Objectives The dose–response relationship for hand-transmitted vibration has been investigated extensively in temperate environments. Since the clinical features of hand-arm vibration syndrome (HAVS) differ between the temperate and tropical environment, we conducted this study to investigate the dose–response relationship of HAVS in a tropical environment.
Methods A total of 173 male construction, forestry and automobile manufacturing plant workers in Malaysia were recruited into this study between August 2011 and 2012. The participants were interviewed for history of vibration exposure and HAVS symptoms, followed by hand functions evaluation and vibration measurement. Three types of vibration doses—lifetime vibration dose (LVD), total operating time (TOT) and cumulative exposure index (CEI)—were calculated and its log values were regressed against the symptoms of HAVS. The correlation between each vibration exposure dose and the hand function evaluation results was obtained.
Results The adjusted prevalence ratio for finger tingling and numbness was 3.34 (95% CI 1.27 to 8.98) for subjects with lnLVD≥20 ln m2 s−4 against those <16 ln m2 s−4. Similar dose–response pattern was found for CEI but not for TOT. No subject reported white finger. The prevalence of finger coldness did not increase with any of the vibration doses. Vibrotactile perception thresholds correlated moderately with lnLVD and lnCEI.
Conclusions The dose–response relationship of HAVS in a tropical environment is valid for finger tingling and numbness. The LVD and CEI are more useful than TOT when evaluating the dose–response pattern of a heterogeneous group of vibratory tools workers.
Statistics from Altmetric.com
Prolonged occupational exposure to hand-transmitted vibration has been known to cause a collection of disabling clinical disorders of the vascular, neurological and musculoskeletal components of the hands and upper limbs known as the hand-arm vibration syndrome (HAVS). The condition has been studied extensively among workers from temperate climatic countries, but there were only limited number of studies from tropical environments.1 In the tropical environment, the clinical presentation of HAVS tends to be predominantly neurological without the occurrence of vibration white finger.1–3
The information on the dose–response relationship between the vibration exposure and the occurrence of HAVS is important for risk management and disease prevention. There has been extensive work done to investigate the dose–response relationship between the hand-transmitted vibration exposure and vibration white finger.4–11 However, the dose–response data on neurological disorders are lacking.8 The currently available studies on the vibration dose–response relationship came mainly from the temperate climate countries,4 ,8 whereas no study has been conducted in the tropical environment. Since the clinical presentation of HAVS differs between the tropical and temperate environments, the dose–response information for the vibration white finger and other symptoms derived from the temperate climatic epidemiological data may not be applicable in tropical environments. Hence, it is important to study the dose–response relationship between the vibration exposure and the outcomes in a tropical environment for effective control of the vibration hazards in the workplace.
In the current study, we investigated three groups of workers using vibratory tools and reported the exposure–response relationship between hand-transmitted vibration levels and relevant outcomes in a tropical environment. The result of this study is important for understanding the risk and severity of vibration exposure without the presence of cold weather provocation.
This study was conducted in Malaysia using a cross-sectional design. Malaysia has a tropical climate with uniform temperature, high humidity and a lot of rainfall. The average annual temperature throughout the country is above 25°C with daily variation from 24°C at night to 33°C during the day. Three groups of vibratory tools workers, the construction workers, forestry workers and automobile manufacturing plant workers were recruited into this study. The construction workers worked at a shopping mall construction site in an urban area of Selangor and the workers used mainly concrete breakers, grinders, impact drills and powered cutters for their daily work. The forestry workers recruited into this study were the tree fellers, located in a logging camp about 30 km from the sea coast of Bintulu, Sarawak and used only chainsaw for the tree felling task. The automobile manufacturing workers used mainly impact wrench and occasionally powered drills for the automobile assembly processes in a factory located at a rural area of Selangor. All subjects recruited into this study were male. The sampling was convenient due to logistic consideration as the subjects recruited into the study required approval from the management of the respective company. For construction workers, the workers were recruited among the manual labour workers, chosen by the management of the company according to the availability of the workers away from work, during the data collection period. For forestry workers, all the tree fellers located in a base logging camp were recruited into the study. The subjects at the automobile manufacturing plant were selected from the operating line of an assembly factory of an automobile manufacturing company and we recruited them based on voluntary participation.
The data collection was carried out between August 2011 and 2012. It consisted of questionnaire interview, hand function evaluation and vibration measurement. Each subject first underwent a face-to-face questionnaire interview by an occupational physician followed by a series of hand function tests. Thereafter, the subject was required to perform a simulated work process using the actual vibratory tools for their daily work, and the vibration level of the respective tool was measured.
The questionnaire interview consisted of detailed information on the extent of vibration exposure , that is, employment history, types of vibratory tool use, daily, yearly and total years of vibration exposure and symptoms related to possible HAVS such as finger tingling, numbness and dullness sensation, white finger, musculoskeletal pain of the upper limbs, finger coldness and detailed information on the frequency, occurrence and severity of each symptom. The interview was carried out face to face by a trained occupational physician to ensure that each subject understood the question correctly so that the accurate exposure and outcome information could be obtained.
The hand function evaluation consisted of the measurement of fingertip skin temperature, fingernail capillary return time, finger vibrotactile perception threshold, pinch strength and grip strength. The skin temperature of the pulp of all fingers was measured using an infrared thermometer (IT-550S, Horiba Ltd, Japan). The finger capillary return time was obtained for all fingers using the nail press test procedure. The fingernail bed for each finger was pressed gently for 10 s, and the time taken for the blood to refill the capillary vasculature was measured using a stopwatch. The grip strength of both hands was obtained using an analogue dynamometer at sitting position with both arms at neutral position, elbow flexed and forearms pronated at 90°. The patient was asked to grip the dynamometer as strongly as possible using one hand followed by the other hand. The procedure was repeated for five rounds and the average grip strength of each hand was then calculated. The pinch strength between the thumb and the index, middle and ring finger was measured, respectively, using a pinch metre (Takei Kiki Kogyo Co. Ltd). The patient was instructed to pinch the gauge using the thumb pad and the distal phalanx of each finger as strongly as possible while preventing other fingers from assisting in the pinching effort. The vibrotactile perception threshold for all fingers was measured using the vibration sensation metre (AU-02, Rion Co Ltd, Japan) with the vibration frequency setting at 125 Hz. During the test, the patient was asked to touch the vibrator probe without a surround with the pulp of a finger while other fingers, hand, wrist and arm positioned to prevent contact with the surface of the vibrator and examination table. The vibration level was then increased gradually at 2.5 dB interval from −10 dB. The patient was asked to lift up the other hand immediately when he felt the vibration. If the responses from the patient were inconsistent, the examiner turned the vibration down and increased gradually again repeatedly to ensure that the exact ascending vibration perception threshold was obtained. The vibration level with at least two out of three similar responses was considered as correct vibration perception threshold.
All subjects were instructed to stop using vibratory tools for at least 12 h, not to smoke for at least 2 h and not to take alcohol for at least 1 day prior to the conduct of the hand function tests. All tests were performed in an indoor environment with the room temperature controlled between 23°C and 25°C. A period of 15 min acclimatisation was allowed for each subject before the commencement of the hand function tests.
Following the questionnaire interview and hand functions evaluation, each worker was asked to perform a vibration work process outside the examination room using the actual vibratory tool and target material used by the worker for his daily work. For construction and forestry workers, a simulated work process was arranged because of logistic issues. The construction workers used the concrete breaker provided in the workplace to hit on an unwanted concrete surface to simulate the actual work processes performed by them. The forestry workers used their own chainsaw to cut the log provided by the company for the purpose of tree felling training. The direction of cutting and the log used simulate very closely the actual tree felling process carried out in the field. For the automobile assembly workers, the vibration measurement of the real work process was carried out on the production line for each worker. Some subjects used more than one tool, especially those among the construction workers. For the subject who used more than one tool, we measured the vibration level for the most frequently used tool.
The vibration level was measured using a triaxial accelerometer (Quest Technologies VI-410 Real-Time Hand-Arm and Whole Body Vibration Analyser Kit). The measurement was conducted by a qualified operator according to the standard of measurements as specified in the ‘ISO5349–2: Mechanical vibration—Measurement and evaluation of human exposure to hand-transmitted vibration—Part 2: Practical guidance for measurement at the workplace’ documentation.12 The accelerometer was firmly attached to the tool handle using a nylon strap at the position nearest to the hand grip according to the recommendation as specified in the ISO5249–2. A minimum of three readings were taken for each tool and each reading was taken for at least 1 min. The result of the measurement was recorded and combined with the vibration exposure information from the questionnaire interview to derive the vibration exposure dose for each worker.
For the current study, we have calculated three types of vibration exposure dose to examine their association with the outcomes. First, we calculated lifetime vibration dose (LVD), originally formulated by Griffin as follows4:
where the ahvi is the vibration total value of the acceleration magnitudes of each tool in three orthogonal axes and is given by the mathematical formula:
where ahwx, ahwy and ahwz are the frequency-weighted root mean square accelerations in the x axis, y axis and z axis, respectively; thi is the daily vibration exposure duration of each tool (hours/day), tdi is the number of working days per year for each tool and tyi is the total number of years used for each tool. Second, we calculated the total hand tool operating time (TOT) using the formula suggested by Miyashita et al13 as follows:
where thi is the daily vibration exposure duration of each tool (hours/day), tdi is the number of working days per year for each tool and tyi is the total number of year use for each tool. Third, we calculated cumulative exposure index (CEI) using the following formula as suggested by Sauni et al5:
where A(8) is the 8-hour time weight average vibration exposure level for each tool and is given by the formula:
where ahvi is the vibration total value for each tool, thi is the daily vibration exposure duration of each tool (h/day), tdi is the number of working days per year for each tool and tyi is the total number of year use for each tool. As the range of the vibration dose values were too wide, we used the logarithm values for the purpose of subsequent statistical analysis. The log LVD, log total operating time (TOT) and log CEI were categorised into three groups respectively according to the quartiles where the second and the third quartiles were combined and regressed against the binary outcomes of the HAVS symptoms. The log vibration dose values were also correlated with the continuous outcomes of the hand function tests to investigate any possible association between the variables.
This study has received written approval from the Medical Ethics Committee of the University of Malaya Medical Centre, the Committee of Ethics of the Wakayama Medical University, Japan as well as the management of the respective company participated in this study. Informed consent was obtained from each subject prior to the data collection. All subjects were informed on the procedure, confidentiality and non-liability of the employer in relation to this study. The subjects were also made to understand that any information provided or obtained in this study would not be used for the purpose of compensation.
Data entry, data cleaning and data analysis were carried out using Statistical Package for Social Sciences (IBM SPSS Statistics) V.19. The data entry was validated using the double-entry method. We analysed the dose–response relationship between the vibration dose and the outcomes of the HAVS by calculating the adjusted prevalence ratio using general linear model Poisson distribution function from STATA Intercooled V.11. The Spearman r correlation was used to find the association between two quantitative variables. The significance level for all statistical tests was set at 0.05.
A total of 173 male subjects were recruited into the study. Two subjects were excluded from the analysis due to the missing data on vibration level measurement. The characteristic of all the subjects in this study is presented in table 1. The majority of the subjects were Malay. Other ethnicity comprised Iban (10%), Indonesian (16%), Bangladesh (6%) and other minor aborigine groups (3%). The long-term medical illnesses suffered by some subjects as shown in table 1 were a mixture of hypertension, diabetes mellitus and low back pain, all of which were well controlled with medication. The types of previous injuries suffered by some of the subjects were superficial skin cuts, non-permanent blunt injuries to the back, finger amputations and elbow fracture not involving nerves. Most of the alcohol consumption was by the Iban ethnic group, but they usually consume alcohol during festive seasons.
The main sources of vibration exposure in this study were from the chainsaw, concrete breaker and impact wrench. Other tools included the grinder, powered cutter and impact drill. The summary of vibration exposure information for all subjects is shown in table 2. The mean (SD) ahv of chainsaw, concrete breaker and impact wrench were 7.16 (3.32) ms−2, 22.98 (6.02) ms−2 and 3.84 (2.59) ms−2, respectively.
The prevalence of tingling, numbness and dullness of the finger among the subjects was 25.1%. None of the subjects reported white finger but 19.3% of the subjects reported experiencing coldness of the fingers during their vibration exposure lifetime. The neurological symptoms occurred mainly during early morning, at night and after work, several times per month regardless of the season of the year. The finger coldness occurred mainly during rainy seasons and in the early morning, several times per month. The prevalence of musculoskeletal pain over the fingers and hands was only 4.7%, whereas over the upper limbs proximal to the wrists it was 10.5%.
Table 3 shows the adjusted prevalence ratio for the symptoms of HAVS according to the three types of vibration exposure dose adjusted for age, race, history of previous injury and history of alcohol intake. These four factors were selected for adjustment due to the significant differences in the values between the quartiles of at least one of the vibration exposure dose (data not shown). Other factors showed no significant difference between the quartiles for all three types of vibration exposure dose.
From table 3, the prevalence of having finger tingling, numbness and dullness, and muscular pain of the fingers and hands increase with increasing lifetime vibration exposure dose and CEI but not with TOT. There is no significant dose–response relationship between the finger coldness or muscular pain of the upper limbs and any of the vibration exposure doses. The TOT does not predict any of the HAVS symptoms.
The correlation between vibration exposure dose and the results of hand function evaluation is shown in table 4. Vibrotactile perception threshold is found to moderately correlate with the log LVD and CEI. The correlation for TOT is weaker than LVD and CEI. The finger skin temperature, fingernails capillary return duration and hand grip strength do not correlate with any of the vibration exposure doses. The strength of the finger muscles has a very weak correlation with the LVD and CEI.
The results of this study require careful interpretation due to limitations of the study design, sample size and the use of logarithmic values. The cross-sectional design inherits potential recall bias, which was present in this study. The subjects were likely to overestimate the answer when giving the information on the duration and frequency of vibration exposure. On the other hand, the inability to measure all vibratory tools for the subjects who used more than one vibratory tool (among the construction workers) was likely to underestimate the extent of vibration exposure. As the number of subjects who used more than one vibratory tool was less than 50%, we expect the net effect on the uncertainty of the vibration exposure information was likely to be an overestimation rather than an underestimation. Overestimation of the vibration exposure could mean that the dose–response relationship reported in this study could occur at a lower level. This is important as it suggested the possible true existence of dose–response relationship for HAVS if any in the tropical environment.
Since the study design was cross-sectional, no concrete evidence can be drawn on the temporal relationship between the exposure and the outcomes. The power of the study was affected by the limited sample size of 173 vibration exposed workers. Hence, dose–response findings in this study require careful interpretation within the context of these limitations.
The subjective outcomes of this study could be affected by recall bias as well as interviewer bias. We tried to minimise the recall biases by providing photographs of vibration white finger and giving explanation on the tingling, numbness and dullness sensation using trained occupational physicians when interviewing the subjects face to face. Besides, the clinical features and pathogenesis of the HAVS and its relationship with the vibration exposure were not explained to the subjects prior to the interview. The interviewer bias could not be overcome effectively, but we masked the vibration measurement results from the interviewing occupational physicians by scheduling the vibration measurement after the interview and hand evaluation process for all participants. Since there was no comparison group for this study, we do not expect any misclassification bias in this study.
The logarithmic values were used for analysis in this study because of the skewed data and wide range of the exposure dose values. Similar method has been used by previous researchers when analysing the LVD in relation to HAVS outcomes.7 ,14 Hence, the dose–response relationship established in this study should be interpreted in terms of the logarithmic values of the vibration doses. As the vibration dose categories are directly divided by its quartile values, the dose–response pattern should be similar when using the corresponding non-logarithmic values.
The automobile assembly process workers were exposed to multiple short intermittent vibrations with relatively longer duration of rest period than the tree fellers and construction workers. The tree fellers were more likely to have a regular exposure and rest periods in a working day, whereas the pattern of exposure for construction workers was more variable and inconsistent depending on the phases of construction. These variations in the pattern of vibration exposure may contribute to the differences in the prevalence of the outcomes as described in this study. Unfortunately, we are unable to control this limitation of the study, and hence, the result of the dose–response analysis of the current study should be interpreted in the context of possible effects of these factors.
In the current study, the dose–response relationship was found for neurological symptoms but not for vascular symptoms. In the correlation analysis, the results were comparable as only the vibrotactile perception threshold correlated moderately with LVD and CEI. The dose–response relationship between the LVD and the neurological symptoms reported in this study is almost similar to the dose–response relationship as reported by Bovenzi et al7 for vibration-induced white fingers among forestry workers. For CEI, a similar dose–response pattern was found for the neurological symptoms, but at higher cut-off levels compared with the study by Sauni et al.5
For vascular symptoms, none of the subjects reported vibration white finger. This is consistent with previous studies in tropical environments.2 ,15 ,16 We had expected the presence of dose–response relationship for finger coldness in this study, but it was not the case. There was no correlation between the vibration exposure and both the finger skin temperature and fingernails capillary return. The reason could be attributed to the short duration of vibration exposure (mean total year of exposure=7.8 years) in the current study sample, whereby the vascular functions of the hand had yet to deteriorate significantly. Another possibility for this observation is that the vascular component of the hand is not effectively damaged by the vibration energy in a tropical environment due to the absence of synergistic effect of cold injury because the normal physiological functions of the peripheral vessels is closely affected by the environmental temperature change.
The occurrence of Raynaud's phenomenon is multifactorial, which could be due to vascular, neural or intravascular abnormalities.17–19 Prolonged contact with vibration energy could damage any or all of the components resulting in cold finger when it is mild or white finger if the damage was severe. Previous researches had found abnormality in the peripheral blood vessels among the vibration-exposed workers, which could explain the cause of vascular symptoms in HAVS. Takeuchi et al20 described arterial smooth muscle hypertrophy and periarterial fibrosis in a study of digital biopsies and Littleford et al21 reported minor structural abnormalities on nailfold microscopy in patients with vibration white finger. In our study, we found significant dose–response relationship for neurological symptoms but not for finger coldness. This could suggest a different pathogenesis for finger coldness from neurological symptoms at least among our studied population.
In the previous study conducted by Miyashita et al,13 the TOT correlated with the severity of the various symptoms of HAVS. However, in the current study, we did not find any relationship between the TOT and any of the outcomes studied. The TOT does not take into account the vibration total value received by each worker, and hence, in the current study, which comprised a heterogeneous group of workers using various types of vibratory tools, the TOT did not adequately represent the severity of the vibration dose received by each subject. In contrast, Miyashita et al13 investigated a homogenous group of forestry workers where the vibration total values received by the subjects were likely to be similar to each other and hence the TOT can be related directly with the severity of the outcomes.
The spectrum of HAVS symptoms was known to manifest differently between warm and cold environments. Since we found significant dose–response relationship only for the neurological outcomes in the warm environment, the existing vibration exposure standard that was developed based on the occurrence of vibration white finger might not be applicable to warm climates. More researches are required to ascertain whether a different set of vibration exposure limit is needed for tropical environment.
In summary, the dose–response relationship between the hand-transmitted vibration and the HAVS in the tropical environment can be expressed in terms of the neurological outcomes. The severity of the finger tingling, numbness and dullness, and vibrotactile perception threshold was directly related to the vibration exposure dose in a tropical environment. The vascular symptoms cannot be used as a proxy for the severity of HAVS in the tropical environment. Besides, the derivation of vibration exposure dose should include the vibration total value and total hand tool operating time that took into account the daily duration, yearly frequency and total years of vibration exposure. The results of this study, however, require careful interpretation within the context of aforementioned limitations.
What this paper adds
The clinical features of hand-arm vibration syndrome (HAVS) differ between tropical and temperate climatic environments.
The dose–response relationship between the hand-transmitted vibration and vibration white finger has been extensively investigated in temperate environment but not in tropical environment
We found that in the tropical environment, the severity of the neurological outcomes of HAVS was directly related to the vibration exposure dose, whereas the vascular outcomes cannot be used to explain the dose–response relationship.
The derivation of vibration exposure dose should include both the vibration total value and total hand-tool operating time that took into account the daily duration, yearly frequency and total years of vibration exposure.
More researches are required to ascertain whether a different set of vibration exposure limit is needed for tropical environment in view of the differences in the dose–response relationship and clinical outcomes of HAVS between tropical and temperate environments
Contributors ATS was the main researcher and was involved in all phases of the study, including study design, literature search, conduct of the study, data analysis and final article write-up. SM did the vibration measurement and was involved in supervisory work and reviewing the manuscript. JF, AD, VCWH, NM, MI, ST, KY were involved in data collection. ABAM was involved in supervisory work, statistical analysis and reviewing the manuscript. KM was the subject matter expert and performed the supervisory work and the reviewing of the manuscript.
Funding This study was funded by research grants from University of Malaya Research Grant (Grant No: RG276/10HTM), Grant-in-aid for Scientific Research (C) (23590750) from the Japan Society for the Promotion of Science (JSPS), and financial support from JSPS's RONPAKU (Dissertation PhD) Programme. The authors’ work is independent of the funders.
Competing interests None.
Patient consent Obtained.
Ethics approval Medical Ethics Committee of the University of Malaya Medical Centre, Kuala Lumpur, Malaysia.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.