It has long been known that volatile chemicals present in the blood can be transferred to the exhaled breath. Breath analysis for trace chemicals offers a non invasive investigation that can be applied in a huge range of conditions. For example, hydrogen and/or methane breath analysis can be carried out following oral administration of poorly absorbable carbohydrate (for example, lactulose). Hydrogen and methane are products of gut microbiota fermentation, hence this breath test is thought to be a measure of gut microbiological activity. This testing is carried out by some in conditions like small intestinal bacterial overgrowth (SIBO ), although somewhat controversial due to imperfect sensitivity. These odourless gases have been absorbed from the gut and transferred to the alveolar breath.The same VSC that are reported to be the most important volatiles in halitosis (H2S, MM and DMS) are implicated as being the greatest contributors to the malodorous character of flatus and faeces. Only DMS is stable in blood and capable of creating blood borne halitosis.

The general mechanism involved in blood borne halitosis is as follows. Volatiles enter the systemic circulation (several routes possible but most usually via the portal system following absorption from the gut), and are then circulated to the lungs, where there is intimate associated between the pulmonary alveoli and capillary networks. During gas exchange of waste carbon dioxide and inhaled oxygen, chemicals present in the blood, may also be exchanged, and may be perceptible on the exhaled breath if they fulfil certain criteria. The detection of a volatile on the breath by objective measurement does not imply its contribution to perceptible malodour. Although all odorants are volatile, not all volatiles are odorants. An odorant must also be present in high enough concentrations to stimulate olfactory detection. For this to happen, great enough quantities of the volatile must be liberated from solution (that is, respiratory tract secretions, saliva). Whether the chemical is liberated into the gas phase, and in what quantities, is dependent upon such factors as pH and temperature of the solution and the concentration of the volatile in the solution.

However, since the blood goes almost everywhere, blood borne volatiles may also be expressed via other routes of excretion, for example, sweat, saliva and urine. An example of a blood borne volatile being expressed via multiple routes, and not just the breath, is trimethylamine (TMA),which has a fish-like odour in vitro, becoming ammoniacal at higher concentrations. Indeed when TMA is present in the urine, the rare disorder trimethylaminuria (TMAU) should be considered, although there are other causes of elevated urine TMA. Dimethylglycinuria is another rare cause of fish odour.

Tangerman and Winkel reported that dimethyl sulphide (DMS, CH3SCH3) was the most common volatile in extra-oral halitosis, which could not be attributed to recent ingestion of volatile foodstuffs. They estimated this DMS-related blood borne halitosis may affect some 0.25-1.5% of the general population. These estimates were based upon a cohort of 58 patients complaining of halitosis, and the findings extrapolated assuming 10-30% prevalence of halitosis in the general population and 5-10% of genuine halitosis being extra-oral in aetiology.

DMS is a volatile organo sulphur compound, constituting a sulphur atom covalently bonded to two methyl groups. The in vitro odour character has been variably described as wild radish or cabbage-like,or unpleasantly sweet,but invariably the terms used have unpleasant connotations. DMS was shown to reduce mood compared to more pleasant scents.100 DMS has an exceptionally low odour detection threshold, meaning that it is detectably malodorous even at very low concentrations (24-100 ppb).The gas can be fatally noxious101 and it is also a mild skin and eye irritant,102 although even at the pathological concentrations reported in disease this is probably not a feature.

In mammals, free DMS is the only form known to be present.DMS can be catabolised from dietary precursors in plants (S-methylmethionine/vitamin U, and dimethylsufoniopropionate).91,103,104 DMS can be synthesised from transamination of methionine, or methylation of methyl mercaptan (which itself can be methylated from H2S). Methylation of H2S and MM is thought to be a detoxification pathway. Dimethylsulphoxide (DMSO) can be reduced to DMS. Bacterial spp. known to synthesise DMS have been demonstrated in the human GI tract. DMS is a component of flatus and faeces and it is assumed that this is the result of intestinal bacterial activity. There is evidence that suggests of all the VSC produced in the colon >90% are absorbed by the lining rather than being emitted as flatus, and H2S and MM are thought to be metabolised by caecal lining tissue (to thiosulphate). DMS, however, is not broken down here, instead passing straight to blood.Free DMS is neutral molecule, it does not contain a reactive thiol group (unlike MM and H2S), hence it is stable in blood as it is unreactive with proteins. MM and H2S both react rapidly in blood and are therefore not implicated in blood borne halitosis.DMS that is absorbed into the bloodstream appears to be a substrate of both CYP450 and flavin containing mono-oxygenases (FMO).

DMS is known to be subject to renal and pulmonary excretion. Dietary loading with either methionine or DMS also produced DMS in the milk of cows. After the work of Tangerman's group, where Tenax trapping and gas chromatography was used to objectively measure free DMS concentrations, the normal range can be taken as <7 nM in peripheral venous blood. Breath DMS is also very low in health. Normal subjects in one study were reported to have a breath DMS range of 0.13-0.65 nM.122 Urine DMS is reported as being generally below the 3 nM detection minimum for gas chromatography.91

The new metabolic condition that has potentially been identified by Tangerman and Winkle, awaits further investigation to identify the exact pathogenesis. Whether the dimethylsulphidemia is related to defects in the methionine cycle, or indeed another cause remains to be elucidated. The prevalence of the potential new condition has been estimated at 0.25-1.25% of the general population,32 which would make it more prevalent than other 'blood borne malodour' conditions previously identified. The extra-oral nature of the halitosis in these patients was deduced by demonstrating equal levels of DMS on the nose and mouth breath. Furthermore, blood DMS was shown to correlate perfectly with breath DMS. Experimentation with solutions of DMS was found to have an odour comparable to the character of halitosis in this group. Patients with this proposed metabolic condition are reported to have breath DMS levels of 0.5–2.5 nM91 and blood DMS levels of 10–80 n mol/l.32

Interestingly, a patient in this group showed significantly raised breath DMS 12 hours after consuming 12 glasses of beer. After this experiment, the patient in question was said to switch to drinking wine, and reported fewer complaints of bad breath.66

These findings could be interpreted with caution as they are based on a small sample, and are as yet uncorroborated by other researchers.