-hypertensive treatment.
Lipids and xenobiotics that were associated with orthostatic blood pressure responses mainly showed a negative beta coefficient in relation to orthostatic blood pressure responses (i.e. higher levels were associated with less pronounced systolic blood pressure decrease) whereas vitamins, cofactors and most amino acids showed positive beta coefficients in relation to orthostatic blood pressure responses (i.e. higher levels of these were associated with more pronounced systolic blood pressure decrease on standing; Tables 2, 3 and Fig. 2).
The eleven metabolites that displayed significant associations with orthostatic blood pressure responses after adjusting for CVD risk factors were diverse. However, these included six lipids involved mainly in the dihydro-sphingomyelins, sphingosines, phosphatidylcholine (PC), and fatty acid metabolism (acyl carnitine, polyunsaturated) pathways. Furthermore, of 3 amino acids that play a role in the creatine, the glutamate and the methionine, cysteine, SAM and taurine metabolism pathways, and a food/plant based component (gucuronide of piperine metabolite C17H21NO3) as well as a glutamine degradant, respectively (Table 3).
Lipid metabolites in relation to orthostatic blood pressure reactions
The six lipid metabolites that were associated with systolic orthostatic blood pressure responses are mainly sphingolipids, or their metabolites, that are found in cell membranes.
Behenoyl dihydro-sphingomyelin (DHSM) is a sphingolipid that is found in 5–10% of cells and plays a role in membrane-related biological processes19. It has been shown to be associated with both high BMI and nonalcoholic fatty liver disease in previous study19.
Sphingomyelin is found mainly in membranous myelin sheaths and is suggested to be an insulator of nerve fibers. It has been discovered that it plays a role in cell signaling pathways, including apoptosis20. Excessive sphingomyelin was previously shown to be associated with insulin resistance20.
Sphingosine and sphingadiene are the backbone of all sphingolipids. They and their derivatives are important second messengers involved in functions such as cell growth, differentiation, and apoptosis.
The negative association between sphingolipids and systolic orthostatic blood pressure decrease may possibly be explained by their previously shown association with high BMI, nonalcoholic fatty liver disease and insulin resistance20,21. In addition, recent studies have shown that the sphingolipid plasma S1P (Sphingosine-1-Phosphate) links to hypertension and biomarkers of inflammation and cardiovascular disease22. S1P has been shown to be released by circulating cells such as leukocytes, erythrocytes and blood platelets facilitated among others by the influence of inflammatory mediators such as TNF-alpha. In our study S1P wasn´t shown to associate with orthostatic blood pressure responses when adjusting for age, gender and supine systolic blood pressure. The role of sphingolipids in cell processes and sphingomyelin’s role as a nerve insulator could on the other hand be regarded as protective features against orthostatic hypotension.
Dihomo-linoleoylcarnitine was another metabolite associated with systolic orthostatic blood pressure response. Previous findings on dihomo-linoleic acid, obesity and metabolic health in humans are mixed23. In rodent studies, beneficial effects of dietary linoleic acid administration were found on insulin resistance and glucose tolerance24. Long-chain acyl fatty acid derivatives accumulate in the cytosol and serum of patients suffering from mitochondrial carnitine palmitoyltransferase II deficiency, the most common inherited disorder of lipid metabolism in adults. Carnitine plays a critical role in energy production. It transports long-chain fatty acids into the mitochondria so they can be oxidized to produce energy. It also transports the toxic compounds generated out of this cellular organelle to prevent their accumulation. Given these key functions, carnitine is concentrated in tissues like skeletal and cardiac muscle that utilize fatty acids as a dietary fuel25.
We found one previous study about metabolic changes preceding cardiac dysfunction that showed an association between linoleoylcarnitine and hypertension and hypertension-induced left ventricular hypertrophy in rodents26.
In our study dihomo-linoleoyl carnitine we found a positive association with systolic orthostatic blood pressure decrease.
1-myristoyl-2-arachidonoyl-GPC is a metabolite of the phosphatidylcholine (PC) metabolism. Not much is known about this molecule, whereas three other metabolites of this pathway choline, trimethylamine N-oxide (TMAO), and betaine respectively, previously showed to predict risk for CVD11,12,13,14,15,25. Phospholipid synthesis has been shown to be vital for cell survival, normal development and the maintenance of health. Alterations in phospholipid levels have previously been shown to be associated with lipid profiles, obesity and insulin resistance20,27,28and with hypertension as well as hypertension-induced left ventricular hypertrophy in rodents26.
We found 1-myristol-2-arachidonoyl-GPC to be inversely associated with systolic orthostatic blood pressure decrease.
Amino acid metabolites in relation to orthostatic blood pressure reactions
In our study, three amino acids and one derivative showed association with systolic orthostatic blood pressure responses.
Creatine is an amino acid that is most abundant in muscles, the heart and brain. Its primary role is to bind with inorganic phosphate in the cell to form phosphocreatine, and thereby serve as a high-energy phosphate source of energy to resynthesize adenosine triphosphate (ATP) that has been degraded to adenosine diphosphate (ADP) + Pi as a source of energy to fuel cellular metabolism. It increases cellular energy availability and is among other consumed by bodybuilders for muscle gain. It has also shown to be neuroprotective and is thought to have anti-inflammatory and immunomodulating effects29.
In our study creatine had a negative association with systolic orthostatic blood pressure responses.
Cysteine-S-sulfate is an abnormal metabolite discovered in the urine and blood of a patient with cysteine oxidase deficiency, a rare disorder of sulfur amino acid metabolism associated with brain damage and mental retardation30. We found it to be inversely associated with systolic orthostatic blood pressure decrease.
Glutamine is the most abundant and versatile amino acid in the body. In health and disease, the rate of glutamine consumption by immune cells is similar or greater than glucose. For instance, in vitro and in vivo studies have determined that glutamine is an essential nutrient for lymphocyte proliferation and cytokine production, macrophage phagocytic plus secretory activities, and neutrophil bacterial killing. Glutamine release to the circulation and availability is mainly controlled by key metabolic organs, such as the gut, liver, and skeletal muscles31. During catabolic/hypercatabolic situations glutamine can become essential for metabolic function. Symptoms of glutamine deficiency include increased susceptibility to infections, bowel changes, diarrhea, ulcers, and weight loss. A side effect of too high levels are vegetative symptoms, skin rash, muscle and joint pain and swelling of hands and feet.
We found glutamine and the glutamine degradant to have a positive association with systolic orthostatic blood pressure decrease.
Other metabolites in relation to orthostatic blood pressure reactions
Metabolites of the glucuronidation of piperine pathway are a black pepper constituent which recently have shown to have antioxidant, anti-inflammatory, antidiabetic, anti-mutagenic, tumor-inhibiting, drug metabolism inhibiting and antidiarrheal properties32,33 in humans. Blood pressure lowering effects were previously reported in animal experiments34.
We found glucuronide of piperine metabolite C17H21NO3 to have a negative association with systolic orthostatic blood pressure responses.
Limitations
This is an exploratory study, meaning that we needed to apply correction for multiple testing. We chose the Bonferroni approach, which may be too conservative, however lowering the risk of type 1 error18. The exploratory nature of our study also means, that the results should be confirmed in independent samples, and that the epidemiological and clinical impact of the results is yet to be determined. Still, we find the results valuable as a base for further studies.
Our findings are from the general population in the age group 50–64 years and may not be generalized to other age groups. Regarding the metabolome and cardio-metabolic health it would be interesting in future studies to compare the metabolome in young and healthy age groups with middle aged and old age groups. Cardio-metabolic disease develops and progresses throughout a lifetime and may become manifest in the age-group of the current study, at least as subclinical disease7,8. An increased prevalence of orthostatic hypotension is seen after the 6th decade and it is rising exponentially in the last decades of life1.
We analyzed possible associations between orthostatic blood pressure responses as a marker of suggested autonomic dysfunction and 818 metabolites to cover as many pathways as possible. We only examined every metabolite for itself, so we did not test for any patterns of differences in clusters of metabolites.
It may be discussed, whether or not the subtle changes in orthostatic blood pressure adaptations, seen in this generally healthy population, reflects overt autonomic dysfunction. However, postural change in blood pressure is one of the five Ewing’s tests, which have been used for assessing autonomic function for several decades35. Impaired orthostatic blood pressure reactions36 as well as deep breathing tests37 have previously been shown to be associated with coronary artery calcium deposits in SCAPIS.
The cross-sectional design limits the interpretation of the results: We cannot comment on the causes and effects of single metabolites on orthostatic blood pressure responses, but solely that an association exists. It is however known, that autonomic dysfunction in most cases is a consequence of cardio-metabolic disease.
A methodological limitation is, that the orthostatic blood pressure responses were only measured once at 3 min after standing up, and there are no additional measurements to record possible delayed abnormal orthostatic blood pressure responses. The low prevalence of manifest orthostatic hypotension is in concordance with the prevalence found in earlier studies1. Nevertheless, the low prevalence is a limitation and it is hard to predict the reliability of just a single visit evaluation. Orthostatic hypotension is known to possibly be affected by acute and transient conditions like dehydration, hypoglycemia, fatigue and low blood count1,3. The reliability of single orthostatic blood pressure recordings should be assessed in future studies. It should also be mentioned, that more precise measurement of orthostatic blood pressure reactions, including the detection of transient orthostatic hypotension, can be achieved by a tilt-test with prolonged continuous blood pressure and heart rate monitoring in up to 20 min after tilting2.
Another limitation is, that we in this exploratory study didn´t adjust for kidney and liver function and other conditions that could possibly influence the absorption, excretion, accumulation and half-life of the metabolites, and thereby the levels of measured concentrations.
On the other hand, we managed to include a vast amount of metabolites in the analyses. The real effects of the metabolites found to associate with orthostatic blood pressure responses should now be addressed more thoroughly in future studies.
Finally, since we assessed linear models in this study, we cannot rule out that some metabolites could be non-linearly related to orthostatic blood pressure reactions.