Search

Circulating autoantibodies (auto-Abs) neutralizing high concentrations (10 ng/ml; in plasma diluted 1:10) of IFN-α and/or IFN-ω are found in about 10% of patients with critical COVID-19 (coronavirus disease 2019) pneumonia but not in individuals with asymptomatic infections. We detect auto-Abs neutralizing 100-fold lower, more physiological, concentrations of IFN-α and/or IFN-ω (100 pg/ml; in 1:10 dilutions of plasma) in 13.6% of 3595 patients with critical COVID-19, including 21% of 374 patients >80 years, and 6.5% of 522 patients with severe COVID-19. These antibodies are also detected in 18% of the 1124 deceased patients (aged 20 days to 99 years; mean: 70 years). Moreover, another 1.3% of patients with critical COVID-19 and 0.9% of the deceased patients have auto-Abs neutralizing high concentrations of IFN-β. We also show, in a sample of 34,159 uninfected individuals from the general population, that auto-Abs neutralizing high concentrations of IFN-α and/or IFN-ω are present in 0.18% of individuals between 18 and 69 years, 1.1% between 70 and 79 years, and 3.4% >80 years. Moreover, the proportion of individuals carrying auto-Abs neutralizing lower concentrations is greater in a subsample of 10,778 uninfected individuals: 1% of individuals 80 years. By contrast, auto-Abs neutralizing IFN-β do not become more frequent with age. Auto-Abs neutralizing type I IFNs predate SARS-CoV-2 infection and sharply increase in prevalence after the age of 70 years. They account for about 20% of both critical COVID-19 cases in the over 80s and total fatal COVID-19 cases. © 2021 The Authors, some rights reserved.

Computerised prescriptions for Hospital Discharge Orders (HDO) are used world-wide to secure medication processes.To evaluate physicians' adoption of computerised provider order-entry (CPOE) for HDO and the prescribing error rate of HDO in an acute medical care unit.A prospective study was conducted in an internal medicine department over a six-month period. The use rate of CPOE for HDO edition, prescription lines concordance between CPOE-edited HDO, exit prescriptions transcribed in the discharge summary (DS), and prescribing error rate in CPOE-edited HDO were all evaluated.A total of 407 patients with HDO were included in the study. HDO were edited via CPOE system for 350 patients (86%), among which 124 (35%) were identically transcribed, 217 (62%) had discrepancies, and nine (3%) were not transcribed in the discharge summary (DS). Prescription errors were analysed using the total of 2,854 drugs prescribed on HDO. Although hospital pharmacists had signalled discrepancies and provided recommendations to the prescribers via alerting pharmaceutical interventions in CPOE 67 prescription errors (error rate of 2.3%) were found. Errors included 53 cases of refractory period disrespected, four cases of drug interactions, three cases of drug redundancies, and two cases of excessive dosage.This study highlights that most HDO were edited via the CPOE system. Together with pharmacist's interventions, the CPOE system contributed to reducing the prescription error rate in HDO. However, discrepancies in the recording process to DS were frequent, calling for reinforcement of error prevention strategies upon the integration of a CPOE system in the hospital's Electronic Health Records. Providing regular training for physicians is also a requirement.

RATIONALE & OBJECTIVE: The KDIGO (Kidney Disease: Improving Global Outcomes) guideline on CKD presented an international classification system that ranks patients' risk for CKD progression. Few data on children informed guideline development STUDY DESIGN: Observational cohort study SETTINGS AND PARTICIPANTS: Children aged 1-18 years enrolled in the North American Chronic Kidney Disease in Children (CKiD) cohort study and the European Effect of Strict Blood Pressure Control and ACE Inhibition on the Progression of CRF in Pediatric Patients (ESCAPE) trial. PREDICTOR: Level of estimated glomerular filtration rate (eGFR) and proteinuria (urine protein-creatinine ratio (mg/mg)) at study entry OUTCOME: A composite event of renal replacement therapy, 50% reduction of estimated GFR (eGFR), or eGFR 2.0 at study entry. Six ordered stages with varying combinations of eGFR categories (60-89, 45-59, 30-44 and 15-29 ml/min/1.73m(2)) and UPCR categories (2.0) described the risk continuum. Median times to event ranged from >10 years for eGFR 45-90 and UPCR 2. Children with glomerular disease were estimated to have a 43% shorter time to event than children with nonglomerular disease. Cross-validation demonstrated risk patterns that were consistent across the ten subsample validation models. LIMITATIONS: Observational study, utilized cross validation rather than external validation CONCLUSION: CKD staged by level of eGFR and proteinuria characterizes the timeline of progression and can guide management strategies in children.

IF 7.191 (2017); International audience; BACKGROUND:Previous studies showed that high and low body mass index (BMI) was associated with worse prognosis in early-stage colorectal cancer (CRC), and low BMI was associated with worse prognosis in metastatic CRC (mCRC). We aimed to assess efficacy outcomes according to BMI.PATIENTS AND METHODS:A pooled analysis of individual data from 2085 patients enrolled in eight FFCD first-line mCRC trials from 1991 to 2013 was performed. Comparisons were made according to the BMI cut-off: Obese (BMI ≥30), overweight patients (BMI ≥ 25), normal BMI patients (BMI: 18.5-24) and thin patients (BMI

Consider the discrete wavelet transform (DWT) of a time series X={Xt,t ∈ Z} with weakly stationary K th differences. Such time series are encountered in many situations, including unit root or long memory processes. If the wavelet has at least K vanishing moments, the DWT is jointly stationary across scales provided that small scales coefficients are reshaped in appropriate blocks to cope with the downsampling embedded in the DWT. Our goal is to compute the covariance matrix or the joint spectral density of the DWT, given the autocovariance of the Kth differences. We assume that the DWT corresponds to a multiresolution analysis (MRA), which allows us to compute wavelet coefficients iteratively through a succession of finite impulse response (FIR) filters and downsampling. This iterative scheme, however, is not suitable for our purpose in the case where the process X itself is not stationary. Hence, we first derive an iterative algorithm with the same DWT output but with input the Kth differences of the time series. An iterative low complexity scheme is then deduced to compute the exact covariance matrix and spectral density of the DWT. This new algorithm is an opportunity to investigate how using exact DWT covariance computations improves previously proposed statistical methods that rely on approximated computations. Numerical experiments are used for comparisons. Two cases are examined: 1) a local semi-parametric likelihood estimation of long memory processes in the wavelet domain and 2) the computation of a test statistic for detecting change points in the wavelet domain for long memory processes. A real data set analysis is also presented. Namely, economic structural changes are investigated by looking for change points in the daily S&P 500 absolute log returns.

Resume L'endotheline 1 (ET-1) possede une puissante action vasoconstrictrice, en particulier au niveau renal. Cette etude decrit la mise au point d'un dosage de l'ET-1 a partir de reactifs commerciaux sur un volume de plasma reduit (0,6 ml) et evalue chez l'enfant l'influence de l'insuffisance renale sur les concentrations plasmatiques de l'ET-1. L'ET-1 est prealablement extraite sur minicolonnes Sep Pack C18 activees par du methanol, equilibrees par de l'eau distillee, eluees par un melange acide trifluoroacetique, acetonitrile, eau (v/v/v). Les rendements d'extraction sont de 90,5 ± 8,9%. Le dosage radio-immunologique utilise un standard ET-1 Novabiochem et un immunserum anti-ET-1 Peninsula. Les coefficients de variation intra- et interessais sont respectivement de 10,7 et 15,8%. Chez les enfants insuffisants renaux non dialyses, les concentrations plasmatiques d'ET-1 (9,8 ± 3,2 pg/ml; moyenne ± DS; n =48) sont plus elevees ( P n = 29) et inversement correlees ( P n = 41; P

Missing value estimation is important in DNA microarray data analysis. A number of algorithms have been developed to solve this problem, but they have several limitations. Most existing algorithms are not able to deal with the situation where a particular time point (column) of the data is missing entirely. In this paper, we present an autoregressive-model-based missing value estimation method (ARLSimpute) that takes into account the dynamic property of microarray temporal data and the local similarity structures in the data. ARLSimpute is especially effective for the situation where a particular time point contains many missing values or where the entire time point is missing. Experiment results suggest that our proposed algorithm is an accurate missing value estimator in comparison with other imputation methods on simulated as well as real microarray time series datasets.

It is well-known (Narula et al., 1998) that the performance of a multiple input single output (MISO) communication system can be improved if channel information is used. In this work we exploit both temporal and spacial correlation of the fading channel to design the best precoder in sense of minimum mean-square error. This performance of the precoder is compared using an EM- like decoding, a PSP decoding and a new proposed decoding algorithm.

In this paper we consider the localization of a mobile station (MS) in time division multiple access (TDMA) based communication systems. We use joint angle and delay measurements of the emitted signals, impinging on an antenna array at different base stations (BSs). Contrary to previously reported work, our technique takes into account not only the measurement noise, but also multipath propagation and possible BSs reporting only non line of sight (NLOS) measurements. Data are processed using the maximum likelihood (ML) approach based on an implementation of the Expectation Maximization (EM) algorithm. We illustrate the proposed approach using simulated data.

The article deals with a technique of voice forgery using the ALISP (automatic language independent speech processing) approach. Such a technique allows the voice of an arbitrary person (the impostor) to be transformed, forging the identity of another person (the client). Our goal is to demonstrate that an automatic speaker recognition system could be seriously threatened by a transformation of this kind. For this purpose, we use a speaker verification system to calculate the likelihood that the forged voice belongs to the genuine client. Experiments on NIST 2004 evaluation data show that the equal error rate for the verification task is significantly increased by our voice transformation.

In this paper, we consider the localization of a source from quantized measurements of time of arrivals (TOA) or time difference of arrivals (TDOA). Applications include, as particular examples, acoustic source localization from a network of microphones under communication constraints, and the localization of a base station using a geolocalized mobile station using timing advance measurements. We use a maximum likelihood approach, based on an efficient implementation of the EM algorithm. Contrary to previously reported work, our technique takes into account not only the measurement noise, but also the presence of outliers (for example, non line of sight propagation) and the quantization. We illustrate our findings using simulated data and real field measurements.

International audience; In this paper, an improvement of the quantization optimization algorithm for the MPEG Advanced Audio Coder (AAC) is presented. This algorithm, given a bit-rate constraint, minimizes the perceived distortion generated by the signal compression. The distortion can be related to the quantization error level over frequency sub-bands through an auditory model. Thus, optimizing the quantification requires knowledge of the rate-distortion function for each sub-band. When this function can be modeled in a simple way, the algorithm can take a one-loop recursive structure. However, in the MPEG AAC, the rate-distortion function is hard to characterize, since AAC makes use of non-linear quantizers and variable length entropy coders. As a result, the standard algorithm makes use of two nested loops with a local decoder, in order to measure the error level rather than predicting its value. We first describe a partial sub-band modeling of the rate-distortion function of interest in the MPEG AAC. Then, using a statistical approach, we find a relationship between the error level and the so-called quantization ``scale-factor'' and propose a new algorithm that is basically similar to a classical one loop ``bit allocation'' process. Finally, we describe the complete algorithm and show that it is more efficient than the standard one.

Orthogonal frequency division multiplex (ODFM) is an attractive way of transmitting signals, by turning a wide-band signal into several narrow band ones. Orthogonal transmultiplexers are the basic tools required for emitting these signals and recovering them at the receiver side. It is well known that the only DFT modulated transmultiplexers are based on a rectangular window as a prototype, if one sticks to critically sampled systems. This paper investigates the possibilities brought by using oversampled FIR transmultiplexers instead of the critically sampled ones.

In a MPEG-4 analytic audio scene, spatialization filters can be used after an audio coder. When properly tuned, a MPEG advanced audio coder (AAC), generates an inaudible distortion if a sufficient output bit-rate is available, and, more important in the context of our study, as long as the listener has direct access to the unprocessed output of the decoder. We show that using a spatialization filter after the decoder, as described in the MPEG-4 standard, may damage the perceived quality. We propose solutions and present a modified encoder working with a hearing model that includes all audio processing (including spatialization), which allows bit-rate reduction without quality damage.

A new class of blind equalization cost functions have been proposed (see. Shtrom, V. and Fan, H., IEEE Trans. on SP., vol.46, no.10, p.2674-83, 1998). They have the particularity of being unimodal, and some of them are sectionally convex in the combined system domain (CDS). Both properties seem very attractive for a cost function, though no proof (or specific definition) was given of the sectional convexity. We show, using a particular case of this class, that sectional convexity not only implies fixing a delay, but that it also requires to fix the value of the coefficient associated with such delay. We show that the CMA (constant modulus algorithm) criterion shares the same property, and that this property is difficult to maintain in actual algorithms.

This paper presents two ways of obtaining more robust and/or efficient blind equalization algorithms. The first one is based on the linear combination of second order statistics methods (SOS) and higher order statistics methods (HOS) in such way that the advantages of both approaches are kept. The new algorithm has better performance than each of its constituents taken separately (pure HOS or SOS method) rather than performance "in between". The second idea consists of taking into account some deterministic knowledge of the input signal in the SOS methods. The results obtained show that this technique also enhances the performance of the algorithm.

The cyclosporine microemulsion formulation Neoral, which allows a better absorption and a more regular pharmacokinetic profile, is proposed for replacing the original formulation, Sandimmun. The present study reports the results of conversion from Sandimmun to Neoral in children with a kidney graft, a population for which information remains limited.Twenty children, 2.5 to 10.5 years of age, who had a kidney graft with a stable renal function for between six months to five years (m = 2.6) were the subjects of this study. The patients were switched from Sandimmun to Neoral at the same dose, adjusted afterwards on a cyclosporine trough level.After six months, the mean dose decreased from 9.1 mg/kg/d to 8.4 mg/kg/d, i.e., 12.5%. After one year, the mean dose was 7 mg mg/kg/d, i.e., 28%. Of the 65% of patients who had a decreased dose, most of them had the highest dose of Sandimmun at the start. Mean serum creatininemia levels slightly increased from 85.6 to 89.5 mumol/L after six months (P = 0.03). None of the patients had a rejection crisis during the first six months under Neoral. Blood pressure did not change significantly, hirsutism improved in two cases but increased or appeared in two cases as well. Gingival hypertrophy increased or appeared in four cases.A decrease in the dose was decided on either to maintain the trough CsA blood level in the desired range or because of the appearance of a symptom suggesting a side effect of cyclosporine, especially the increase of creatinemia. The trough level did not appear to be the best index for adapting the dose.In stable pediatric kidney transplant recipients, the switch from Sandimmun to Neoral provided a reduction in drug dosage in 65% of cases without an increase in adverse events.

From 1976 through 1991, 50 renal transplants (with 7 kidneys from living related donors and 43 cadaver kidneys) were performed in 47 children under five years of age (range 11 to 59 months) at the Enfants-Malades Hospital, Paris, France. Donor age ranged from 3 months to 53 years. Six of the seven kidneys from living related donors are currently functioning after a follow-up of 6 months to 8 years, whereas actuarial survival of cadaver kidneys was 70% at one year and 66% at five years. The main cause of graft loss was vascular thrombosis (40% of lost kidneys). In the most recent years of the study period, graft survival was substantially improved by routine prophylactic anticoagulant therapy with low-molecular weight heparin: one-year graft survival rate was 83% in the 23 children grafted between 1989 and 1991. Posttransplantation growth was closely related to quality of graft function: among the 29 children with sufficiently long follow-up the 19 patients with normal renal function exhibited normal or catch-up growth, whereas the ten patients with chronic renal failure or recent rejection had poor growth. Complications were uncommon with the exception of hypertension. Mortality rate was 12%, i.e., only slightly higher than in older pediatric kidney recipients. Achievement at school was normal in most cases (with a lag in only five patients). Provided effective therapy is given to prevent the main adverse outcome (i.e., vascular thrombosis), renal transplantation does not involve excessive risks even in infants as young as one year of age.

The authors deal with frequency-domain (FD) adaptive identification with application to acoustic echo cancellation. Several proposed frequency-domain LMS algorithms are reviewed and unified in the common framework of fast convolution using the short-term Fourier transform. A more flexible identification scheme is then proposed, making it possible to independently control the delay and the adaptation rate. The theoretical proof of convergence of this new scheme is provided together with simulations demonstrating the efficiency of this method in both stationary and nonstationary environments. >

Among 149 kidneys transplanted in 141 children and teenagers aged 4 1/2 to 20 years, from November 1972 through December 1979, 70 (47%) were still functioning after ten years (67/145 cadaver donor grafts and 3/4 living related donor grafts). Medical and social status at the last follow-up visit (10 to 16 1/2 years after transplantation; m = 12 years) was analyzed. Patients were divided into five groups on the basis of glomerular filtration rate (GFR; ml/min/1.73 m2) and blood pressure (BP): 1) GFR greater than 80 ml and normal BP: 23 patients (33%); 2) GFR in the 60-80 ml range and/or high BP: 24 patients (34%/3) GFR in the 40-60 range: 6 patients; 4) GFR in the 15-40 range: 7 patients; and 5) hemodialysis restarted 10 1/2 to 13 years after transplantation. Mean adult stature was 155.7 +/- 10.4 cm in males and 149.8 +/- 10 cm in females. Osteoporosis was found in 88% of patients who underwent bone density quantitation. Twenty-four per cent of patients had aseptic osteonecrosis with variable degrees of impairment as a result. Chronic HBsAg carrier status was found in 37% of patients and was accompanied with persistent cytolysis in half the cases. Only one malignancy was seen (carcinoma of the urinary bladder in a child under cyclophosphamide). Six deaths were recorded between 10 and 13 years after transplantation; causes included septicemia (2 cases), cancer (1 case), hepatitis B (1 case), cerebral cystinosis (1 case), and unexplained sudden death (1 case).(ABSTRACT TRUNCATED AT 250 WORDS)

A clinicopathologic study of 41 biopsy specimens from renal grafts transplanted at least ten years earlier showed normal findings or only mild tubulointerstitial lesions in 63% of cases. Results confirmed the good prognostic value of renal graft biopsy performed ten years after transplantation. A very close correlation was demonstrated between histologic evidence of chronic rejection (interstitial fibrosis, tubular atrophy, and fibrous endarteritis) and long-term graft outcome. Severity of interstitial fibrosis was apparently a more reliable indicator of prognosis than severity of fibrous endarteritis. Among 16 patients with a history of acute rejection, none had residual parenchymatous lesions. In contrast, among patients with major tubulointerstitial and/or vascular lesions, most had previously experienced one or several episodes of acute rejection. Evidence of allograft glomerulopathy was seen in approximately one case in four (11 patients). Among these patients, only eight had proteinuria and only four lost their graft. Among the four patients who developed recurrent membranoproliferative glomerulonephritis (type I in three cases and type II in one case), three lost the deposits previously detected by immunofluorescence, probably as a result of immunosuppressive therapy.

Objectifs La stenose des arteres renales est systematiquement recherchee en cas d’hypertension arterielle (HTA) de novo chez l’enfant. L’echo-Doppler est l’examen realise de premiere intention, mais sa sensibilite est faible. L’arteriographie est actuellement l’examen de reference. Cet examen est invasif et comporte une morbidite non negligeable. Le but de ce travail est d’evaluer l’interet du scanner multicoupe de derniere generation dans le diagnostic et l’exploration des stenoses de l’artere renale chez l’enfant. Materiels et methodes Sept enfants hospitalises pour bilan d’HTA de novo ont beneficie d’un angioscanner des arteres renales. Les resultats d’imagerie ont pu etre correles a l’angiographie numerisee ou aux donnees chirurgicales. Resultats Dans tous les cas, le scanner a confirme la presence d’une stenose des arteres renales uni ou bilaterale, a pu en evaluer l’etendue, l’importance et la localisation en vue d’un eventuel geste interventionnel radiologique ou chirurgical. Il permet egalement de faire le bilan d’autres anomalies arterielles associees et de realiser si besoin une cartographie vasculaire pre-operatoire. Conclusion Le scanner multicoupe de derniere generation est l’examen de choix dans l’exploration des stenoses de l’artere renale chez l’enfant pour en faire le diagnostic et guider la conduite therapeutique

Results of receiver performance for the detection of a coherent light signal in the presence of background radiation are obtained and compared for two different types of background noise: a Gaussian field and a Gaussian field whose complex analytic signal is not circularly symmetric. It is pointed out that the randomness in the noise radiation does not significantly diminish the probability of correct detection.

Quantum communication is studied under the cutoff-rate criterion for transmission of M coherent states of real amplitude (M=2, . . ., 7). Comparison with other measurement operators of interest, such as the number and the quasi-classical operators, is made. The effect of quantization of the decision level on the cutoff rate is considered. >

Results on the efficient capacity of a noisy PPM photon-counting channel are presented. Optimization of the optical system over the number of slots, the average power of the signal and the frequency bandwidth of the channel is considered and studied. The regions of quantum-limited operations and weak background noise are analysed in detail.

The cutoff rate R/sub 0/ of an optical channel transmitting M-ary phase-shift-keyed coherent states is calculated for a receiver using the optimum quantum detector. This is shown to be nearly equivalent, when M>>1, to a system that measures the phase of the received signal. A closed-form expression yielding a good approximation to R/sub 0/ is derived. >

Cut-off rate performance for phase modulated coherent states in a noiseless optical channel, using a phase measurement operator is calculated. The results obtained deviate significantly from those given by the upper bound, shown previously to be close to the optimal in several practical cases.

Pharmacokinetic studies of amikacin were investigated concomitant with the kidney and auditory function and with the minimal inhibitory concentrations on the causal bacteria in 23 premature neonates with confirmed or suspected infections. It appears that the usual doses administered to newborns or preterm neonates must be decreased and calculated in function to the gestational ages and life's days. Every preterm neonates received a first infusion of 7.5 mg/kg. When premature have less than 14 days of life, amikacin was given at a dose of 3.75 mg/kg every 12 hr for a gestational age less than 32 weeks, or 5 mg/kg every 12 hr for a gestational age between 32 and 37 weeks. After 14 days of life, considering our results it must be necessary to administer 5 or 7.5 mg/kg every 12 hr.

Fifty-eight patients with colorectal liver metastases were treated by intra-arterial hepatic chemotherapy (IAHC) containing 5 FU (n = 42) or FUDR (n = 16). Twenty-three patients (39.6 p. 100) complained of abdominal pain. In three of these patients, the course was complicated by digestive hemorrhage. Endoscopic explorations and angioscintigraphy were normal in 4, showed oesophagitis in 3, superficial gastritis or duodenitis in 8 (34.7 p. 100) and gastric (2) or duodenal ulceration (6) in 8 (34.7 p. 100). The duodenal ulceration was extensive and considered to be cause of hemorrhage in two cases. Duodenal perforation due to the catheter was discovered in two other cases, one of which was secondary to tumoral extension revealed by forceps biopsy. This patient died 3 months later. Surgical treatment was mandatory in the other case due to digestive hemorrhage but did not prevent death. Angioscintigraphy performed in 15 patients with gastroduodenal inflammation or ulceration was normal in 7 patients, revealed arterial thrombosis in 5 and an extra-hepatic perfusion in the gastroduodenal area in 3 : this was related to a small pyloric artery which was occluded secondarily. IAHC was continued there after. This experience underlines the importance of exploring patients with digestive symptoms during IAHC so that it may be temporarily discontinued while an inadequately positioned infusion catheter may be corrected should gastroduodenal ulceration occur.

The pharmacokinetic studies of anticancer drugs still go on after they are out on the market. The therapeutic protocol is then more precise and depends on the dosage results of these drugs. Some examples of dosage adjustment according to their plasmatic level are reported here for high-dose methotrexate infusion and for CDDP infusion over five days. The test dose and the bayesian method (pharmacokinetic population) are used to predict the adapted individualized dosage for each patient.

Numerical computation of the capacity of binary optical channel using photon-counting techniques have been performed. The results obtained show that the so-called two-photon coherent source (TCS) provides a better transmission of the optical information than ideal laser as already demonstrated by Yuen and Shapiro and also by Helstrom from the probability of error point of view. On the other hand rather than using capacity as the criterion, the R cut-off rate is taken as the practical measure of communication performance of the channel. This cut-off rate is calculated in different cases and for soft and hard decisions. A method of calculation to bound R is developed from the relationship between the channel reliability rate function and the probability generating function. This simple method is thus utilized to compare the performance of the channel with TCS and those obtained with ideal laser.

In the course of chronic renal failure, aluminium may deposit and accumulate in different tissues. The aluminium content of parathyroid glands was measured in 31 haemodialysis patients at the time of a parathyroidectomy. The values were compared with those obtained from ten control patients with primary hyperparathyroidism without renal failure, and were related to bone remodelling. Of the 31 patients, 27 had a bone biopsy after double tetracycline labelling, at the time of parathyroidectomy. Twenty-one patients had severe hyperparathyroidism, three patients had hyperparathyroidism associated with osteomalacia, three patients had mild hyperparathyroidism with reduced bone formation. Seven patients had bone aluminium deposits, associated with osteomalacia in one case. The parathyroid aluminium was 62 +/- 35.7 (mumol/g glandular dry weight) in haemodialysis patients and 14.3 +/- 6.3 in control patients (P less than 0.001). A significant positive correlation existed between parathyroid aluminium and serum aluminium (P less than 0.01). The parathyroid aluminium was not different in the patients with and without bone aluminium deposits. A weak correlation was found between parathyroid aluminium and plasma parathyroid hormone. A significant negative correlation existed between parathyroid aluminium and osteoblastic surfaces (P less than 0.05), but no correlation was found with bone formation rate at tissue and bone multicellular units levels. We conclude that aluminium accumulates in parathyroid glands of dialysed patients. Severe hyperparathyroidism may coexist with aluminium overload of parathyroid glands. A marked aluminium overload, however, may cut short the course of hyperparathyroidism and may decrease parathyroid function and cellular activity in bone.