Overweight/obesity and physiological variables in workers exposed to rotating shift in a food production factory

Introduction Shift work has been hypothesized as a potential risk factor for overweight/obesity or other metabolic changes. We examined the relationship between work shift and body mass index, waist-hip ratio, lipid profile, and glucose concentration in workers from a food manufacturing factory in Colombia. Objectives To investigate the association between shift work and changes in physiological variables in food manufacturing industry workers in Medellín, Colombia. Methods This cross-sectional study was conducted with 763 employees from a food manufacturing factory. Information was collected from the medical records from the occupational health provider institution in charge of workers’ periodic follow-up. Results The study sample consisted of 637 (83.5%) men and 126 (16.5%) women. Mean age was 43.35 ± 9.8 years, and mean body mass index was 25.49 ± 3.23 kg/m2. After adjusting for potential confounders, logistic multivariate regression revealed a statistically significant association between shift work and higher body mass index and higher total cholesterol levels compared with dayshift (p < 0.05). Finally, the analysis of waist-hip ratio for each shift scheme and sex showed that this ratio was higher for rotating shift workers, with a significant difference for women. Conclusions Significant associations were observed between shift work and overweight/obesity and hypercholesterolemia. However, these findings should be confirmed by longitudinal studies.


INTRODUCTION
Shift work, either in fixed or rotating schemes, is an important condition for millions of workers in Colombia and worldwide.Due to the growing social demand for 24-hour provision of goods and services, companies have been required to implement strategies to maximize productivity, including rotating shift work.Consequently, the percentage of workers in this scheme has increased over the last decades, being estimated at 20% of the worldwide population 1 and almost 29% in Colombia, according to the 2007 first national survey on health and working conductions. 2n extensive literature published in the last decade suggest a possible association between shift work and changes in both workers' physiological and psychological parameters.These effects are based on two crucial variations in their vital conditions: changes in the circadian cycle and sleep deprivation. 3][11][12] Night workers experience a constant conflict between the pattern governed by the internal circadian cycle and the pattern imposed by the external biological rhythm, which generates an imbalance in normal body functions.The issue is even more complex in the case of rotating shift workers, because not only they are subjected to an inadequate circadian rhythm, but also the variability of their work scheme makes any adaptive attempt difficult. 4ne of the determining factors for changes in the physiological parameters of night or rotating shift workers is lack of counterregulatory activity in cortisol secretion.Glucocorticoid secretion is under strong influence of the suprachiasmatic nucleus circadian clock, which functions as a modulator of sensitivity of soft tissues to cortisol.In normal conditions, cortisol levels reach their peak in the early morning and their nadir at midnight, in order to adjust body activities and functions to the external biological rhythm.Night or rotating shift workers develop functional hypercortisolism, with high cortisol levels in the afternoon and consequent complications resulting from metabolic syndrome.Glucocorticoids are required for the functioning of all organs and tissues, especially in respiratory, cardiovascular, immune, and musculoskeletal function; therefore, dysregulation in their levels affects multiple body system and changes several physiological parameters. 12he aim of the present study was to analyze the influence of fixed and rotating work schedule on blood pressure, body mass index (BMI), abdominal circumference, waist-hip ratio, lipid profile, and glucose levels.For this purpose, differences in the prevalence of overweight/obesity and/or variations of physiological parameters were compared between rotating shift workers and those with a fixed work schedule, as well as the association of these variables with other demographical and labor factors in food industry manufacture workers from the metropolitan region of Medellín.

PARTICIPANTS
This cross-sectional study was conducted at a food industry factory in Medellín, Colombia.
Based on a formula for comparing proportions in independent groups and on the results of a report by Nakamura et al., 13 who found a percentage of obesity of 27.7% among shift workers, and of 18.8% among daytime workers, with a 95% confidence interval and a power of 80%, the recommended sample size was 357 workers.However, there was available information on 947 workers; thus, all periodic occupational health surveillance records available were assessed.A total of 947 medical records were selected from the database of an Occupational Health Services Provider Institution (Institución Prestador de Servicios de Salud, IPS) with headquarters in Medellín, Colombia, which is responsible for the periodical follow-up of the health status of workers from the factory.Individuals working in the factory for less than one year were excluded, in order to avoid the period of circadian adaptation (n = 150).Workers diagnosed with arterial hypertension, heart failure, cardio or cerebrovascular atherosclerotic disease, diabetes mellitus, or other metabolic or endocrinologic diseases before start working in the factory (n = 28) were excluded.The study also excluded workers whose more than 10% of data of interest were missed (n = 10).The custody of information was maintained through a letter of commitment and a confidentiality agreement, signed both by the researchers and the IPS.Furthermore, the present study is in compliance with Law 1,266 of 2008, which laid down the general provisions for habeas data and regulates the handling of information contained in personal databases.

MEASUREMENTS
Information on demographics (age, gender, marital status, educational level, socioeconomic stratum), as well as on habits and lifestyle (smoking, alcoholism, and hours of aerobic exercise) were obtained through the occupational medical records used by the physicians from the IPS.BMI was calculated as body weight (kg), divided by height squared (m 2 ).The parameters established by the World Health Organization (WHO) were applied to define the BMI classification groups. 14Waist circumference was measured at the midpoint between the superior border of the iliac crest and the inferior margin of the rib.WHO parameters were also applied in this case (central obesity when circumference is >88 cm in women and >102 cm in men). 15Hip circumference was measured at the largest circumference of buttocks.Likewise, WHO parameters were used to determine the risk threshold for cardiovascular disease (value >0.95 for men and >0.85 for women). 15BMI, waist circumference, and waist-hip ratio were measured by a physician during physical examination.

STATISTICAL ANALYSIS
Participants' characteristics were summarized for day shift workers and rotating shift workers.
Variables were expressed as mean ± standard deviation or median ± interquartile range, according to data distribution, and categorical variables as percentages.The Chi-square test and the independent t test, or the Mann-Whitney test, were used to compare means or medians between groups for normally and asymmetrically distributed data, respectively.

RESULTS
This study analyzed information on 763 workers with daytime and rotating work schedules, who were predominantly male, with a mean age of 43.35 ± 9.8 years.There were significant differences between day shift workers and rotating shift workers in terms of sex and educational level.With regard to alcohol consumption, smoking, and physical activity, no statistically significant differences were observed between the groups (p > 0.05) (Table 1).
An analysis of physiological measurements from workers participating in the study according to their work schedule showed that the BMI of rotating shift workers is higher than that of day shift workers, representing a statistically significant difference.As for the lipid profile, higher levels of total cholesterol were found in participants working in rotating shifts compared to those working in day shifts (p < 0.05).There was no significant difference in the other blood lipid measurements between the groups (Table 2).
An analysis of the waist-hip ratio for each shift schedule and according to participant's sex revealed that this ratio was higher among rotating shift workers compared to day shift workers, with a significant difference for female workers (Table 3).

DISCUSSION
This study aimed to analyze the association of rotating and day work schemes with BMI, waist-hip ratio, lipid and glucose profile in a group of workers from a food industry factory located in Medellín, Colombia.In the present study, the overall prevalence of workers with overweight or obesity was 53.1%, being thus lower than the overall prevalence of overweight or obesity in adults from 18 to 64 years of age, which corresponds to 56.4% of the Colombian population, according to the 2015 Colombian National Nutrition Survey. 16hese results may be explained by the healthy workers bias, 17 meaning that the working population in contact with risk factor has lower morbidity and mortality rates compared with the general population, since workers are selected based on their employability status, which depends mostly on their health state before entering the labor market.Another factor to consider is that 83.5% of the sample consisted of men, who, as reported in 2015 Colombian National Nutrition Survey, have a lower prevalence of overweight or obesity compared to women. 16n the present study, the prevalence of overweight or obesity among rotating shift workers was 48.2%, a result close to that found among rotating shift workers from the industrial canteens of Sercoinfal CA, a company in the state Lara, Venezuela, which was 48.3%. 18It was also consistent with findings by Macagnan et al. 19 showing a prevalence of overweight of 42.2% among nightshift workers in a production line in a poultry processing.1][22][23] The mechanism that associates shift work with higher BMI has not been completely understood yet, but different studies mention changes in circadian rhythm and sleep deprivation as triggering factors.Sleep deprivation causes a decrease in leptin levels and an increase in ghrelin levels, which seems to coincide with increased appetite and weight gain. 24n the sample of this study, the BMI of rotating shift workers (25.58 ± 3.1 kg/m 2 and 24.8 ± 3 kg/m 2 , respectively) was higher than that of dayshift workers.This difference was statistically significant and is in line with findings from prospective studies, such as that of Ishizaki et al., 22 which also found a difference in BMI increase among rotating shift and dayshift male employees in a Japanese metal product factory after a 10-year follow-up.Similarly, Suwazono et al. 20 reported a significantly higher increase in BMI after 14 years for rotating shift workers compared with dayshift workers in a steel company in Japan.
However, multiple studies did not find statistically significant differences in BMI among rotating shift and dayshift workers, including a local study by Vásquez-Trespalacios et al. with 200 workers in a clinical care setting, which did not reveal any statistically significant association between overweight, obesity, or waist-hip ration and shift work. 11In the international scenario, a study by Nakamura et al. conducted to determine whether there is an association between shift work and risk factors for coronary disease in male Japanese workers reported a mean BMI of 23 ± 3.2 kg/m 2 for dayshift workers and of 23.2 ± 2,6 kg/m 2 for twoshift workers. 25imilarly, Geliebter et al. 26 evaluated the frequency of weight gain in rotating shift workers compared with dayshift workers and did not find significant differences in BMI between the two groups.In a systematic review of eight studies, five of which were considered to be high-and three of them low-quality studies.Seven studies presented crude results for an association between shift exposure and change in body weight.
Five studies presented weight-related outcomes adjusted for potentially relevant confounders (age, sex, bodyweight at baseline, and physical activity).Consequently, the evidence for a confoundersadjusted relationship between shift work exposure and body was considered to be insufficient; therefore, the development of more and higher-quality studies about the subject was deemed necessary to further clarify the underlying mechanisms. 27ith regard to lipid profile, higher total cholesterol levels were observed in rotating shift workers compared to dayshift workers (p < 0.05).This result is coherent with the findings from a study with male workers from a facture that manufactures semiconductors and related components in Kota Bharu, Malaysia, which reported a significantly higher prevalence of hypercholesterolemia (47.4%) and hypertriglyceridemia (42,1%) among rotating shift workers compared with dayshift workers. 28ásquez-Trespalacios et al. 11 and Knutsson 29 found that rotating shift workers are not less likely to perform regular physical activity compared with dayshift workers, which gains relevance when it is known that this factor may potentially act as a mediator in the relationship between work shift and weight gain. 29owever, the two studies show contrasting results for the waist-hip ratio, with the present study revealing a higher ratio among female rotating shift workers, with a statistical significant specific for women (p = 0,035). 29aking only fasting glycemia as a reference, no statistically significant difference concerning glycemia was found, which is in line with findings reported by Sharma et al., 30 who studied 12 otherwise healthy nurses performing rotating-shift work to evaluate its effect on glucose metabolism.As previously mentioned, as with the present study, there was no difference in glucose levels between the two shift schemes.However, based on the other parameters studied, particularly on the higher postprandial glucose concentrations observed during the night shift, the authors were able to conclude that the circadian variation that occurs in the night shift could impair beta cell function.
The present study had some limitations to be considered.Firstly, although sample size is statistically relevant, there is a remarkable prevalence of the male gender in the rotating shift (>80%).Furthermore, the cross-sectional design of the study limits the analysis of a possible relationship between shift work and overweight/obesity, which prevents understanding the causality and the temporal influence of work shift exposure.Additionally, information was collected from medical records obtained in an occupational information system, where there may be discrepancies in recording the type of patient's shift, especially when this information can be considered clinically irrelevant at the time of consultation.
Conversely, it is important to mention that some patients started to work in the day shift following a recommendation made by the occupational physician, but they had been previously exposed to rotating shift (for years, in some cases).In turn, some rotating shift workers were advised by the physician to limit night shifts as a strategy to protect their sleep, in some cases of patients with baseline epilepsy history.This was also indicated for patients who were previously healthy and developed remarkable changes in physiological variables, possibly as a strategy to prevent complications.Both scenarios did not represent a large portion of the sample population; however, these cases are described in order to acknowledge them as potential sources of information bias.

CONCLUSIONS
The analyzed information showed significant associations between work shift and changes in physiological parameters in terms of BMI and total cholesterol for both men and women, and in terms of waist-hip ratio for women.However, these findings should be confirmed by longitudinal studies that ideally integrate the analysis of the impact of work shift on sleep, at least from the patient's subject perspective, in order to allow for this additional correlation.

Table 2 .
Physiological measurements according to shift schedules * Changes in lipid profile without considering VLDL.LDL = low-density lipoprotein; HDL = high-density lipoprotein; SD = standard deviation; VLDL = very low-density lipoprotein.

Table 3 .
Mean waist-hip ratio stratified by sex