Heat Stress and Cardiac Strain in French Vineyard Workers

Abstract

Agricultural workers often produce considerable excess heat due to the physically demanding nature of their activities, increasing their risk of thermal stress in even moderately warm conditions. Few studies have examined the physiological responses to heat load in agriculture. We aimed to assess the heat strain experienced by vineyard workers during canopy management in dry field conditions, and to disentangle the effects of the heat produced by the body and the thermal environment. Thirty workers from five Bordeaux vineyards of southern France were monitored during vine-lifting and trellising (June 2012). The mean heart rate, net cardiac cost, relative cardiac cost, and cardiac workload score were assessed during field activity. As the workers were nested within vineyards, multilevel linear regression models were used for correct inference. Skin temperature increased by an average of 1.0°C. Cardiac indices showed marked differences between individuals. The workload was evaluated as ‘heavy’ or ‘very heavy’ for more than one-third of the workers, of whom one experienced heat exhaustion. Above some individual characteristics, we highlighted a contextual effect (air temperature) for the mean heart rate (P = 0.03), the relative cardiac cost (P = 0.01) and, to a lesser extent, a cardiac workload score (P = 0.07). Canopy management by hand in vineyards causes considerable cardiac and thermoregulatory strain. Appropriate instruments should be developed to simultaneously evaluate work intensity, work quality, and productivity at the vineyard level to raise the awareness of both managers and employees about taking preventive measures.

agriculture, cardiac strain, heat stress, vineyard workers

What’s important about this paper

Agricultural workers can produce excess heat because of the physical demands of their work, putting them at risk of thermal stress even in moderately warm conditions. Among vineyard workers, cardiac strain during canopy management tasks varied widely between individuals, with more than one-third of workers studied experiencing heavy or very heavy workload. Dry-bulb temperature and task duration were associated with cardiac strain. Heat stress can be prevented among vineyard and other agricultural workers.

Introduction

Heat exposure is common among workers in the agriculture sector because most tasks are performed outdoors, where the climate cannot be controlled. Because of the physically demanding nature of their activities, agricultural workers often produce considerable excess heat, increasing their risk of thermal stress in even moderately warm conditions. Moreover, many activities in this sector (especially pesticide spraying) require personal protective equipment (PPE), reducing the body’s ability to dissipate heat by hampering the exchange of heat with the environment (Staal Wästerlund, 2018).

Changes in the body’s core temperature can alter the absorption, distribution, metabolism, and excretion of chemicals. Heat stress, with or without exercise, activates thermo-effectors (e.g. skin blood flow, sweating, respiration), which, in turn, lead to more efficient transcutaneous absorption of pesticides in humans (Gordon and Leon, 2005). Increases in respiration can lead to further toxicant exposure through inhalation, as high temperatures accelerate dispersion and increase the concentration of airborne particles (Gordon, 2003; Leon, 2008).

Many heat-stress indices, such as the wet-bulb globe temperature (WBGT) index and predicted heat strain (PHS), have been developed to establish work conditions that reduce the risk of heat-related illnesses. However, the WBGT index and the PHS method were both developed to determine the risk of heat stress in places where people work for extended periods in normal work clothing (Parsons, 2013). These indices cannot, therefore, be used in work situations in which the person is required to wear protective clothing, the thermal conditions rapidly change, or people who are not physically fit perform the work (Staal Wästerlund, 2018). Under such circumstances, physiological measurements (such as mean skin temperature and/or heart rate [HR]) (ISO9886, 2004) provide very good insight into the extent of heat strain experienced by a subject (NIOSH, 2016).

Although ergonomic and physiological studies have long investigated the risks associated with activities conducted in hot environments (Parsons, 2000; Crandall and González-Alonso, 2010), few studies have examined the physiological responses to the heat load in agriculture (Costa et al., 1989; Nigg et al., 1992; Singh et al., 2011; Sahu et al., 2013; Mitchell et al., 2017, 2018; Wagoner et al., 2020). Even less is known about the physiological cost in vineyard farming activities. We recently compared the physiological burden of vineyard workers wearing three different types of PPE in humid field conditions (Grimbuhler and Viel, 2018). Here, we took advantage of this experience to further assess heat strain experienced by vineyard workers during canopy management in dry field conditions, and to disentangle the effects of the heat produced by the body and the thermal environment.

Methods

Study area and population

The study population has been fully described elsewhere (Grimbuhler and Viel, 2018). Briefly, the study took place during the 2012 pesticide treatment season (19 June to 27 June) in seven Bordeaux vineyards of southern France. A total of 42 workers (six per vineyard) accepted to participate and provided written informed consent. Canopy management consisted of lifting (raising the fruit-bearing shoots from the ground and attaching them vertically to wires running above the vine) and trellising (separating the shoots from each other and stapling them to overhead wires). Trellis height ranged from 1.50 to 2.10 m, depending on the vineyard.

Each worker’s practice was observed throughout his/her task by a field staff member. Sociodemographic characteristics, potential determinants of physiological strain, and task-related factors were collected using a structured questionnaire. Subjects also provided subjective evaluations.

Participants and environmental conditions

Vineyard workers were paid by salary rather than by the line of work completed in the field. On arrival at the vineyard, pairs of workers were assigned a number of rows by the supervisor, depending on the weather conditions, the trellising system, the vine vigor, etc. Partners worked each side of the vines, keeping pace with each other, one slightly ahead to make trellising easier, and reduce the risk of injury. At the end of the row, the quickest pairs looked down the rows and helped the slowest pairs get the job done. This group cohesiveness fostered homogeneous work speed. Depending on how long the rows were, workers usually took a one/two-minute break for hydration before moving on to another row.

Participants were first observed working during humid conditions (corresponding to dewy or damp conditions or even a short episode of rainfall). After this re-entry task, which lasted between 71 and 120 min, depending on the vineyard, they took a snack break or had lunch. Upon return from the break, the researchers and workers decided consensually whether the environmental conditions had changed from humid to dry, based on canopy appearance and field relative humidity (<60%). As a result, a dry context was acknowledged for five of the seven vineyards, leaving 30 of the 42 initial workers for consideration in the present study (12 workers from the other 2 vineyards being removed because of on-going humid conditions).

According to the guidance document from the European Food Safety Authority (EFSA) on the assessment of operator, worker, resident, and bystander exposure, PPE is required for re-entry tasks (such as canopy management) (EFSA, 2014). So, the 30 remaining workers donned a Costal aluminum garment (Brisa®), composed of a shirt and long pants (with aluminized leggings) made of poly-cotton blend (35%/65%). Additional gear items were a cotton cap, Mapa Ultrane 553 protective gloves (composed of polyamide and seamless textile liner with palm and fingers covered with nitrile), and rubber boots. For undergarments, all subjects wore identical disposable two-piece long johns (100% cotton). The work was self-paced, allowing workers to adapt their workload to the body’s production of heat and its capacity to exchange heat with the surroundings.

The outdoor dry-bulb temperature ranged from 26.3 to 31.2°C and relative humidity ranged from 39 to 54% across field sites (Table 1). In two estates (n°2 and 3), changes in wind speed (up to 5.9 and 4.7 m/s, respectively) were observed, which are typical during this season.

Table 1.

General conditions of the vine-lifting task in dry conditions (Bordeaux vineyards, southern France, June 2012) (extended from Grimbuhler and Viel, 2018).

	Estate 1	Estate 2	Estate 3	Estate 4	Estate 5
Vine-growing area	Entre-deux-Mers	Entre-deux-Mers	Medoc	Graves	Blayais
Date	June 20	June 21	June 22	June 25	June 27
Starting time (hh:mm)	13:40	14:20	13:53	14:28	10:30
Task duration (mn)	46	142	88	38	100
Dry-bulb temperature (°C)	31.2	26.3	28.2	30.4	30.5
Relative humidity (%)	52	48	39	52	54
Wind speed (m/s)	1.3–3.0	1.2–5.9	1.5–4.7	0.0–0.9	0.6–1.8

Assessment of physiological cost

Skin temperature was measured with a Proges Plus® 22L thermo-button (with an accuracy of ±0.1°C) placed against the skin of the tibia (just above the ankle) of each worker. Measurements were recorded once per minute throughout the task. The skin temperature was assessed at rest before starting vine-lifting and the mean skin temperature was calculated during work.

All workers were fitted with a Polar® RS800 HR monitor before putting on their poly-cotton garment. It included an adjustable chest strap equipped with a heartbeat detector and a wrist receiver that displayed and recorded the HR transmitted at 5-s intervals. Resting HR (HR_rest) was estimated through the first percentile value of the HR recording (i.e. when vineyard workers resumed work after their snack or lunch break) (Malchaire et al., 1986). The net cardiac cost (NCC) represented the difference between the mean HR (HR_mean) and HR_rest. The theoretical maximum HR (TMHR) was defined by the Gellish formula (207 − 0.7 × age) (Gellish et al., 2007). Heart rate reserve (HRR) was the difference between the TMHR and HR_rest. The relative cardiac cost (RCC) expressed the NCC as a percentage of the HRR (NCC/HRR). A cardiac workload score was calculated according to Meunier et al. (1994). This summed composite index is based on the HR_mean, RCC, and 99th percentile of the HR (HR99) and ranges from 1 to 15. This quantitative score was then transformed into a five-level classification of work intensity (1–3: light; 4–6: moderate; 7–9: rather heavy; 10–12: heavy; 13–15: very heavy) (Meunier et al., 1994).

Statistical analyses

The following data were considered at the worker-level: age (years), gender (male/female), body mass index (BMI, kg/m²), experience in vine-lifting (years), handedness (left-handed, right-handed), skin temperature at rest (°C), mean skin temperature during work (°C), and cardiac strain indices (mean HR, bpm; NCC, bpm; RCC, %; cardiac workload score, unitless). Because one worker lost her thermos-button, the missing temperature values were replaced by the modal values from remaining participants. Additional information was collected at the vineyard-level: task duration, air temperature, and relative humidity.

When considering heat-stress, two dimensions must be taken into consideration: the amount of heat produced (depending on the individual physiological characteristics and workload) and the thermal environment of the worker (which determines the possibilities for heat dissipation). In an attempt to disentangle the effects of these two dimensions, we drew upon the hierarchical structure of the data for analysis as the workers were nested within vineyards. Associations between various physiological indices (dependent variables) and various risk factors (defined at either the worker or vineyard-level) were examined using multilevel linear regression models. After careful review of the literature, four individual-level covariates (age, gender, BMI, and experience in vine-lifting) were forced into the models, as well as the task duration and the air temperature (vineyard-level covariates). The remaining covariates were introduced into the models provided they were associated with the physiological index under study (P < 0.20). All analyses were performed using MLwiN software version 2.24 (Centre for Multilevel Modelling, Bristol University, UK).

Results

Description of the population

Ages ranged from 21 to 59 years with a median of 39 years. Most of the workers involved in the study were female (19 of 30 workers) (Table 2). According to their BMI, 33% of the workers could be considered overweight and 10% were obese. Experience in vineyard lifting ranged from zero to more than 40 years. Most workers were right-hand dominant (90%). No worker reported current use of medications that could have contributed to the early onset of heat stress.

Table 2.

Characteristics of vineyard workers (n = 30, number [percentage], Bordeaux vineyards, southern France, June 2012) (extended from Grimbuhler and Viel, 2018).

Age (years)
<30	6 (0.20)
30–39	10 (0.33)
40–49	5 (0.17)
≥50	9 (0.30)
Gender
Female	19 (0.63)
Male	11 (0.37)
Body mass index (kg/m  2)
≤25	17 (0.57)
25–29.9	10 (0.33)
≥30	3 (0.10)
Experience in the vine-lifting task (years)
<5	5 (0.17)
5–9	8 (0.26)
10–14	7 (0.23)
15–19	2 (0.07)
20–24	3 (0.10)
≥25	5 (0.17)
Handedness
Left-handed	3 (10.0)
Right-handed	27 (90.0)

Physiological costs and work intensity

Physiological characteristics are reported in Table 3. During canopy management, skin temperature increased by an average of 1.0°C (from 33.9 to 34.9°C). In terms of the average cardiac indices, the mean HR was 113.5 beats per mn (bpm), the mean NCC was 24.1 bpm, the mean RCC was 23.1%, and the mean cardiac workload score was 8.8.

Table 3.

Physiological strain of workers (n = 30, Bordeaux vineyards, Southern France, June 2012).

	Mean	Standard deviation	Minimum	Maximum
Skin temperature at rest (°C)	33.9	1.1	31.7	35.8
Mean skin temperature during work (°C)	34.9	1.0	32.34	36.1
Mean heart rate (bpm)a	113.5	17.3	78.7	165.3
Net cardiac cost (bpm)a	24.1	13.3	7.1	62.7
Relative cardiac cost (%)a	23.1	13.6	10.1	87.2
Cardiac workload score (unitless)a	8.8	3.3	2.0	15.0

^a Cardiovascular parameters were missing for two workers because of dysfunction of the cardiometer.

When transforming the latter into the five-level classification, the workload associated with canopy management was evaluated as ‘heavy’ or ‘very heavy’ for more than one-third of the workers (10 of 28). Among the five individuals facing a very heavy workload, a woman (aged 50, TMHR of 172 bpm, NCC of 63 bpm, RCC of 87%) experienced heat exhaustion. Alerted by apparent weakness and possible giddiness, the field monitor checked her cardiac frequency displayed on the wrist-unit (195 bpm) and immediately interrupted her job (17 min before the other 5 workers of Estate 5). The worker was then moved to a cool environment for rest and provided with fluids to drink.

Risk factors for physiological costs

Risk factors for physiological strain of workers appear in Table 4. Independent variables were identical whatever the physiological index considered, as no remaining covariate was associated at P < 0.20. Regarding the worker-level variables, age was not significant for any outcome. Male gender was associated with the mean skin temperature at rest (P < 10^–7), as well as BMI (P = 0.02). Experience in vine-lifting was significantly linked to the mean skin temperature during work (P = 0.01), and near-significantly with the NCC (P = 0.08). Regarding the vineyard-level variables, the task duration was significantly associated with the RCC (P = 0.02). Finally, a significant link was found between the air temperature and both the mean HR (P = 0.03) and the RCC (P = 0.01), while the link with the cardiac workload score was of borderline significance (P = 0.07).

Table 4.

Risk factors for physiological strain of workers (n = 30, multilevel regression model, beta coefficient [standard deviation], Bordeaux vineyards, Southern France, June 2012).

	Skin temperature at rest (°C)	Mean skin temperature during work (°C)	Mean heart rate (bpm)	Net cardiac cost (bpm)	Relative cardiac cost (%)	Cardiac workload score (unitless)
Worker-level variables
Age (years)	−0.006 (−0.047,0.035)	−0.018 (−0.047,0.011)	0.142 (−0.538,0.822)	0.296 (−0.212,0.804)	0.404 (−0.106,0.914)	0.030 (−0.099,0.159)
Gender (male)	0.250 (−0.514, 1.014)	1.447**(0.922,1.972)	−0.605 (−12.884,11.674)	−4.344 (−13.511,4.823)	3.473 (−5.851,12.797)	−0.570 (−2.897,1.757)
Body mass index (kg/m2)	0.114**(0.020, 0.208)	0.035 (−0.030,0.100)	−0.162 (−1.736,1.412)	−0.982 (−2.156,0.192)	0.488 (−0.708,1.684)	−0.136 (−0.434,0.162)
Experience in vine-lifting (years)	0.000 (−0.043,0.043)	0.038**(0.009,0.067)	0.088 (−0.616,0.792)	−0.476* (−1.001,0.049)	−0.332 (−0.865,0.201)	0.010 (−0.123,0.143)
Vineyard-level variables
Task duration (mn)	−0.010 (−0.026,0.006)	−0.006 (−0.016,0.004)	0.097 (−0.148,0.342)	0.118 (−0.064,0.300)	0.219**(0.033, 0.405)	0.016 (−0.031,0.063)
Dry-bulb temperature (°C)	−0.132 (−0.359,0.095)	0.057 (−0.100,0.214)	4.084**(0.442, 7.726)	2.040 (−0.679,4.759)	3.652**(0.888, 6.416)	0.644* (−0.046,1.334)

^a Cardiovascular parameters were missing for two workers because of dysfunction of the cardiometer.

*P < 0.10.

^**  P < 0.05.

Discussion

This study shows that the cardiac strain recorded during canopy management varied greatly between individuals. One-third of vineyard workers were heat-stressed to nearly critical physiological levels (heavy or very heavy workload), with one worker experiencing heat exhaustion. Above some individual characteristics, we highlighted a contextual effect (air temperature) for three cardiac indices (mean HR, RCC, and to a lesser extent cardiac workload score).

This study had three main strengths. First, the impact of heat on people performing routine agricultural activities has been little studied. Work conditions and practices were representative of the conditions typically encountered during hand canopy management in vineyards. Second, temperature and HR monitoring provided a more objective measurement of the physiological reactions and intensity of the effort of vineyard workers than the use of perceived discomfort (such as the Borg scale of perceived exertion), which is generally determined by a combination of physiological, psychological, and physical factors. Third, we used a sound statistical methodology, as the potential correlation of the data within vineyards was accounted for by the two-level regression models for correct inference.

Our study also had a few limitations. First, the sample size (30 workers) was modest, although the statistical power was sufficient to highlight a contextual effect (air temperature) above some individual characteristics. Second, we were well aware of the ISO 9886 4-point method used in hot conditions for estimating mean skin temperature (ISO9886, 2004). However, only one thermo-button was placed against the skin of the tibia (one of the four recommended sites together with neck, right scapula, and left hand) to ensure worker acceptance and avoid interfering with the demanding canopy management task. Although a distinction has to be made between local skin temperature measured at a specific point on the body surface and mean skin temperature measured over the entire body surface, comparisons across workers were appropriate as skin temperature was measured at the same anatomical site. Third, environmental heat was only measured as dry bulb temperature and not WBGT (as no black bloc thermometer was available in the field), although the WBGT is used in many heat-stress alert limits to protect most healthy workers. However, the Outdoor Heat Exposure Rule from the Washington State Department of Labor and Industries can be a useful standard for our study (NIOSH, 2016). It is applicable to employees performing work in an outdoor environment during summertime. For those wearing double-layer woven clothes (e.g. coveralls, jackets, and sweatshirts) it applies to temperatures at or above 25°C. According to this alert limit, our study clearly took place under heat-stress conditions. However, the absence of a black bloc sensor precluded the measurement of the mean radiant temperature although solar radiant load is known to contribute to heat stress. Forth, the hydration status could unfortunately not be measured. We did not monitor the consumption of fluids because our main concern was to minimize the effect of our study on the regular work, most workers reported drinking before coming to work, and task duration (increasing fluid intake) varied between vineyards. Moreover, for cultural and practical reasons, managers were reluctant to facilitate urine collection (to measure specific gravity or osmolality) or weight monitoring (before and after the shift).

The difficulty of accurately determining which workers are most at risk of excessive occupational heat exposure is that heat tolerance varies broadly between individuals and even within an individual on a day-to-day basis. This is because environmental conditions, intensity of the activity, the worker’s behavior, and individual physiological factors can shift and modulate the risk of occupational heat exposure (Schlader et al., 2011; Staal Wasterlund, 2018). For example, workers may adjust their work pace, such as taking short rests between rows. Cardiac strain can also vary according to other personal criteria, such as physical fitness or physiological acclimatization to the work conditions.

In this study, average cardiac estimates were not far below or just above the individual recommended limits (110 bpm, 30 bpm, and 30%, for mean HR, NCC, and RCC, respectively) (Saha et al., 2008), revealing environmental heat stress and the high physiological demand of the job. It was expected that the vineyard-level variables had no influence on the mean skin temperature at rest (as the vine-lifting had not started at that time), suggesting internal validity. The experience in vine-lifting was negatively linked to the NCC. Being related to the resting HR, this index is considered to more specifically evaluate the workload strictly connected to the job (Costa et al., 1989). This negative association of borderline significance suggests that the more experienced workers adapted their performance strategies. Conversely, the RCC, which takes into consideration the maximal HR of the subjects, provides an evaluation more related to individual strain (Costa et al., 1989). This is in line with our results, showing that both heat stress and length of task deeply influenced this cardiac strain.

Unfortunately, our results cannot be easily compared to those of other studies because very few (if any) field studies in the scientific literature have monitored the cardiac burden of farming activities under identical conditions.

Conclusion

Our results show that canopy management by hand in vineyards causes considerable cardiac and thermoregulatory strain. Heat stress management strategies are well known (training about heat illness prevention, implementation of appropriate work-rest schedules, provision of shade at the outdoor setting, availability of fluids for hydration, etc.) (Jackson and Rosenberg, 2010). However, it appears necessary to develop appropriate instruments for simultaneously evaluating work intensity, work quality, and productivity at the vineyard-level, which could provide an integrative view and raise the awareness of the necessity to take preventive measures among both managers and employees.

Acknowledgments

We are grateful to the vineyard workers, vineyard managers, and field monitors who made this study possible.

Funding

The Union des Industries de la Protection des Plantes (French Crop Protection Industry Association) funded the National Research Institute of Science and Technology for Environment and Agriculture for the field study within the framework of the Safe Use Initiative project. The funder had no role in the design or conduct of this study, the analysis or interpretation of the data, or the preparation of this manuscript.

Conflict of interest

The authors have no conflicts of interest or financial relationships to disclose.

Ethical approval

Subjects included in the study volunteered, gave their written informed consent, and were observed in the course of their normal work activities. Therefore, no approval from an ethics committee was required by French regulations at the time of the study.