Epidemiology, clinical presentation, treatment, and follow-up of mercury poisoning in China: a retrospective analysis

Background: There are no reports on the incidence of mercury poisoning in a large population in China. This study investigated the epidemiology, clinical manifestations, treatment, and follow-up of Chinese patients with mercury poisoning. Methods: Data for 288 mercury poisoning patients were collected at our hospital from July 2014 to September 2019, including sex, age, admission time, blood mercury content, urine mercury content, creatinine, urinary mercury/creatinine ratio, 24-h urinary protein levels, electromyography (EMG) ndings, renal biopsy, and follow-up. Patient characteristics were evaluated by statistical and correlation analyses. Results: First, mercury poisoning in China mainly occurred through occupational exposure and the inappropriate use of mercury-containing cosmetics and Chinese folk remedies (CFRs). Second, the most common symptoms were nervous system (50.3%), kidney (16.4%) and breathing (8.0%). Mercury poisoning-induced Nephrotic syndrome (NS) and peripheral neuropathy are common long-term complications. The complications of occupational and cosmetics-induced mercury poisoning are consistent with international belief. However, the NS caused by CFRs is mainly membranous nephropathy and the probability of peripheral neuropathy caused by CFRs is higher than other pathogens. Third, follow-up data shows that 13 patients with EMG-conrmed neurological injury, 10 showed full recovery after 38.5 ± 8.03 months. Furthermore, among 18 patients with NS, 15 had normal urine protein and serum albumin levels after 22.67 ± 10.26 months. Conclusions: Regulation of skin-lightening cosmetic products, safety surveillance of CFRs, and prevention and control of occupational exposure must be improved to decrease the incidence of mercury poisoning in China. distributed parametric data the mean ± standard Parametric data with skewed distributions are presented as the median ± interquartile range (IQR). Differences between groups were compared using the Student’s t-test and Kruskal–Wallis test for normally and non-normally distributed variables, respectively. Correlation analysis between multi-factor and peripheral neuropathy: For the categorical data, we did chi-square test. For the continuous variables, we logistic analysis. Two-sided p values are presented, p IQR, interquartile range; MCN, minimal-change nephropathy; MN, membranous nephropathy; MSPGN, mesangial proliferative glomerulonephritis; Reporting of Observational studies

used for centuries for commercial and medical purposes and are a common constituent of traditional Chinese medicines (TCMs) [11,12]. Unlike TCMs, which are formally approved medical treatments and have reasonably safe levels of bioaccessible Hg [13], Chinese folk remedies (CFRs) involve an informal use of TCMs, rely mostly on experience rather than on formal teachings, and have not been approved by the government [14]. Although incidents of mercury poisoning in China have been occasionally reported domestically and abroad [15,16], there have been no clinical studies involving a large sample size to date. In the present study, we carried out a retrospective analysis of the clinical manifestations, treatment, and follow-up of Chinese patients with mercury poisoning. Our ndings provide important information for clinicians and health authorities on the pathogeny of mercury poisoning in China.

Patients
This study was a single-center, retrospective analysis conducted at a 52-bed poisoning treatment center in Fengtai District, Beijing, China. Poisoning patients were mainly admitted from northern, central, and eastern China. The STROBE (strengthening the Reporting of Observational studies in Epidemiology) criteria were followed. In this retrospective study, we analyzed the data for patients treated for mercury poisoning between July 2014 and September 2019 at The Poisoning Treatment Department of The Fifth Medical Center of the PLA General Hospital in China. In total, 288 cases were screened according to the owchart (Fig. 1). The inclusion criteria were as follows: (1) mercury was detected in the blood and urine, and the urinary mercury/creatinine (UMC) ratio was > 20 µmol/mol [17]; and (2) at least one of the following symptoms was present: neurasthenia syndrome (dizziness, fatigue, insomnia, dreaming, forgetfulness, and inattention), tremor, limb or trunk pain, oral gingivitis, proteinuria, etc. The exclusion criteria were as follows: (1)

Statistical analysis
Statistical analyses were performed using the STATA v.13.0 software. Categorical data are reported as numbers and differences were evaluated using the χ 2 test. Continuous variables were assessed for departure from normality using the Shapiro-Wilk W-test with α = 0.10. Normally distributed parametric data are presented as the mean ± standard deviation. Parametric data with skewed distributions are presented as the median ± interquartile range (IQR). Differences between groups were compared using the Student's t-test and Kruskal-Wallis test for normally and non-normally distributed variables, respectively. Correlation analysis between multi-factor and peripheral neuropathy: For the categorical data, we did chisquare test. For the continuous variables, we did logistic analysis. Two-sided p values are presented, which were considered statistically signi cant at p < .05.

Baseline characteristics
Between 2014 and 2019, 288 patients (189 females and 99 males) were hospitalized for mercury poisoning. The mean age was 38.1 ± 12.2 years (range: 3-71 years). In 84 patients (29.27%), the poisoning was related to cosmetics: the products were not registered in the Administration for Industry and Commerce; in some cosmetic products, an excessive mercury content was detected at professional institutions; and some users presented a phenomena that deactivating the same cosmetics products for 6 to 12 months, their UMC was keep in the normal range. In 66 patients (23%), the poisoning was occupational (forestry work, mercury mining, metallurgy, and machine manufacturing), and mercury and inorganic mercury could clearly be detected at the factory. In 46 patients (16%), the poisoning was related to CFRs; these patients took CFRs during or before the onset of symptoms, mostly for simple, chronic, and benign illnesses, including the management of psoriasis (n = 19, 41.3%), venous thrombosis (n = 1), cervical lymphadenopathy (n = 1), acne removal (n = 2), insomnia (n = 4), migraine (n = 3), depression (n = 1), glioma (n = 1), facial paralysis (n = 2), allergic purpura (n = 1), urticaria (n = 2), rubella (n = 1), pruritus of the perineum (n = 1), and cervical spondylopathy (n = 2), but also as an anodyne (n = 1) or for health preservation (n = 4). Most CFRs contain cinnabar, calomel, or other mercury-containing Chinese herbal medicines and even directly use mercury beads for heating inhalation. In the remaining 92 patients (32%), the source of poisoning was unclear; however, accessory examination showed HUM, accompanied by systemic symptoms such as headache, dizziness, tremor, or obvious proteinuria (++).
Comparison of patients with occupational mercury poisoning, mercury poisoning caused by cosmetics, and mercury poisoning caused by CFRs showed no obvious differences in their blood mercury content (χ 2 = 1.05, p = .59), urinary mercury content (χ 2 = 3.29, p = . 19), and UMC ratio (χ 2 = 3.86, p = .14). The occupational mercury poisoning in male is more susceptible to females. (OR = 1.75, χ 2 = 3.86, p = .049). Moreover, the age distribution showed that the pathogenesis in men had an earlier onset than that in women (χ 2 = 3.883, p = .049). The men were between 20 and 50 years old, while the women with occupational mercury poisoning ranged in age from 30 to 50 years old. Meanwhile, female patients between the ages of 20 and 50 years old were more likely to experience mercury poisoning through the use of cosmetic products (χ 2 = 21.22, p < .001). Among the patients with mercury poisoning caused by CFRs, there was no sex difference (χ 2 = 0.004, p = .949). Moreover, there was no difference in the age distribution (χ 2 = 6.4 < χ 2 [0.25, 5] = 6.63, p > .25; Fig. 2).

Clinical ndings
The median duration of hospitalization was 14 days (range: 3-67 days). There was no correlation between the duration of hospitalization and the source of mercury poisoning (χ 2 = 1.86, p = .396). Table 1 shows the symptoms of the 288 patients with mercury poisoning. The most common general symptoms were neurological, nephrological, respiratory, gastrointestinal, and dermatological.   Fig. 3B). EMG was performed in 39 patients with neuroparesthesia (Table 3). We examined the relationships between the EMG ndings and age, sex, duration of hospital stay, pathogeny, blood mercury concentrations, urine mercury concentrations, creatinine, and UMC ratios and found that as the age, the probability of mercury-induced peripheral neuropathy is increased (OR = 1.07, χ 2 = 3.93, p < .001). Patients with mercury poisoning caused by improperly used CFRs were more likely to have peripheral neurogenic damage (χ 2 = 13.6, p = .003).

Diagnosis and treatment
All 288 patients were diagnosed with mercury poisoning based on the clinical manifestations and UMC values and underwent symptomatic treatment, which included intramuscular injection (0.25 g twice a day) of sodium 2,3-dimercapto-1-propanesulfonate (DMPS) for 3 consecutive days, followed by 4 days of intermittent treatment, for mercury excretion, and detecting once UMC ratio after twice courses.
Combining some symptomatic treatments such as administration of calcium carbonate (3 g twice a day orally), Shenkang injection (100 mL once a day as an intravenous infusion), painkillers (10 mg of tramadol once or twice a day or 5 mg of oxycodone), glucocorticoids (prednisone), and neurotrophic drugs (20 µg/day mouse nerve growth factor for injection, 1.5 mg/day cobamamide, or 0.5 g/day citicoline sodium). For treatment of the 39 patients with NS, prednisone was administered at 1 mg/kg/day for 8 weeks, followed by a progressive decrease by 5 mg (1 tablet) every 2 weeks, based on the 24-h urinary protein level. Prednisone administration usually resulted in a rapid (within 2-3 days) improvement of the patient's clinical status, with a decrease in fever. In the cases with nerve injury, a neurotrophic factor was administered to reduce the myelin swelling and prevent the nerve ber degeneration.

Follow-up
Some patients were from underdeveloped regions, of which 280/288 patients were rechecked within 6 months to 1 year after discharge, and their UMC values were found to be in the normal range (outpatient query system). Unfortunately, after the rst re-examination, most patients did not continue to seek treatment and were followed up via telephone calls in September 2018. Of the 13 patients with EMG-con rmed neurological injury, who were followed up, 10 patients showed full recovery after 38.5 ± 8.03 months (EMGs of eight patients con rmed recovery, and two patients con rmed that they had no discomfort), two showed improvements relative to their previous state, and one showed no change.

Discussion
We report the epidemiology, treatment, and prognosis of patients admitted with mercury poisoning to a hospital in China. We mainly clari ed the following three aspects: 1) different mercury poisoning-induced pathogenies show sex and age difference; 2) there was a certain relationship between different pathogenies and mercury poisoning-induced complications; and 3) a clear diagnosis and timely treatment of mercury poisoning are necessary.
The gender and age distribution of occupational mercury poisoning patients are as described above, but most of the occupational exposure are hospitalized in occupational hospitals throughout the country, so the distribution is worth further discussion. Meanwhile, we report that mercury-containing cosmetic products are the main cause of mercury poisoning in females. Mercury inhibits the melanin pigment production by competing with the copper in tyrosinase [10]. Mercury is absorbed by the skin and can cause many complications such as nephrotic syndrome and peripheral nerve injury. Unfortunately, skinlightening creams containing inorganic mercury are widely used by dark-skinned people to achieve a lighter skin tone [10,18]. In China, TCM can be confused with CFRs. Most patients from this study improperly used CFRs for simple, chronic, and benign illnesses [14], the situation of this study are consistent with above. Since ancient times, patients taking mercury-containing Chinese medicines represent a considerable proportion in China, regardless of whether they are directly prescribed or improperly use mercury-containing TCMs [19].
Long-term mercury exposure causes damage to several organ systems, including the nervous, urinary, digestive, and respiratory systems [20]. Neurological manifestations of inorganic mercury intoxication include weakness, numbness, paresthesia, muscle cramps or atrophy, diminished muscle stretch re exes, paresis, fasciculations, and sensory loss [21]. The conventional belief is that occupation-induced mercury poisoning is the most common [22][23][24]. There are only a few domestic and overseas reports related to mercury poisoning caused by TCMs [14]. In this study, EMG showed that most nerve damage in patients was caused by CFRs, likely because occupational disease patients tend to be admitted to occupational disease hospitals. Mercury has a strong a nity for renal tissue which can lead to NS [18,[25][26][27]. Mercury-induced MN results from long-term use of mercury-containing skin-lightening cosmetics or from occupational contact with mercury [28][29][30]. Minimal-change glomerular lesions were detected in women in Kenya, who used mercury-containing skin-lightening creams [31], and four cases of MCN were described following the use of mercury-containing skin-lightening cream for 2-6 months [32]. Our study further supported the belief that MN and MCN occur following the use of mercury-containing skinlightening creams. However, MN was also observed following the inappropriate use of CFRs [13]. The results of this study are consistent with those of most international mercury poisoning cases and reports [33] that used small samples. In addition, it contributes to the treatment and prognosis of mercury poisoning induced by mercury-containing CFRs.
As previously mentioned [34,35], patients with mercury poisoning are often misdiagnosed in China as having digestive system diseases (acute abdomen and acute gastroenteritis), nervous system diseases (neurasthenia and vegetative nerve functional disturbance), and diseases of the urinary system (acute nephritis and nephritic syndrome). The reasons for the top two misdiagnosed diseases are incorrect history taking by doctors and a shortage of relevant knowledge about mercury poisoning. In China, mercury poisoning can only be de nitively diagnosed and treatments provided at prevention and treatment centers for occupational diseases, such as the Centers for Disease Control [14,17,35]. Prior to the establishment of poison treatment centers, mercury poisoning cases were mainly treated at the neurology, nephrology, and gastroenterology departments of hospitals, with most patients receiving symptomatic and not etiological treatment [36]. Moreover, although overseas researchers report some long-term follow-up prognoses for nephritic syndrome and peripheral nerve injury caused by mercury poisoning, domestic researchers do not report relevant aspects. Here, the follow-up results showed that the average recovery periods of nerve injury and nephritic syndrome caused by mercury poisoning were 38.5 and 22.67 months after dispelling mercury, respectively. Although there are speci c therapeutic drugs and a sound prognosis for mercury poisoning, the state of the illness is prolonged for most patients because of the lack of education in mercury poisoning and a shortage of national poison testing centers, most of which are only used for occupational disease groups. Public awareness of the main causes of mercury poisoning, including the use of cosmetic products and CFRs and occupational exposure, must be increased to relieve the heavy burden of mercury poisoning on both rural Asian communities and their healthcare systems.
This study has several limitations. First, there is a lack of multicenter studies, and there are only a few poisoning treatment centers in China. Although our center is the largest poisoning treatment center, it mainly receives patients from North, Central, and East China, while receiving only few patients from South and West China. Second, owing to the sparse distribution of domestic mercury poisoning centers, low incomes, and high mobility of patients, the second follow-up was only conducted over the telephone, with some proportion lost to follow-up. Third, to comply with China's national requirements, we adopted the GBZ-2007 National Mercury Poisoning Diagnostic Standards, which have not been updated for a long time and differed from international standards. Fourth, patients with NS caused by mercury poisoning did not receive renal biopsy and could only be evaluated in terms of clinical cure.

Conclusions
In this article, we report the epidemiology, treatment, and prognosis of patients with mercury poisoning at a hospital in China. Our study found that 1) poisoning in China mainly occurred through occupational exposure and the use of cosmetics and CFRs, as previously reported; 2) mercury poisoning due to occupational exposure was more common in males in the 20-50 year-old age group, while female patients aged 20-50 years were more likely to experience cosmetics-related poisoning. Mercury poisoning due to inappropriate use of CFRs was observed at all ages, regardless of the sex; 3) patients with mercury poisoning caused by the inappropriate use of CFRs were more likely to have peripheral neurogenic damage; NS caused by mercury poisoning was mostly of the MN and MCN types following the use of mercury-containing skin-lightening creams and of the MN type following the inappropriate use

Declarations
Ethics approval and consent to participate: All human participants signed a written informed consent.
Consent for participation in the study was obtained where participants are children (under 16 years old) Figure 1 Flow chart of patient screening in this study. UMC, urinary mercury/creatinine.

Figure 2
Distribution of mercury poisoning patients by disease etiology, sex, and age. CFRs, Chinese folk remedies.