The Association Between Vibrotactile and Thermotactile Perception Thresholds and Personal Risk Factors in Workers Exposed to Hand-Transmitted Vibration

Abstract

Background: Hand–arm vibration syndrome (HAVS) is a well-recognized occupational condition resulting from prolonged exposure to hand-transmitted vibration (HTV), characterized by vascular, neurological, and musculoskeletal impairments. While vibration exposure is a known risk factor for HAVS, less is understood about the role of personal risk factors and, particularly regarding neurosensory dysfunction. This study aimed to examine the association between vibrotactile (VPT) and thermotactile perception thresholds (TPT) and individual risk factors and comorbidities in HTV-exposed workers. Methods: A total of 235 male HTV workers were evaluated between 1995 and 2005 at the University of Trieste’s Occupational Medicine Unit. Personal, occupational, and health-related data were collected, and sensory function was assessed in both hands. VPTs at 31.5 and 125 Hz and TPTs (for warm and cold) were measured on fingers innervated by the median and ulnar nerves. Results: Multivariable regression analysis revealed that impaired VPTs were significantly associated with age, higher daily vibration exposure (expressed as 8 h energy-equivalent A(8) values), BMI ≥ 25, smoking, vascular/metabolic disorders, and neurosensory symptoms. In contrast, TPTs showed weaker and less consistent associations, with some links to smoking and alcohol use. Conclusions: These findings suggest that, in addition to vibration exposure, individual factors such as aging, overweight, smoking, and underlying health conditions significantly contribute to neurosensory impairment and may exacerbate neurosensory dysfunction in a context of HAVS. The results underscore the importance of including personal health risk factors in both clinical assessment and preventive strategies for HAVS and may inform future research on its pathogenesis.

1. Introduction

Hand–arm vibration syndrome (HAVS) is a complex and multifaceted pathological condition caused by prolonged exposure to hand–arm vibrations, typically experienced by individuals operating electrically, pneumatically, or combustion-powered tools and machinery that transmit hand–arm vibration in occupational settings [1,2,3]. The vascular symptoms of HAVS, such as vibration-induced white finger (VWF), have been extensively studied [4,5,6]. However, the neurosensory component of the syndrome also plays a critical role in the clinical profile of affected individuals. Vibration-induced neuropathy is characterised by deterioration of distal sensory functions, particularly tactile and thermal sensitivity in the hands and fingers, with symptoms including hypoesthesia (numbness), paraesthesia (tingling), reduced tactile discrimination, and impairment of fine sensorimotor functions and manual dexterity [4,5,6]. This aspect of HAVS presents significant challenges for individuals in terms of fine motor control, both in the workplace and in daily activities. Despite its clinical relevance, the mechanisms underlying neurosensory damage in HAVS are not yet fully understood, particularly with respect to the interaction between vibration exposure and individual risk factors.

To objectively detect early-stage alterations induced by exposure to hand-transmitted vibration (HTV) at the neurosensory level, the assessment of thermal perception thresholds (TPTs) and vibrotactile perception thresholds (VPTs), measured at the distal phalanges of the hands, are considered useful methods for detecting peripheral neurosensory dysfunction [7,8,9].

The purpose of VPT analysis is to evaluate the function of various populations of cutaneous mechanoreceptors and their large-diameter myelinated afferent fibres [10]. The measurement of TPT is regarded as a valuable test for assessing the function of small-diameter myelinated and unmyelinated fibres: specifically, unmyelinated C-fibres for heat sensation and myelinated Aδ-fibres for cold [11]. The methodology for VPT threshold analysis is standardised by the international standard ISO 13091-1 (2001), which also proposes reference values based on a limited number of cross-sectional studies [12,13].

Currently, there is no universally accepted method for the measurement and evaluation of TPT, although recommendations for the execution of thermal sensory testing procedures have been provided by national academies, scientific associations, and research networks [8,9,14].

As previously described, the measurement of VPTs and TPTs allows for the assessment of peripheral neurosensory alterations and the correlation of perceptual responses with specific physiological or pathological conditions of the somatosensory system. Individuals affected by hand–arm vibration syndrome (HAVS) often exhibit elevated perception thresholds to both tactile and thermal stimuli [15]. It has been suggested that such neurosensory impairment may result from both direct damage to nerve endings caused by mechanical vibratory energy and secondary ischaemic damage due to vibration-induced vascular changes, such as reduced blood flow and oxygen supply to peripheral nerves [5,6,16]. While the role of hand–arm vibration in the deterioration of neurosensory thresholds is well established, the contribution of individual risk factors to the onset, progression, and severity of neurosensory deficits in HAVS remains less well-understood and under-investigated. Personal factors such as age, anthropometric characteristics, lifestyle habits (e.g., tobacco and alcohol use), and comorbidities (e.g., diabetes, pre-existing neuropathies) may exacerbate the impact of vibration exposure, predisposing individuals to greater neurosensory damage. Current knowledge regarding the role of personal risk factors remains incomplete, and research in this area is limited. Only a few studies have thoroughly explored how these factors modulate the neurosensory component of HAVS or the specific ways in which they influence vibrotactile and thermal perception thresholds [17,18,19,20]. As a result, the quantitative contribution of personal risk factors to the deterioration of these sensory thresholds remains uncertain [6].

This uncertainty hinders both comprehensive risk assessment and the implementation of targeted interventions for individuals at increased risk of neurosensory damage due to a combination of vibration exposure and personal vulnerability [3]. A deeper understanding of the role of personal risk factors in the development of HAVS would enhance both prevention and management of the syndrome. Furthermore, it would be practically useful in improving the design and application of occupational health surveillance protocols, work fitness criteria, and the development of appropriate health promotion programmes aimed at the most vulnerable populations. The aim of this study was to evaluate the potential effect of personal risk factors on VPTs and TPTs in a population of workers occupationally exposed to HTV.

2. Materials and Methods

2.1. Study Population

Between 1995 and 2005, a total of 235 male workers exposed to HTV were consecutively recruited and examined at the Occupational Medicine Unit of the University of Trieste. For each participant, detailed medical history data were collected by specialists or trainees in occupational medicine. During the medical examination, the following personal and health-related information was obtained: age, body mass index (BMI), tobacco smoking habits (pack-years), daily alcohol consumption (1 alcohol unit = 12 g ethanol), industrial sector of employment and specific job task, presence of vascular and/or metabolic disorders (diabetes, hypertension, overweight/obesity), history of trauma or surgical interventions involving the neck and/or upper limbs, and cervical spine conditions.

In addition, based on the results of the medical history and a clinical neurological examination, peripheral neurosensory disturbances in the fingers and hands were classified according to the stages of the Stockholm Workshop Scale for Sensorineural Symptoms (SWS_SN): SWS_SN 0 = no neurosensory symptoms; SWS_SN 1 = intermittent numbness, with or without tingling; SWS_SN 2 = intermittent or persistent numbness and reduced tactile and/or thermal sensory perception; SWS_SN 3 = persistent numbness, reduced neurosensory perception, and impaired manual dexterity [16,21,22].

2.2. Definition and Measurement of Outcomes

VPT measurements were performed using an HVLab tactile vibrometer, while TPT for both warmth (W-TPT) and cold (C-TPT) were assessed using an HVLab thermal aesthesiometer. Both diagnostic instruments were developed and manufactured by the Institute of Sound and Vibration Research, University of Southampton, UK. VPT thresholds are defined as the minimum vibration intensity that a subject can perceive when a vibratory stimulus is applied to the fingertips. VPT measurements assess somatosensory tactile sensitivity, with the aim of evaluating the functional and anatomical integrity of cutaneous mechanoreceptors and their neurosensory afferent pathways.

W-TPT and C-TPT thresholds are defined, respectively, as the minimum temperature change (°C) perceived as warmth or cooling by the subject, measured using a Peltier cell-based thermal aesthesiometry system [6,12]. VPT measurements were conducted in accordance with the recommendations of international standard ISO 13091-1 [12] and were taken at frequencies of 31.5 and 125 Hz, which correspond to the response characteristics of Meissner’s corpuscles and Pacinian corpuscles, respectively [2]. Measuring VPTs at 31.5 Hz and 125 Hz allows for the assessment of rapidly adapting mechanoreceptors—Meissner’s corpuscles and Pacinian corpuscles. The primary rationale for this focus is that rapidly adapting receptors are the most affected by hand–arm vibration exposure, as they are directly responsible for detecting vibration stimuli within the frequency range most commonly emitted by power tools (i.e., 31.5–125 Hz). This makes them clinically and occupationally relevant targets for neurosensory assessment in the context of HAVS [5,12,23]. VPT values were expressed in decibels (dB) relative to an acceleration of 10⁻⁶ ms⁻² root mean square (r.m.s.).

Both VPT and TPT values were measured on the palmar surface of the distal phalanx of the II digit (innervated by the median nerve) and the V digit (innervated by the ulnar nerve) of both hands. The procedures for recording vibrotactile and thermal stress thresholds have been described in detail in a previous study by our group [2].

2.3. Covariates

To estimate the effect of personal, non-occupational risk factors on VPT and TPT, the following covariates were considered, as they have been suggested in the literature to potentially influence the neurosensory component of HAVS [5,19,24,25]: age (years); body mass index (BMI, kg/m²) classified as follows: <25 or ≥25 kg/m² [26]; smoking habits classified as follows: never smokers, former, or current smokers; alcohol consumption classified as non-consumers or consumers; presence of vascular and/or metabolic disorders (MVD) classified (Yes or No); history of trauma or surgical procedures involving upper limb and/or neck traumas (including surgeries)/disorders (UNTD) (Yes or No); and severity of peripheral neurosensory symptoms in the fingers and hands (SWS_SN = No in the absence of neurological symptoms, SWS_SN = Yes in the presence of symptoms at stage ≥ 1).

Daily exposure to vibration was assessed in terms of the frequency-weighted root mean square (r.m.s.) acceleration equivalent to an 8 h workday with vibrating tools, in accordance with ISO 5349-1:2001 [27] (A(8), ms⁻² r.m.s.). A(8) values were estimated individually, based on self-reported daily exposure durations to vibrating tools collected during medical history interviews, and using the median frequency-weighted vibration amplitudes (ms⁻² r.m.s.) of the specific tools used, as recorded in the Occupational Medicine Unit of the University of Trieste’s vibration database. The A(8) value was calculated using the following formula:

$$A(8)=\sqrt{\sum _{i=1}^n{a}_{hvi}^2{\displaystyle \frac{T_i}{T_0}}}\hspace{1em}\hspace{1em}\left(m{s}^{-2}r.m.s.\right)$$

where a_hvi is the total r.m.s. acceleration of the i-th tool, T_i is the duration of use of tool i (in hours), and T₀ is the reference period of 8 h.

For the purposes of data analysis, individual A(8) values were categorised as follows: A(8) < 2.5 ms⁻² r.m.s. and A(8) ≥ 2.5 ms⁻² r.m.s., in accordance with the action value proposed by the EU Physical Agents Directive 2002/44/EC for hand–arm transmitted vibration [28].

2.4. Statistical Methods

The main characteristics of the subjects enrolled in our study are described using the median, first quartile (Q1), third quartile (Q3), and by the mean and standard deviation (SD) for continuous variables, and by absolute frequencies and percentages for categorical variables. The associations between VPT and TPT values and occupational and personal risk factors were assessed using multivariable linear regression analysis. In the final models, the VPT and TPT values (dependent variables) for the II and V fingers were calculated as the mean of the values measured in both hands of the subjects. In the final models, VPT values were transformed to a logarithmic scale (ln-VPT) due to their skewed distribution. For this reason, the regression coefficients (β) represent the expected change in the ln-outcome per unit change in the predictor and the estimated relative change in the original outcome variable associated with a one-unit increase in the predictor is given by the following formula:

$$\%\hspace{0.17em}\mathrm{change}=\left(e^{\beta }-1\right)\times 100$$

In our final models, the presence of heteroscedasticity was tested using the Breusch–Pagan [29] method; to ensure reliable inferences in the presence of heteroscedasticity, robust standard errors were applied for the estimation of the multivariable linear regression parameters using the Huber–White method [30,31]. Statistical analyses were conducted using STATA software (StataCorp. 2019. Stata Statistical Software: Release 16. College Station, TX: StataCorp LP).

3. Results

Table 1. Characteristics of 235 male workers exposed to hand-transmitted vibration (HTV): personal, medical, and HTV exposure data.

	Median	Q1	Q3	Mean (SD)
Age (y)	46.0	39.0	54.0	46.5 (10.4)
BMI (kg/m2)	25.8	23.7	28.4	26.2 (3.8)
<25, n (%)	99.0 (42.1)
≥25, n (%)	136.0 (57.9)
Smoking in smokers (pack-y)	15.0	7.5	24.8	18.9 (16.7)
Non-smokers, n (%)	80.0 (34.0)
Former/current smokers, n (%)	155.0 (66.0)
Drinking in consumers (alcohol units/day)	2.0	1.0	4.0	2.5 (2.1)
Alcohol non-consumers, n (%)	122.0 (51.9)
Alcohol consumers, n (%)	113.0 (48.1)
Daily exposure to HTV, A(8) (ms−2 r.m.s.)	3.0	2.5	4.3	3.4 (1.3)
<2.5, n (%)	91.0 (38.7)
≥2.5, n (%)	144.0 (61.3)
Metabolic and/or vascular diseases, n (%)
Yes	7.0 (3.0)
No	228.0 (97.0)
Upper limb and/or neck traumas/disorders
Yes	28.0 (11.9)
No	207.0 (88.1)
Vibrotactile Perception Threshold (ms−2)—125 Hz (Digit II)	0.35	0.19	0.97	1.36 (3.51)
Vibrotactile Perception Threshold (ms−2)—31.5 Hz (Digit II)	0.25	0.16	0.41	0.54 (1.14)
Vibrotactile Perception Threshold (ms−2)—125 Hz (Digit V)	0.55	0.31	1.22	1.76 (3.89)
Vibrotactile Perception Threshold (ms−2)—31.5 Hz (Digit V)	0.26	0.16	0.42	0.63 (1.49)
Warm Perception Threshold (°C)—(Digit II)	40.2	37.6	43.6	40.55 (4.82)
Warm Perception Threshold (°C)—(Digit V)	40.3	37.8	44.1	40.90 (5.02)
Cold Perception Threshold (°C)—(Digit II)	25.2	22.0	27.1	24.11 (4.11)
Cold Perception Threshold (°C)—(Digit V)	23.0	19.8	25.7	22.09 (5.08)

summarises the main characteristics of the 235 male workers examined. The mean age of the subjects at the time of the medical examination was 46.5 years (SD ± 10.4), and the mean BMI was 26.2 kg/m² (SD ± 3.8). Regarding smoking habits within the study population, 66% of the subjects were current or former smokers (N = 155), with a mean consumption of 18.9 pack-years (SD ± 16.7). A total of 48.1% of the examined subjects reported consuming alcoholic beverages, even occasionally (N = 113); among these, the average number of alcohol units consumed per day was 2.5 (SD ± 2.1). The estimated daily occupational exposure to hand–arm transmitted vibrations was 3.4 m/s² r.m.s. (SD ± 1.3). Specifically, 38.7% of the workers were exposed to A(8) values < 2.5 m/s² r.m.s., whereas 61.3% were exposed to values ≥ 2.5 m/s² r.m.s., i.e., at or above the EU action value. According to medical history, 3.0% of the subjects reported metabolic and/or vascular diseases (N = 7, of whom four had metabolic diseases—including diabetes mellitus—and three had vascular diseases), while 11.9% had upper limb and/or neck traumas or disorders (N = 28). The VPT and TPT measurements for the digit II (median nerve) and digit V (ulnar nerve) are presented in Table 1. The median and mean VPT values at 125 Hz were 0.35 and 1.36 ms⁻² for digit II, and 0.55 and 1.76 ms⁻² for digit V. At 31.5 Hz, the median and mean VPT values were 0.25 and 0.54 ms⁻² for digit II, and 0.26 and 0.63 ms⁻² for digit V. The differences between the mean and median values indicate a skewed distribution of VPT thresholds among the examined subjects.

In the digit II, the median and mean W-TPT thresholds were 40.2 °C and 40.6 °C, respectively, while the C-TPT thresholds were 25.2 °C and 24.1 °C. In digit V, the median and mean W-TPT thresholds were 40.3 °C and 40.9 °C, while the C-TPT thresholds were 23.0 °C and 22.1 °C, respectively.

Table 2. Strata-specific daily vibration exposure A(8) (ms⁻² r.m.s.) by industry and occurrence of peripheral sensorineural disorders staged according to the Stockholm Workshop ’86 scale (SWS_SN).

		Total Hours of Exposure	Total Years of Exposure	A(8)
Industry	N (%)	Median	Median	Median	Q1	Q3	Mean (SD)	K-Wallis
Forestry	47.0 (20.0)	9240.0	9.0	5.2	4.6	5.7	5.0 (0.8)	0.0001
Construction	90.0 (38.3)	8360.0	24.5	4.3	2.5	4.3	3.7 (1.00)
Shipbuilding	22.0 (9.4)	8140.0	18.5	2.6	1.8	2.6	2.6 (0.7)
Engineering	62.0 (26.4)	13,200.0	25.0	2.5	2.0	2.7	2.4 (0.3)
Iron and Steel	14.0 (6.0)	11,220.0	15.5	2.5	1.9	3.0	2.4 (0.7)
SWS_SN (Stockholm Workshop ’86)
0—None	125.0 (53.2)	7700.0	18.0	3.5	2.5	4.6	3.6 (1.3)	0.0371
1—Mild	69.0 (29.4)	11,880.0	21.0	2.8	2.5	4.3	3.4 (1.3)
2—Moderate	31.0 (13.2)	10,560.0	20.0	2.6	2.0	3.5	2.9 (1.0)
3—Severe	9.0 (3.8)	16,500.0	15.0	3.0	2.5	3.7	3.2 (1.3)
Not reported	1.0 (0.4)	--		--

presents the distribution of A(8) values, stratified by industrial sector, and the prevalence of neurosensory disorders among the 235 examined male workers. The estimated A(8) exposure by industrial sector showed the highest values among forestry workers (median 5.2 and mean 5.0 ms⁻² r.m.s.), while the lowest values were observed in workers from the mechanical and steel industries (median 2.5 and mean 2.4 ms⁻² r.m.s.), with statistically significant differences between sectors (p = 0.0001). According to the medical history, 53.2% of workers reported no neurosensory disturbances in the fingers or hands, 29.4% had a mild SWS_SN score, 13.2% a moderate score, and 3.8% a severe SWS_SN score.

VPT and TPT measurements, stratified by industrial sector, A(8) exposure, and SWS_SN stages, are reported in in the Supplementary Materials (Tables S1–S4). Univariate analyses of the relationship between vibrotactile thresholds and risk factors showed significant crude associations between VPT at 125 Hz and both daily vibration exposure and severity of neurosensory symptoms, particularly in the second digit. VPT at 31.5 Hz and thermal thresholds (W-TPT and C-TPT) showed significant differences only across workers from different industrial sectors.

Multivariate linear regression analyses revealed that VPT values, expressed as ln-VPT, at both 31.5 and 125 Hz, were significantly and positively associated with age and vibration exposure (A(8)) in both digits II and V (

Table 3. Results of multivariable linear regression analysis of log vibrotactile perception thresholds (ln VPT) in digit II at 125 Hz and 31.5 Hz to personal and occupational risk factors.

ln VPT Digit II, 125 Hz (N = 214); R-Squared = 0.189		Adjusted β (95% CI)	p	ln VPT Digit II, 31.5 Hz (N = 213); R-Squared = 0.151		Adjusted β (95% CI)	p
Age (y)	-	0.03 (0.02; 0.04)	0.000	Age (y)	-	0.02 (0.00; 0.03)	0.002
BMI (kg/m2)	<25	Ref	--	BMI (kg/m2)	<25	Ref	--
	≥25	0.27 (−0.04; 0.58)	0.084		≥25	0.15 (0.05; 0.51)	0.018
Smoking	No	Ref	--	Smoking	No	Ref	--
	Yes	0.34 (0.02; 0.66)	0.035		Yes	0.40 (0.17; 0.62)	0.001
Alcohol	No	Ref	--	Alcohol	No	Ref	--
	Yes	0.06 (−0.26; 0.37)	0.719		Yes	0.16 (−0.07; 0.38)	0.177
Upper limb and/or neck traumas/disorders	No	Ref	--	Upper limb and/or neck traumas/disorders	No	Ref	--
	Yes	−0.11 (−0.70; 0.48)	0.719		Yes	−0.02 (−0.39; 0.36)	0.937
Metabolic and/or vascular disorders	No	Ref	--	Metabolic and/or vascular disorders	No	Ref	--
	Yes	0.73 (−0.28; 1.74)	0.157		Yes	0.76 (0.16; 1.35)	0.013
SWS_SN—Sensorineural disorders	No	Ref	--	SWS_SN—Sensorineural disorders	No	Ref	--
	Yes	0.34 (0.03; 0.66)	0.034		Yes	0.10 (−0.14; 0.34)	0.418
A(8) (ms−2 r.m.s.)	<2.5	Ref	--	A(8) (ms−2 r.m.s.)	<2.5	Ref	--
	≥2.5	0.51 (0.18; 0.83)	0.002		≥2.5	0.28 (0.05; 0.51)	0.018

and

Table 4. Results of multivariable linear regression analysis of log vibrotactile vibration thresholds (ln VPT) in digit V at 125 Hz and 31.5 Hz to personal and occupational risk factors.

ln VPT Digit V, 125 Hz (N = 214); R-Squared = 0.227		Adjusted β (95% CI)	p	ln VPT Digit V, 31.5 Hz (N = 213); R-Squared = 0.164		Adjusted β (95% CI)	p
Age (y)	-	0.04 (0.02; 0.05)	0.000	Age (y)	-	0.02 (0.01; 0.03)	0.003
BMI (kg/m2)	<25	Ref	--	BMI (kg/m2)	<25	Ref	--
	≥25	0.18 (−0.11; 0.48)	0.230		≥25	0.30 (0.01; 0.50)	0.040
Smoking	No	Ref	--	Smoking	No	Ref	--
	Yes	0.36 (0.06; 0.66)	0.019		Yes	0.39 (0.15; 0.64)	0.001
Alcohol	No	Ref	--	Alcohol	No	Ref	--
	Yes	0.30 (0.01; 0.59)	0.040		Yes	0.20 (−0.05; 0.45)	0.123
Upper limb and/or neck traumas/disorders	No	Ref	--	Upper limb and/or neck traumas/disorders	No	Ref	--
	Yes	−0.04 (−0.47; 0.39)	0.865		Yes	−0.21 (−0.54; 0.13)	0.223
Metabolic and/or vascular disorders	No	Ref	--	Metabolic and/or vascular disorders	No	Ref	--
	Yes	1.06 (0.13; 2.00)	0.026		Yes	0.77 (0.16; 1.38)	0.013
SWS_SN—Sensorineural disorders	No	Ref	--	SWS_SN—Sensorineural disorders	No	Ref	--
	Yes	0.18 (−0.11; 0.47)	0.221		Yes	0.08 (−0.17; 0.34)	0.519
A(8) (ms−2 r.m.s.)	<2.5	Ref	--	A(8) (ms−2 r.m.s.)	<2.5	Ref	--
	≥2.5	0.45 (0.16; 0.74)	0.003		≥2.5	0.25 (0.01; 0.50)	0.040

). Specifically, the detrimental percentage change in the VPT per unit increase of age varied between 2.0% (β = 0.02) and 4.1% (β = 0.04) (Table 3 and Table 4). The change in the VPT for those with an A(8) ≥ 2.5 varied between 28.4% (β = 0.25) and 66.5% (β = 0.51) (Table 3 and Table 4). Other associations found were (i) with overweight (BMI ≥ 25) for VPT at 31.5 Hz in both digits tested, with an increase of 16.2% when VPT was tested in digit II (β = 0.15) and of almost 35.0% when VPT was tested in digit V (β = 0.30); with tobacco smoking for VPT at 31.5 and 125 Hz in both digits, with a minimum VPT detected increase of 40.5% (β = 0.34) and a maximum of 49.2% (β = 0.40); with vascular/metabolic disorders for VPT at 31.5 Hz in digit II (113.8% in VPT increase; (β = 0.73)) and at 31.5 and 125 Hz in the digit V, with a minimum of almost 116.0% (β = 0.77) and a maximum of 188.6% (β = 1.06) in VPT increase; (ii) alcohol consumption showed a 35.0% in VPT increase of at 125 Hz in the digit V (β = 0.30) and peripheral neurosensory disorders a 40.5% VPT increase at 125 Hz in digit II (β = 0.34); (iii) no associations were observed with musculoskeletal disorders affecting the cervical spine or upper limbs (Table 3 and Table 4).

W-TPTs were found to be significantly associated with BMI and tobacco smoking in both digits, while an inverse relationship was observed with musculoskeletal disorders of the neck and/or upper limbs due to trauma or surgical interventions when W-TPTs have been tested in digit V (

Table 5. Results of multivariable linear regression analysis of warm thermal perception thresholds (W-TPT, °C) and cold thermal perception thresholds (C-TPT, °C) in digit II to personal and occupational risk factors.

W-TPT Digit II (N = 222); R-Squared = 0.073		Adjusted β (95% CI)	p	C-TPT Digit II (N = 222); R-Squared = 0.041		Adjusted β (95% CI)	p
Age (y)	-	0.02 (−0.04; 0.08)	0.493	Age (y)	-	−0.04 (−0.09; 0.02)	0.159
BMI (kg/m2)	<25	Ref	--	BMI (kg/m2)	<25	Ref	--
	≥25	1.61 (0.30; 2.92)	0.016		≥25	0.07 (−1.06; 1.20)	0.904
Smoking	No	Ref	--	Smoking	No	Ref	--
	Yes	1.68 (0.34; 3.03)	0.014		Yes	−0.76 (−1.94; 0.41)	0.199
Alcohol	No	Ref	--	Alcohol	No	Ref	--
	Yes	−0.45 (−1.77; 0.86)	0.495		Yes	−1.02 (−2.16; 0.12)	0.078
Upper limb and/or neck traumas/disorders	No	Ref	--	Upper limb and/or neck traumas/disorders	No	Ref	--
	Yes	−1.31 (−3.32; 0.71)	0.203		Yes	−0.02 (−1.78; 1.72)	0.978
Metabolic and/or vascular disorders	No	Ref	--	Metabolic and/or vascular disorders	No	Ref	--
	Yes	0.44 (−3.26; 4.13)	0.815		Yes	0.16 (−3.05; 3.37)	0.924
SWS_SN—Sensorineural disorders	No	Ref	--	SWS_SN—Sensorineural disorders	No	Ref	--
	Yes	1.02 (−0.27; 2.31)	0.120		Yes	−0.26 (−1.38; 0.86)	0.651
A(8) (ms−2 r.m.s.)	<2.5	Ref	--	A(8) (ms−2 r.m.s.)	<2.5	Ref	--
	≥2.5	−0.17 (−1.50; 1.16)	0.801		≥2.5	−0.82 (−1.98; 0.33)	0.161

and

Table 6. Results of multivariable linear regression analysis of warm thermal perception thresholds (W-TPT, °C) and cold thermal perception thresholds (C-TPT, °C) in digit V to personal and occupational risk factors.

W-TPT Digit V (N = 222); R-Squared = 0.089		Adjusted β (95% CI)	p	C-TPT Digit V (N = 223); R-Squared = 0.050		Adjusted β (95% CI)	p
Age (y)	-	0.02 (−0.04; 0.09)	0.475	Age (y)	-	−0.08 (−0.15; −0.01)	0.017
BMI (kg/m2)	<25	Ref	--	BMI (kg/m2)	<25	Ref	--
	≥25	1.88 (0.54; 3.23)	0.006		≥25	0.78 (−0.60; 2.16)	0.267
Smoking	No	Ref	--	Smoking	No	Ref	--
	Yes	1.58 (0.19; 2.97)	0.026		Yes	−0.35 (−1.78; 1.08)	0.631
Alcohol	No	Ref	--	Alcohol	No	Ref	--
	Yes	0.46 (−0.90; 1.81)	0.508		Yes	−1.13 (−2.52; 0.26)	0.111
Upper limb and/or neck traumas/disorders	No	Ref	--	Upper limb and/or neck traumas/disorders	No	Ref	--
	Yes	−2.24 (−4.32; −0.17)	0.034		Yes	0.83 (−1.28; 2.93)	0.441
Metabolic and/or vascular disorders	No	Ref	--	Metabolic and/or vascular disorders	No	Ref	--
	Yes	0.44 (−3.36; 4.23)	0.822		Yes	1.29 (−2.65; 5.23)	0.520
SWS_SN—Sensorineural disorders	No	Ref	--	SWS_SN—Sensorineural disorders	No	Ref	--
	Yes	1.15 (−0.19; 2.48)	0.092		Yes	−0.17 (−1.55; 1.20)	0.803
A(8) (ms−2 r.m.s.)	<2.5	Ref	--	A(8) (ms−2 r.m.s.)	<2.5	Ref	--
	≥2.5	0.29 (−1.08; 1.67)	0.674		≥2.5	−0.33 (−1.74; 1.09)	0.650

). C-TPTs showed only a negative association with age in digit V, and an inverse but non-significant relationship with alcohol consumption in both digits. No associations were observed between TPTs and vibration exposure (A(8)) or peripheral neurosensory disorders.

Finally, multivariate linear regression models did not reveal any statistically significant interactions between occupational vibration exposure (A(8)) and personal risk factors in the predictive estimation of VPT or TPT (data not shown).

4. Discussion

In the present study, the associations between occupational and personal risk factors and the neurosensory component of HAVS were investigated in individuals occupationally exposed to hand–arm vibrations.

The median values of VPTs in digit II and in digit V, measured bilaterally at frequencies of 31.5 Hz and 125 Hz, were overall consistent with those reported in the literature [32].

Similarly, the values of TPTs for both warm and cold stimuli were in line with those previously obtained by us [32], demonstrating good external concordance of our study results.

In a Swedish study conducted to measure TPTs in an adult population exposed to HTV, the mean C-TPTs, measured in digits II and V, were higher than those observed in the present study [33] (Table 1). This may be partly explained by a peripheral physiological adaptation to cold climates developed by individuals living in Nordic countries, and by the effect of lower average ambient temperatures on thermal perception testing. These factors may contribute to subjects perceiving cold sensations at higher temperatures. Several clinical and epidemiological studies have reported neurosensory alterations in the fingers of workers exposed to vibration; for example, among shipyard and dock workers, forestry workers, miners, and metalworkers [11,16,32,34,35,36,37]. Our results confirm a direct association between the deterioration of peripheral neurosensory function and daily exposure to hand-transmitted vibration (HTV), expressed in terms of A(8) [5,38,39,40]. These findings support clinical and experimental evidence that peripheral neurosensory dysfunctions in individuals exposed to HTV are caused by the damaging action of the mechanical energy from vibration on cutaneous mechanoreceptors (Meissner and Pacinian corpuscles) and their Aβ afferent fibres, as well as on unmyelinated C fibres (responsible for heat sensation) and unmyelinated Aδ fibres (responsible for cold sensation).

4.1. Vibrotactile Perception Thresholds

This study investigated the potential role of personal, non-occupational risk factors in the aetiopathogenesis of peripheral neurosensory disorders, in addition to the effects of HTV. These risk factors included age, smoking habits, alcohol consumption, certain metabolic conditions (such as diabetes or hypertension), vascular diseases, and a history of trauma affecting the cervical spine and/or upper limbs.

Overall, the results of this study showed that age, smoking habits, and a positive medical history for metabolic and/or vascular disorders were positively associated with alterations in VPTs at 31.5 and 125 Hz in the areas innervated by the median (index finger) and ulnar (little finger) nerves (see Table 3 and Table 4). This suggests that both occupational and individual factors may contribute to the deterioration of neurosensory function in the distal parts of the upper limbs. In our population, the impact of smoking status (smokers or ex-smokers vs. non-smokers) on the percentage worsening of VPT ranged from 40.5% to 59.2%, which is comparable to the deterioration observed in subjects with higher vibration exposure, whose VPT worsening ranged from 28.4% to 66.5% (Table 3 and Table 4).

From a histopathological perspective, ageing leads to a progressive degeneration of peripheral nerve fibres, particularly the large-calibre myelinated fibres responsible for transmitting vibratory sensation. It also results in reduced neuroplasticity and a diminished capacity for peripheral nerve tissue repair [41], as well as thinning of blood vessels and increased arterial stiffness, ultimately leading to reduced perfusion of peripheral nerve fibres [42]. With regard to the positive association between altered VPT and cigarette smoking, it is well known that nicotine and carbon monoxide induce peripheral vasoconstriction, worsening the reduction in blood flow to the hands and promoting ischaemia of the nerve fibres. Furthermore, increased oxidative stress caused by cigarette smoke exposure may render peripheral nerve cells more vulnerable to vibration-induced damage.

Lastly, comorbidities such as diabetes mellitus, dyslipidaemia, and vascular diseases may contribute to the development of peripheral neuropathies due to metabolic and/or vascular damage, thereby accelerating the deterioration of vibratory sensitivity. In our study, the presence of metabolic and/or vascular disorders was associated with a 113.8% worsening of the vibrotactile perception threshold (VPT) in digit II and up to a 188.6% increase in digit V (Table 3 and Table 4). These alterations in VPT were greater than those observed in the population with higher A(8) exposure, suggesting a significant role of such comorbidities in the deterioration of the neurosensory component of HAVS, independent of occupational exposure to HTV.

In the present study, a high BMI was found to be positively associated with deterioration in VPT at 31.5 Hz in digit II and at 31.5 and 125 Hz in digit V (see Table 3 and Table 4), with a respective percentage VPT increase of 16.2% and 35.0%. Excess body weight may negatively influence vibrotactile perception due to mechanical insulation of cutaneous receptors; moreover, obesity is frequently associated in the literature with reduced peripheral perfusion [43,44], which may in turn impair the sensory function of mechanoreceptors. However, findings in the literature appear to be contradictory: one study [45] found no effect of BMI on the results of neurosensory testing, while another [46] reported that a high body mass index, in combination with around 10 years of occupational exposure to hand-transmitted vibrations (HTVs), increased the perception of neurosensory symptoms in the fingers resembling those seen in hand–arm vibration syndrome (HAVS). As previously observed, the results of our study highlight a positive correlation between altered vibrotactile perception thresholds and metabolic and/or vascular diseases. This finding is supported by several studies that have examined the association between VPT and diabetic neuropathy affecting large-diameter nerve fibres [47]. The disease often begins with damage to small nerve fibres, leading to altered sensory processing of temperature and pain, as well as potential impairment of sweating and local blood flow in peripheral tissues. Subsequently, large-diameter nerve fibres may become involved, serving as a possible early indicator of the risk of foot ulceration [48]. Significant deterioration in nerve conduction velocity and increased sensory thresholds across various modalities have been reported in diabetes mellitus [49]. A reduction in motor nerve conduction velocity has also been found in individuals with glucose intolerance, even in the absence of diagnosed diabetes mellitus [50], possibly due to the pro-oxidative effects of chronic hyperglycaemia on nerve cell structures [51].

In our study, no significant associations were observed between alcohol consumption and VPT thresholds; only one statistically significant association was found at 125 Hz in digit V corresponding to a 35.0% increase in VPT (Table 4). A similarly non-significant influence of alcohol consumption on vibrotactile threshold determination was reported in other studies [49]. Some national and international guidelines provide specific thresholds for alcohol consumption in the general population. For example, the World Health Organization (WHO) defines high-risk drinking in men as the consumption of more than 60 g of pure ethanol per day. In our study population, 48.1% (N = 113) reported current alcohol consumption: among them, 92.0% (N = 104) consumed fewer than six alcohol units per day, and 64.0% (N = 73) reported consuming up to two alcohol units per day. Only 8.0% of subjects (N = 9) reported consuming from more than six alcohol units per day. Given this limited variability, a broader range of alcohol exposure would be necessary to more accurately assess the potential association between alcohol consumption and deterioration of the neurosensory component of HAVS in populations exposed to HTV.

In summary, the results of this study suggest that various individual risk factors, such as cigarette smoking, age, and the presence of comorbidities, may independently contribute to the onset of neurosensory symptoms, regardless of the well-known biological effect of occupational exposure to hand–arm vibrations.

4.2. Thermotactile Perception Thresholds

In contrast to the findings related to VPTs, the models concerning TPTs (both warm and cold) yielded less consistent results: neither the occupational risk factors nor the personal risk factors considered (lifestyle habits and pathological conditions) showed significant associations with the deterioration of thermotactile thresholds, with the exception of the effect of smoking habits on warm (W-TPT) and cold (C-TPT) thresholds—although this was limited to the median nerve (digit II) (Table 5 and Table 6).

A more detailed analysis of the data showed that ageing produced statistically significant associations only for C-TPT thresholds in digit V (Table 6). Studies on the effects of age on thermotactile perception thresholds have produced contradictory results [7,52,53]. In general, as age increases, there is a worsening in the perception of both warmth and cold, although this decline is less marked than that observed for vibrotactile thresholds [54,55]. This appears to be related to a lower susceptibility of unmyelinated and small myelinated fibres to various alteration mechanisms, including mechanical ones, compared to other types of nerve fibres [54,56].

The weak correlation between metabolic disorders and altered TPTs observed in our study does not align with data from the literature, which reports greater degeneration of small-diameter fibres compared to large-diameter fibres. This finding has also been confirmed histologically, where two subpopulations, classified as neurographically normal and abnormal, respectively, differed in both intraepidermal nerve fibre density and QST tests [57,58]. In the presence of a combination of neuropathy affecting both small and large diameter fibres, the process involved small-diameter fibres more extensively [59]. It is well known that diabetic peripheral neuropathy is distinguished into large-fibre neuropathy and small (or fine)-fibre neuropathy. It has previously been shown that damage to unmyelinated small nerve fibres (manifesting as abnormalities in pain and cold perception) precedes that of large-diameter fibres and serves as an early indicator of peripheral neuropathy [60].

It should be noted that our results showed an unexpected inverse association between altered TPT thresholds and previous trauma or surgical procedures involving the upper limbs, with an improvement in the W-TPT threshold in digit V (Table 6). However, it is important to consider that the methods used to measure VPT and TPT thresholds are not comparable. The former is a validated and recognised tool for assessing neurosensory function in individuals exposed to hand-transmitted vibration (HTV), providing useful data for both the diagnosis and ongoing monitoring of vibration-induced neuropathy. A technical standard is available which outlines in detail the methods for measuring VPT thresholds in individuals exposed to HTV [12,13]. The ISO standards offer detailed recommendations to ensure the consistency, reliability, and reproducibility of VPT testing.

Given that conventional electroneurophysiological methods are not suitable for investigating the response of small-calibre nerve fibres, and that the assessment of thermotactile thresholds is considered a valid alternative method for evaluating both the anatomical integrity and functional capacity of the small fibres responsible for thermal perception [33,61], it is worth noting that no international standard currently exists for the measurement and assessment of thermotactile thresholds. Moreover, although this does not directly limit the internal validity of our findings, the limited availability of normative reference values in the general population [33,62] may affect the comparability of our results and may have contributed to the weak and sometimes inconsistent association estimates observed, which were not always in line with expectations. Indeed, it is reasonable to expect that trauma or surgical interventions to the upper limbs could exacerbate or contribute to the development of the neurosensory component of HAVS through pathophysiological mechanisms such as nerve damage or sensitisation, microvascular impairment, inflammatory processes, and structural neuroanatomical changes [63,64].

4.3. Limitations

Our study also has some limitations. It is a cross-sectional study in which data were collected from individuals at a single point in time over a ten-year period. Moreover, the consecutive recruitment of participants may introduce selection bias, as the sample may not fully represent the target population. Additionally, when participants are enrolled consecutively, reliance on self-reported information (such as lifestyle factors and personal habits) can introduce bias due to inaccurate or incomplete recall (recall bias). In our study, we used information on medical history and exposure to personal risk factors that may change over time. Furthermore, given that individual exposure in our study population was estimated rather than measured and assessed directly in the field, the calculated A(8) values may have been imprecisely quantified. In this context, the use of data derived from the literature or industrial hygiene databases may not accurately reflect the actual individual exposure to hand–arm vibration within the study population, which could have led to exposure misclassification and, consequently, imprecise effect estimates. However, it should be noted that the results of our study showed good consistency between daily vibration exposure (A(8)) values and deterioration of VPT thresholds, revealing a 32.3–66.5% worsening of thresholds detected in digit II and a 28.5–56.8% worsening detected in digit II among those with higher exposure, thus supporting our individual A(8) exposure estimates.

Finally, it is important to note that data for the present study were collected between 1995 and 2005, using protocols aligned with the then-current ISO 13091-1 standard [12,13]. However, a more recent update, ISO 13091-2:2021 [65], includes expanded guidelines for vibrotactile testing, including specifications on probe size, contact force, calibration methods, and interpretation of threshold variability. Although these refinements could not be applied retrospectively to our dataset, they represent a valuable advancement in the standardization of sensory testing and future studies should adopt these updated protocols to enhance methodological consistency and improve the clinical and occupational relevance of VPT assessments.

4.4. Strengths

Our study also has several strengths. Both vibrotactile and thermotactile perception thresholds were examined to assess the role of personal risk factors in the deterioration of the neurosensory component of HAVS. To this end, we considered both hands and separately analysed the VPT and TPT thresholds for the second and fifth digits, as different pathophysiological pathways and varying neurosusceptibility may be involved in the peripheral neurosensory damage of the median (digit II) and ulnar (digit V) nerves. In our study, we accounted for multiple covariates to evaluate the role of personal risk factors in the deterioration of VPT and TPT values. Furthermore, with particular reference to vibrotactile thresholds, the measurement method employed aims to assess neurosensory function in order to detect the early signs of vibration-induced neuropathy. The methods used to measure VPTs are validated and represent a recognised tool for assessing nerve fibre function in individuals exposed to HTV, providing useful data for both diagnosis and longitudinal monitoring of the progression of vibration-induced neurosensory disorders.

5. Conclusions

Despite the limitations associated with the cross-sectional design of the study, our results provide important insights into the relative contribution of individual risk factors to neurosensory deterioration in workers exposed to hand-transmitted vibration. Although many of these risk factors are known to contribute to sensory impairment, our analysis quantifies their specific impact on the neurosensory component of HAVS, particularly in relation to vibrotactile perception thresholds (VPTs) in the context of professional exposure to HTV. The observed effect sizes suggest that some personal risk factors—such as smoking, overweight, and vascular or metabolic comorbidities—may exacerbate neurosensory dysfunction in a magnitude comparable to, or even exceeding, vibration exposure. These findings support the potential value of integrating comorbidity screening into occupational health assessments, with the aim of identifying high-risk individuals who may benefit from earlier intervention or more intensive monitoring. However, further studies with larger study populations are needed to confirm these findings.