Optimizing linguistic materials for feature-based intelligibility assessment in speech impairments.

Assessing the intelligibility of speech-disordered individuals generally involves asking them to read aloud texts such as word lists, a procedure that can be time-consuming if the materials are lengthy. This paper seeks to optimize such elicitation materials by identifying an optimal trade-off between the quantity of material needed for assessment purposes and its capacity to elicit a robust intelligibility metrics. More specifically, it investigates the effect of reducing the number of pseudowords used in a phonetic-acoustic decoding task in a speech-impaired population in terms of the subsequent impact on the intelligibility classifier as quantified by accuracy indexes (AUC of ROC, Balanced Accuracy index and F-scores). A comparison of obtained accuracy indexes shows that when reduction of the amount of elicitation material is based on a phonetic criterion-here, related to phonotactic complexity-the classifier has a higher classifying ability than when the material is arbitrarily reduced. Crucially, downsizing the material to about 30% of the original dataset does not diminish the classifier's performance nor affect its stability. This result is of significant interest to clinicians as well as patients since it validates a tool that is both reliable and efficient.

French adaptation of the "Frenchay Dysarthria Assessment 2" speech intelligibility test.

OBJECTIVES: Speech intelligibility can be defined as "the degree to which a speaker's intended message is recovered by a listener". Loss of intelligibility is one of the most frequent complaints in patients suffering from speech disorder, impairing communication. Measurement of intelligibility is therefore an important parameter in follow-up. We developed a French version of the "Frenchay Dysarthria Assessment, 2nd edition" (FDA-2), an intelligibility test recognized internationally in its English version. The present study details the construction of the test and its preliminary validation. MATERIALS AND METHODS: We first compiled a set of words and phrases in French, based on the criteria defined in FDA-2. In a second step, we validated the test in healthy subjects in normal and noisy conditions, to check sensitivity to speech signal degradation. RESULTS: The test proved valid and sensitive, as scores were significantly lower for noise-degraded stimuli. CONCLUSION: This French-language intelligibility test can be used to evaluate speech disorder: for example, in dysarthria, head and neck cancer or after cochlear implantation.

Particle exposures and infections.

Particle exposures increase the risk for human infections. Particles can deposit in the nose, pharynx, larynx, trachea, bronchi, and distal lung and, accordingly, the respiratory tract is the system most frequently infected after such exposure; however, meningitis also occurs. Cigarette smoking, burning of biomass, dust storms, mining, agricultural work, environmental tobacco smoke (ETS), wood stoves, traffic-related emissions, gas stoves, and ambient air pollution are all particle-related exposures associated with an increased risk for respiratory infections. In addition, cigarette smoking, burning of biomass, dust storms, mining, and ETS can result in an elevated risk for tuberculosis, atypical mycobacterial infections, and meningitis. One of the mechanisms for particle-related infections includes an accumulation of iron by surface functional groups of particulate matter (PM). Since elevations in metal availability are common to every particle exposure, all PM potentially contributes to these infections. Therefore, exposures to wood stove emissions, diesel exhaust, and air pollution particles are predicted to increase the incidence and prevalence of tuberculosis, atypical mycobacterial infections, and meningitis, albeit these elevations are likely to be small and detectable only in large population studies. Since iron accumulation correlates with the presence of surface functional groups and dependent metal coordination by the PM, the risk for infection continues as long as the particle is retained. Subsequently, it is expected that the cessation of exposure will diminish, but not totally reverse, the elevated risk for infection.

[F0 characteristics in Parkinsonian speech: Contrast between the effect of hypodopaminergy due to Parkinson's disease and that of the therapeutic delivery of L-Dopa]. / Les anomalies de la fréquence fondamentale chez le locuteur Parkinsonien: contraste entre les effets respectifs de l'hypodopaminergie dûe à la maladie de Parkinson et de l'apport thérapeutique par L-Dopa.

Voice and speech impairments are frequent in Parkinson's disease, particularly when the disease is at an advanced stage. These impairments affect spoken communication and may become a serious disability for someone with Parkinson's disease. Many studies based on auditory-perceptual or acoustic methods have been carried out to characterize dysarthria. The heterogeneity of evaluation methods and experimental bias however make results difficult to understand. For instance, in terms of phonatory impairments and with regard to F0, results are contradictory: PD speech may be characterized by either higher F0 or lower F0 compared to control subjects, or there may be no difference at all between the two population. In this study, we aim to provide a conceptual and methodological framework which allows for interpreting the results obtained from 44 speakers (29 PD and 15 control subjects) in relation to physiological (gender, age, PD subjects' pharmacologic state) and linguistic (speech production tasks) constraints. For the present corpus, we did not observe any F0 mean difference between the two groups. Our results however reveal a significant increase in F0 mean in PD subjects under L-dopa. We assume a double and opposite effect on F0 mean during drug withdrawal: low sub-glottal pressure, due to PD, results in a decrease in F0, while laryngeal rigidity leads to an increase in F0. These two effects thus mutually annihilate. Under L-Dopa, however, the drug effect increases sub-glottal pressure, which combined with an increase in F0 due to rigidity, leads to a global increase in F0.

[Prosody and reading: Temporal and melodic characteristics in the dyslexic child in reading and narration]. / Prosodie et lecture: particularités temporelles et mélodiques de l'enfant dyslexique en lecture et en narration.

Dyslexia is widely associated to a massive phonological awareness deficit. This deficit leads to difficulties in grapheme to phoneme conversions in reading and difficulties in words and sentences productions. The origin of the phonological deficit is partially explained by the magnocellular, cerebellar and articulatory theories. Recently, an increasing number of studies demonstrated the relationship between prosody and reading and, more specifically, the potential key role of suprasegmental phonology in the healthy development of phonological representations. The aim of this study is to explore part of prosodic features in dyslexics and normal developing children in reading and narration tasks, in French. We examined reading accuracy, reading rate, pauses frequency and duration, inter-pausal units (IPUs) duration and instantaneous variations of F0. Results show correct decoding skills for all subjects but a lack of automation of this procedure for dyslexics. Differences in pauses frequency and duration, IPUS duration and F0 variations observed between dyslexics and controls confirm the link between prosodic reading and automaticity. The longer pause duration in narrative form a temporal feature of dyslexic's production. This temporal characteristic reflects the cognitive cost done by a speech generation task involving lexical selection, syntactic planning and articulatory programming processes. This result is a first step towards evidence of a suprasegmental phonological deficit in spoken language that could be an early marker of later reading difficulties.

[From intelligibility to comprehension, which measurement in practice?]. / De l'intelligibilité à compréhensibilité de la parole, quelles mesures en pratique clinique?

UNLABELLED: The modelling of the tests assessing speech disorders in touch with the international classification of the functioning implies a reflection on measurement tools. The access to the sense of the message from decoding of the signal to the action infers at the interlocutor's justifies various methods. The objective of this work is to demonstrate the limit of the usual tests of intelligibility. MATERIAL AND METHOD: A population of 10 subjects presenting a speech disorder secondary to a Head and Neck cancer and 2 subjects controls recorded a corpus of 80 syllables developed with the most frequent 16 French consonants and 5 vowels. The test of intelligibility consisted of a task of identification of consonants by a jury of 11 listeners. RESULTS: About is the method used to estimate the informative value of the consonants tested, the uvular consonant /r/ and the nasal In/, /m/ are very well perceived. The weighting by the rate of confusion makes tipped over the position of labiodentales If/, /v/ and that of / p/. The just perception of the sound/p/ would be underestimated while sounds /v/ and /f/ would be overestimated. Sounds /g/, /k/, /j/ remain the least well received. This observation results probably from the profile of distortions produced by the selection of the population and probably from the frequency of the syllables containing these consonants in the french language for the sample of the corpus. CONCLUSION: The rate of confusion appears as a fundamental element, inciting to look for more relevant methods of analysis of the results of the tests of intelligibility.

[Social consequence of a dysphonic voice, design and validation of a questionnaire and first results]. / Conséquences sociales de la dysphonie, conception et validation d'un questionnaire et premiers résultats.

In our society, based on communication, dysphonia becomes a handicap that could be responsible of work discrimination. Actually, several commercial services are provided by phone only, and voice quality is mandatory for the employees. This work aim was to determine the social picture relayed by dysphonia. Our hypothesis was that dysphonia sounds pejorative compared to normal voice. 40 voice samples (30 dysphonic and 10 normal) were presented randomly to a perceptual jury of 20 naïve listener. The task was for each of them to fill a questionnaire, designed specifically to describe the speaker's look and personality. 20 items were evaluated, divided into 4 categories: health, temperament, appearance, and way of life. The results showed significant differences between normal subjects and dysphonic patients. For instance, the pathological voices were depicted as more tired, introverted, sloppy than normal voices, and less trustable. No significant differences were found according to the severity of voice disorders. This work is presently continued. It allowed to validate our questionnaire and has offers great perspectives on patient's management and voice therapy.

[Prosodic features of computer-assisted presentations: a preliminary study]. / Caractéristiques prosodiques des présentations assistées par ordinateur: une etude préliminaire.

INTRODUCTION: Computer-assisted presentations (CAP) seem to be characterized by changes of rhythm and melody of speech, compared to spontaneous speech, in many cases with periods with rapid and excessively regular breaks interspersed with long pauses. METHODS: We analyzed the prosody and especially the pace in the temporal organization of prominences (distribution of stress, articulatory speed, pause duration) in 37 inexperienced subjects in a CAP corresponding to an actual review of graduation. The results were compared with a presentation without support (PWS) conducted immediately after the first. A jury had to propose a gradation of prosody (typical CAP, intermediate, non-typical). Accentual distribution was analyzed from a perceptual experiment. The articulatory speed and length of breaks were analyzed from the speech signal. RESULTS: We found the main elements of prosodic style assumed in three-quarters of subjects. This style was characterized by a faster articulatory rate and a more regular and rhythmic breaks due to excessive elongation of silent pauses when changing view of the slide. DISCUSSION: The assumption seems validated because the majority of subjects in our study are of particular prosodic rhythm (faster and more regular). Subjects with archetypal prosody of CAP appear to be those whose speech followed closely their slide presentation, which disrupted their communication.

Influence of phonetic context on the dysphonic event: contribution of new methodologies for the analysis of pathological voice.

Widely studied in terms of perception, acoustics or aerodynamics, dysphonia stays nevertheless a speech phenomenon, closely related to the phonetic composition of the message conveyed by the voice. In this paper, we present a series of three works with the aim to understand the implications of the phonetic manifestation of dysphonia. Our first study proposes a new approach to the perceptual analysis of dysphonia (the phonetic labeling), which principle is to listen and evaluate each phoneme in a sentence separately. This study confirms the hypothesis of Laver that the dysphonia is not a constant noise added to the speech signal, but a discontinuous phenomenon, occurring on certain phonemes, based on the phonetic context. However, the burden of executing the task has led us to look to the techniques of automatic speaker recognition (ASR) to automate the procedure. With the collaboration of the LIA, we have developed a system for automatic classification of dysphonia from the techniques of ASR. This is the subject of our second study. The first results obtained with this system suggest that the unvoiced consonants show predominant performance in the task of automatic classification of dysphonia. This result is surprising since it is often assumed that dysphonia occurs only on laryngeal vibration. We started looking for explanations of this phenomenon and we present our assumptions and experiences in the third work we present.

Vascular and thrombogenic effects of pulmonary exposure to Libby amphibole.

Exposure to Libby amphibole (LA) asbestos is associated with increased incidences of human autoimmune disease and mortality related to cardiovascular diseases. However, the systemic and vascular impacts are less well examined because of the dominance of pulmonary disease. It was postulated that regardless of the type of exposure scenario, LA exposure might produce systemic and vascular inflammogenic and thrombotic alterations in healthy and cardiovascular compromised rat models. Samples from three independent studies were examined. In the first study, male Wistar Kyoto (WKY), spontaneously hypertensive (SH), and SH heart failure (SHHF) rats were intratracheally instilled once with 0 (vehicle), 0.25, or 1 mg/rat of LA. In the second study, F344 rats were instilled with vehicle or LA at 0.5, 1.5, or 5 mg/rat. In the third study, F344 rats were instilled with the same mass concentrations of LA delivered by biweekly multiple instillations over 3 mo to simulate an episodic subchronic exposure. Complete blood count, platelet aggregation, serum cytokines, and biomarkers of systemic and aortic effects were examined. LA reduced adenosine diphosphate (ADP)-induced platelet aggregation and decreased circulating platelets in WKY (1 mg/rat) and F344 (5 mg/rat) at the 3-mo time point but did not do so in SH or SHHF rats. A decline in circulating lymphocytes with age appeared to be exacerbated by LA exposure in F344 rats but the differences were not significant. Aorta mRNA expression for biomarkers of oxidative stress (HO-1, LOX-1), inflammation (MIP-2), and thrombosis (tPA, PAI-1, vWf) were increased at baseline in SH and SHHF relative to WKY. LA exposure upregulated several of these biomarkers and also those involved in aortic contractility of WKY rats at 3 mo, suggesting thrombogenic, vasocontractile, and oxidative stress-mediated impairments. The aorta changes in F344 rats were less remarkable than changes noted in WKY following LA exposure. In conclusion, exposure to LA decreased circulating platelets and platelet coagulability while increasing the expression of oxidative stress, thrombosis, and vasoconstriction biomarkers in the aorta of healthy rats. These changes were similar to those noted at baseline in SH and SHHF rats, suggesting that LA-induced pulmonary injury might increase the risk of developing cardiovascular disease in healthy individuals.

Transcriptional activation of inflammasome components by Libby amphibole and the role of iron.

The induction of the NALP3 inflammasome complex is shown to be necessary for the development of fibrosis after asbestos exposure. Libby amphibole (LA) induces lung inflammation and fibrosis, while complexation of iron (Fe) on fibers inhibits inflammation. In this study we examined the ability of LA to induce the inflammasome cascade and the role of Fe in modulating inflammasome activity. Spontaneously hypertensive rats were exposed intratracheally to either saline (300 µl), deferoxamine (Def) (1 mg), FeCl(3) (21 µg), LA (0.5 mg), Fe-loaded LA (Fe + LA), or LA + Def. Activities of oxidative stress-sensitive enzymes, expression of inflammasome-specific genes, and cytokine proteins in bronchoalveolar lavage fluid were analyzed. Lung enzymes at 4 h and 24 h post-exposure were unchanged. LA increased lung expression of genes including interleukin-1ß (IL-1ß), cathepsin-B, ASC, NALP3, interleukin (IL)-6 and NFκB. LA+Fe significantly reduced IL-1ß and NFκB with a trend of reduction in ASC, NALP3, cathepsin-B and IL-6 expression. Def treatment did not reverse the inhibitory effect of Fe on IL-1ß and ASC but reversed IL-6 expression. CCL-7, CCL-12, CXCL-3 and COX-2 were induced by LA while LA+Fe tended to reduce these responses. Phosphorylation of ERK but not MEK was increased at 4 h after LA but not LA+Fe exposure. In conclusion, components of the NALP3 inflammasome are transcriptionally activated acutely during LA-induced inflammation. The key inflammatory regulators IL-1ß and NFκB were inhibited in the presence of surface-complexed Fe possibly through decreased ERK signaling upstream of the NALP3 inflammasome. The inflammasome activation by LA may contribute to fibrosis, and Fe may reduce this response and alter compensatory mechanisms in individuals exposed to LA.

[Context influence on the perception of dysphonia: when the knowledge of the patient's clinical state can modify the results of perceptual voice quality assessment]. / Influence du contexte d'écoute dans la perception des dysphonies: effets de la connaissance de l'état thérapeutique du patient sur le jugement de sa qualité vocale.

Two experiments were conducted to examine how knowledge of the patient's clinical state affects the results of perceptual voice quality assessment performed by specialists in voice therapy. This study involved 53 patients with a dysphonia. For each speaker, we selected a pair of recordings made in different circumstances. These pairs of voices were presented to seven listeners (ENT surgeons or speech therapists). The task was to perceptually compare the severity of the dysphonia between the 2 recordings of the pair. Stimuli were presented first in a blind test, then several weeks later with accompanying information about the patient; in particular, whether the voice was pre- or post- treatment was explicitly specified. We balanced this artificial contextual information in order to (alpha) reinforce the blind judgment (for example, voices perceived as better in the blind test were indicated as post treatment); (beta) be inconsistent (in a clinical point of view) compared to the blind test (for example, voices perceived as more disordered during the blind test were indicated as post treatment). Results revealed that in the clinical-consistent context alpha, the preference was amplified in a significant way. In clinical-inconsistent condition beta, we observed an inhibition effect or a change of decision. In this condition, the judgment was more dependant on the contextual information (pre/post treatment) than on the auditory sensation obtained in blind condition. These findings are discussed with reference to results in the literature on visual, olfactory or auditory perception in context. In the frame of perceptual voice assessment, results revealed that only blind tests can provide reliable results.

[Perceptual assessment of dysphonia: a training protocol with natural speech]. / Mise au point et évaluation d'un protocole d'apprentissage de jugement perceptif de la sévérité de dysphonies sur de la parole naturelle.

INTRODUCTION: In this study, we proposed and tested the effectiveness of a training procedure on perceptual evaluation of dysphonia. Participants were naive listeners who learned to judge the severity of vocal disorders. MATERIALS AND METHODS: The corpus consisted of 142 female voices ranging from normal to severely dysphonic voices. A total of 38 naive listeners were tested, and the experiment was divided in 3 phases: (1) A pretest to assess the level of our listeners in the categorization of dysphonia, (2) the training phase and (3) a post-test to assess the improvement associated with the training. To examine the persistence of the training, the post-test was administrated twice: Immediately after the training session and one week later. RESULTS: Improvement in the performance between the pretest and the post-test was found for moderate and severe dysphonia. No improvement was observed in the categorization of slightly dysphonic voices, which seem the more resistant to learning. Normal voices also showed no improvement, which is likely due to the high performance on these voices during the pre-test. The improvement observed in the categorization of moderate and severe dysphonia was still present even one week after the training. DISCUSSION: The protocol tested in this study appears to be effective and could constitute an element of training courses for speech and voice therapists. The perception of the intermediate levels of dysphonia, however, remains difficult to quantify and needs to be improved.

Glucagon-like peptide-1 secretion in women with gestational diabetes mellitus during and after pregnancy.

UNLABELLED: Gestational diabetes mellitus (GDM) predisposes women to future development of Type 2 diabetes mellitus (DM2) and the two conditions share similar metabolic alterations. Recent observations suggest that a defective glucose stimulated insulin secretion by glucagon-like peptide-1 (GLP- 1) plays a role in the pathogenesis of DM2. Whether such a defect is impaired in GDM remains to be ascertained. AIM: We have determined GLP-1 secretion in response to oral glucose tolerance test (OGTT) in GDM and normal glucose tolerance (NGT) during and after pregnancy. MATERIALS AND METHODS: 100-g-3h OGTT was performed in 12 GDM and 16 NGT women at 27.3 ± 4.1 weeks of gestation, for determination of plasma GLP-1, glucose, insulin, and C-peptide. Insulin sensitivity (ISI) and insulin secretion (first and second phase); as well as ISI-secretion index (ISSI) were also derived. RESULTS: NGT and GDM women were comparable for age pre-pregnancy body mass index (BMI) and weight gain. GDM had higher glucose area under the curve (AUC): 27,575.5 ± 3448 vs 20,685.88 ± 2715 mg/dl min (p<0.01), but lower first-phase insulin secretion (993.12±367 vs 1376.61 ± 423, p<0.05) and ISSI compared to controls (3873.23 ± 1185 vs 6232.13 ± 1734, p<0.001). When we examined GLP-1 mean levels in relation to mean glycemic values, GLP-1 secretion was inappropriately low with respect to mean glycemic values in GDM compared to NGT. At follow-up, AUCGLP-1 was significantly lower in post-partum GDM compared to post-partum NGT women (2542 ± 273 vs 10,092 ± 7367 pmol·l-1·min-1, p<0.05, respectively). CONCLUSIONS: Our study suggests that GLP-1 secretion in GDM women is inadequate for the prevailing glycemic levels both in pregnancy and post partum. Moreover, we cannot exclude that other important aspects of the incretin effect may be involved in GDM development.

[Preliminary experimental studies of whispered voice: subglottal pressure and postural aspects]. / Etudes expérimentales préiminaires de la voix chuchotée: pression sous-glottique et étude posturale.

OBJECTIVES: Whispered voice is investigated from the energetic point of view. Our hypothese is that whispered voice, and mainly the forced whispers are very similar to vocal effort, for non acoustic vocal parameters, and body movements. DESIGN: Two experiments are combined. The first experiment focused on the laryngeal aspects, mainly the subglottal pressure, and also the electroglottography. The second experiment analyzed the body movements associated with loud whispered voice. RESULTS: No electroglottograhic wave was observed during the two experiments. The vocal intensity was always lower during whispered voice than during comfortable voice (p < 0.001). In experiment 1, subglottal pressure was significantly higher during forced whisper than during the other conditions (p < 0.001). In experiment 2, the amplitude of the trunk movements in the sagittal plane were greater during loud whisper than during comfortable voice (p < 0.001). CONCLUSION: Whispered voice can be close to vocal loading, especially the loud whispered voice.

[Dysarthria across Parkinson's disease progression. Natural history of its components: dysphonia, dysprosody and dysarthria]. / La dysarthrie au cours de la maladie de Parkinson. Histoire naturelle de ses composantes: dysphonie, dysprosodie et dysarthrie.

Dysarthria refers to a collective name for a group of neurologic motor speech disorders, resulting from central and/or peripheral nervous system abnormalities. Speech alteration in Parkinson's disease, so-called hypokinetic dysarthria, presents with prosodic insufficiency, related to a monotony of pitch and intensity, a reduction of accentuation, variable speech rate and possible phoneme imprecision. In most cases, voice is harsh and breathy. This symptom can affect both voice and speech quality, as well as prosody and intelligibility. As a consequence, many patients complain about speech impairments, which affect their communication in daily living activities. Perceptual and instrumental assessments require different and numerous investigation methods, which use may help to further understand the specific dysarthria pathophysiology. This is of importance in order to adjust treatments for dysarthria; as a matter of fact, dopa-therapy, functional neurosurgery or even behavioural speech therapy have variable effects on voice and speech quality in Parkinson's disease.

Cumulative exposure assessment for trace-level polycyclic aromatic hydrocarbons (PAHs) using human blood and plasma analysis.

Humans experience chronic cumulative trace-level exposure to mixtures of volatile, semi-volatile, and non-volatile polycyclic aromatic hydrocarbons (PAHs) present in the environment as by-products of combustion processes. Certain PAHs are known or suspected human carcinogens and so we have developed methodology for measuring their circulating (blood borne) concentrations as a tool to assess internal dose and health risk. We use liquid/liquid extraction and gas chromatography-mass spectrometry and present analytical parameters including dynamic range (0-250 ng/ml), linearity (>0.99 for all compounds), and instrument sensitivity (range 2-22 pg/ml) for a series of 22 PAHs representing 2-6-rings. The method is shown to be sufficiently sensitive for estimating PAHs baseline levels (typical median range from 1 to 1000 pg/ml) in groups of normal control subjects using 1-ml aliquots of human plasma but we note that some individuals have very low background concentrations for 5- and 6-ring compounds that fall below robust quantitation levels.

The effects of ambient particulate matter on human alveolar macrophage oxidative and inflammatory responses.

Epidemiologic and occupational studies demonstrated that ambient particulate matter (PM) and diesel exhaust particles (DEP) exert deleterious effects on human cardiopulmonary health, including exacerbation of pre-existing lung disease and development of respiratory infections. The effects of ambient PM on lung cell responsiveness are poorly defined. Human alveolar macrophages (AM) were exposed to SRM 1649 (Washington, DC, urban dust; UD), SRM 2975 (forklift diesel exhaust particles; DEP), and fine or coarse ambient PM collected in Chapel Hill, NC, during the late fall (November) and early summer (June) of 2001-2002. AM were subsequently incubated with lipopolysaccharide (LPS), phorbol myristate acetate (PMA), or calcium ionophore A23817 for 6 or 24 h after PM exposure. UD and DEP markedly suppressed O2- release 24 h post-PM exposure. UD exposure significantly inhibited tumor necrosis factor (TNF)-alpha, interleukin (IL)-6, and IL-8 release after exposure to 10 nanog/ml LPS. DEP significantly suppressed only TNF-alpha and IL-6 release. Suppressed cytokine release may also be produced by reduced cellular cytokine production. Data suggested that decreased cytokine release is not produced by the presence of benzo[a]pyrene (BaP), a polycyclic aromatic hydrocarbon. Comparison of TNF-alpha release after LPS, PMA, or A23817 revealed that suppressive effects of UD are LPS dependent, whereas inhibitory effects of DEP may work across multiple mechanistic pathways. November and June Chapel Hill PM exposure stimulated TNF-alpha and IL-8 release before LPS exposure. Fine and coarse November PM exposure markedly suppressed TNF-alpha release 6 h after LPS stimulation, but appeared to exert a stimulatory effect on IL-8 release 24 h after LPS exposure. June fine and coarse PM suppressed IL-8 release after LPS exposure. Data suggest that seasonal influences on PM composition affect AM inflammatory response before and after bacterial exposure. Overall, delayed or inhibited AM immune responses to LPS after PM exposure suggest human exposure to ambient PM may enhance pulmonary susceptibility to respiratory infections.