Multimorbidity and complex multimorbidity in Brazilian rural workers.

OBJECTIVE: To estimate the prevalence of multimorbidity and complex multimorbidity in rural workers and their association with sociodemographic characteristics, occupational contact with pesticides, lifestyle and clinical condition.
METHODS: This is a cross-sectional epidemiological study with 806 farmers from the main agricultural municipality of the state of Espírito Santo/Brazil, conducted from December 2016 to April 2017. Multimorbidity was defined as the presence of two or more chronic diseases in the same individual, while complex multimorbidity was classified as the occurrence of three or more chronic conditions affecting three or more body systems. Socio-demographic data, occupational contact with pesticides, lifestyle data and clinical condition data were collected through a structured questionnaire. Binary logistic regression was conducted to identify risk factors for multimorbidity.
RESULTS: The prevalence of multimorbidity among farmers was 41.5% (n = 328), and complex multimorbidity was 16.7% (n = 132). More than 77% of farmers had at least one chronic illness. Hypertension, dyslipidemia and depression were the most prevalent morbidities. Being 40 years or older (OR 3.33, 95% CI 2.06-5.39), previous medical diagnosis of pesticide poisoning (OR 1.89, 95% CI 1.03-3.44), high waist circumference (OR 2.82, CI 95% 1.98-4.02) and worse health self-assessment (OR 2.10, 95% CI 1.52-2.91) significantly increased the chances of multimorbidity. The same associations were found for the diagnosis of complex multimorbidity.
CONCLUSION: We identified a high prevalence of multimorbidity and complex multimorbidity among the evaluated farmers. These results were associated with increased age, abdominal fat, pesticide poisoning, and poor or fair health self-assessment. Public policies are necessary to prevent, control and treat this condition in this population.

Introduction

Exposure to dust, toxic chemicals, ultraviolet radiation, noise, and venomous animals in the daily routine of rural work represents potential sources of health problems for farmers [1]. Besides these, the transformations brought about by the mechanization and modernization of agricultural activities have modified the form of work organization in the field, with direct consequences to the physical and psychological domains, on the lifestyle and food consumption of these workers [2, 3].

This reality, aggravated by the reduced supply of health diagnosis and treatment services in rural areas [4], may increase farmers' vulnerability to chronic morbidity. Some evidence suggests worse health conditions and more disease among rural populations compared to other population groups [5,6,7]. It is noteworthy that these diseases may be isolated or coexist in the same individual, a condition known as multimorbidity [8]. Multimorbidity leads to a reduction in quality of life, higher mortality, polypharmacy, and an increase in the need for medical care, thus affecting health costs, and the productivity and functional capacity of individuals [9].

Knowing the distribution of diseases and the prevalence of multimorbidity in specific communities and populations is of fundamental importance for the planning and organization of health services and policies [10]. In this sense, when compared to the traditional criterion of classification of multimorbidity, the use of the concept of complex multimorbidity has been considered by some authors as a more effective way to identify people with priority care and plan the investment of health resources [11]. Nevertheless, no Brazilian study has been identified that has used this approach for the study of multimorbidity.

Given the above, and considering all the risk factors in the reality of rural work, the impact of chronic diseases on health, productivity and care costs, as well as the scarcity of data on multimorbidity in these professionals, this study aims to estimate the prevalence of multimorbidity and complex multimorbidity in rural workers and their association with sociodemographic characteristics, occupational contact with pesticides, lifestyle, and clinical condition.

Materials and methods

Data source

This is an cross-sectional epidemiological study derived from a larger study conducted in the municipality of Santa Maria de Jetibá, located in the state of Espírito Santo, southeastern Brazil, titled “Health condition and associated factors: a study of farmers in Espírito Santo—AgroSaúdES”, funded by the Espírito Santo Research Support Foundation (FAPES)—FAPES Notice / CNPq / Decit-SCTIE-MS / SESA—PPSUS—No. 05/2015.

Study population

The original study involved a representative sample of male and female farmers who met the following inclusion criteria: aged 18 to 59 years, not pregnant, having agriculture as their main source of income, and being in full employment for at least six months.

Sample size calculation

To identify the eligible farmers in the original study, data available in the individual and family records, as collected by the Family Health Strategy teams, were used to cover 100% of the 11 health regions in the municipality. Through this survey, we identified 7,287 farmers out of a total of 4,018 families. From this universe, the sample size calculation for the original project was performed considering 50% prevalence of outcomes (to maximize sample), 3.5% sampling error, and 95% significance level, making up a minimum sample of 708 farmers. 806 farmers were invited to compensate for possible losses. All sample size calculations were performed using the EPIDAT program (version 3.1). The participants were selected by a stratified lot, considering the number of families by health region and by Community Health Agent (CHA), in order to respect proportionality among the 11 regions and among the 80 CHAs. Only one individual per family was admitted, thus avoiding the interdependence of information. In case of refusal or non-attendance, a new participant was called from the reserve list, respecting the sex and the health unit of origin of the person who gave up/refused.

It should be noted that, due to the characteristics of the investigated municipality in which family farming predominates, the farmers who participated in this study had farming practices characterized by the predominance of polyculture and low degree of mechanization.

For the analytical developments proposed in this paper, the minimum sample size was calculated considering an estimated prevalence of multimorbidity in rural populations of 18.6% [12], 3% error, and a 95% confidence interval, resulting in a minimum required sample of 594 individuals. However, to improve sample representativeness and statistical relevance, we used data from all farmers who participated in the original project.

Data collection

Data collection of the original study took place between December 2016 and April 2017 in the dependencies of the health units of the municipality. A semi-structured questionnaire was applied, containing questions about socioeconomic, demographic, and occupational characteristics, occupational contact with pesticides, lifestyle, eating habits, and health condition, including the presence of chronic diseases and self-rated health. All this information was obtained through self-report. Anthropometric measurements were also collected, such as waist circumference, hemodynamic data such as systolic blood pressure (SBP), diastolic blood pressure (DBP), and blood drawn for biochemical examinations for markers such as thyroid stimulating hormone (TSH) and total cholesterol and fractions. To obtain biochemical data, 10 mL of blood was collected by venipuncture after 12 hours of fasting.

Only the variables of interest for this article were selected.

Variables selected for this study

Multimorbidity was evaluated in two different ways: through the traditional concept defined as the presence of two or more chronic diseases in the same individual (Multimorbidity ≥ 2 CD) [8] and through the concept of “complex multimorbidity”, classified as the occurrence of three or more chronic conditions affecting three or more body systems or different domains [13].

Chronic diseases were identified by counting morbidities reported by farmers from the question: “Has a doctor or other health professional ever told you that you had any of these diseases?”. Chronic diseases investigated in this study were: arrhythmia, infarction, stroke, diabetes mellitus, herniated disk, arthrosis, Repetitive Strain Injuries/Work Related Musculoskeletal Disorders (RSI/WMSD), renal disease, Parkinson's, Alzheimer's, cirrhosis, infertility, cancer, thyroid diseases, asthma, bronchitis, and pulmonary emphysema. In addition to the diseases referred to through self-report, we also considered the diagnoses of hypertension, dyslipidemia, thyroid disorders, and depression, performed through this research.

To determine the organic systems or domains affected according to each disease, we used the International Classification of Diseases– 11th revision (ICD-11), namely: circulatory system (hypertension, stroke, infarct, cardiac arrhythmia), endocrine, nutritional or metabolic disorders (diabetes, dyslipidemia, thyroid changes), musculoskeletal or connective tissue system (RSI/WMSD, arrhythmia), mental, behavioral or neurodevelopmental disorders (Alzheimer’s, depression), genitourinary system (infertility, kidney diseases), digestive system (liver cirrhosis), pulmonary system (bronchitis, asthma, pulmonary emphysema), and neoplasms (cancer).

The classification of blood pressure levels was performed based on the values of SBP and DBP according to the classification established in the VII Brazilian Hypertension Guidelines [14]. Thus, subjects with SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or who reported the use of blood pressure medications were considered hypertensive. These measurements were measured during the interview at least three times for each individual using the Omron® Automatic Pressure Monitor HME-7200, calibrated and validated by the National Institute of Metrology, Quality and Technology (INMETRO). To avoid interference with the results, subjects were instructed to sit and rest for about five minutes, empty their bladder and not consume food, alcohol, coffee or cigarettes for 30 minutes prior to the assessment. For data analysis, the average of two measurements was considered and a third measurement was performed whenever the difference between the first two was greater than 4 mmHg [15].

To investigate dyslipidemia, the levels of total cholesterol, HDL-c, LDL-c and triglycerides were measured. Total cholesterol and HDL cholesterol were determined, respectively, by the enzymatic colorimetric method with the Cholesterol Liquicolor Kit (In Vitro Diagnostica Ltda) and the Cholesterol HDL Precipitation Kit (In Vitro Diagnostica Ltda). To determine LDL cholesterol, we used the Friedewald formula [16]. Triglycerides were determined by the enzymatic colorimetric method with the Triglycerides Liquicolor mono® Kit (In Vitro Diagnostica Ltda). The results were classified according to the V Brazilian Guidelines on Dyslipidemias and Prevention of Atherosclerosis [16]. Individuals who reported the use of lipid-lowering drugs were also considered dyslipidemic.

In addition to self-report, the thyroid alteration was also evaluated by measuring the TSH through the chemiluminescence method. Farmers who had TSH values of 0.34 to 5.60 μUI/mL were considered as having “no thyroid alteration”, and individuals that had values above or below the reference range were classified as “with thyroid alteration”.

To evaluate the symptoms of depression, the Major Depressive Episode Module of the Mini-International Neuropsychiatric Interview (MINI) version 5.0 [17] was used. We considered "With Depression" farmers classified through the MINI with "Current Depression Episode" or "Recurrent Depression Episode".

Independent variables included socioeconomic variables (sex, age, race/color, marital status, schooling, socioeconomic class, and land tenure), occupational characteristics related to exposure to pesticides (use of Personal Protective Equipment, frequency and number of pesticides used), lifestyle (smoking, physical activity, alcohol consumption) and clinical conditions (previous intoxication by agrochemicals, waist circumference, and self-assessment of health). All these variables were collected by self-report.

Socioeconomic class was determined according to the Brazilian Economic Classification Criterion [18], in which A and B are the highest economic levels, C is intermediate, and D or E are low economic levels. Schooling was assessed by the number of years of study reported by the farmer.

Regarding lifestyle-related variables, all were obtained by self-report. It was considered that a “smoker” would be a farmer who reported smoking, an “ex-smoker” one who did not smoke, but who had smoked in the past, and a “non-smoker” would be a farmer who had reported never having smoked. Alcohol intake was assessed by asking, "How often do you drink alcohol?" Farmers who reported consuming alcohol, regardless of time or amount, were categorized as "Consuming." Those who reported not drinking alcohol were classified as "Not consuming". Farmers were also asked if they performed any other physical activities than those related to agricultural work. Answers were categorized as “Yes” or “No”, regardless of the type, time, or intensity of the exercise performed.

Health self-assessment was assessed by the question “In general, compared to people your age, how do you consider your own health status?”, assuming “very good”, “good”, “fair” and “poor.” Subsequently, we categorized the variable as “good/very good” and “fair/poor”. Waist circumference was classified according to the World Health Organization [19], considering values ≤ 94cm for men and ≤ 80cm for women as “without metabolic risk”, and “increased metabolic risk” for the other values. To collect this measurement, a 1cm wide Sanny® brand inextensible tape measure was used in triple measurement. The subject was instructed to stand, arms outstretched and feet together. The tape was positioned at the smallest curvature located between the last costal arch and the iliac crest. When it was impossible to locate the smallest curvature, we used the midpoint between these two anatomical points as the reference.

Statistical analyses

The absolute and relative frequencies of the independent variables were calculated according to the presence or absence of multimorbidity (≥ 2 CD) and multimorbidity complex outcomes. Then, the chi-square test was performed to verify the association between them. Variables with p-value < 5% in this test were included in the logistic regression analysis. The odds ratio was adjusted with respective 95% confidence intervals. The quality of the model was accounted for by the Hosmer-Lemeshow test.

The study was approved by the Research Ethics Committee of the Health Sciences Center of the Federal University of Espírito Santo, Opinion no. 2091172 (CAAE 52839116.3.0000.5060). All participants signed the Informed Consent Form.

Results

Of the 806 participants, 790 individuals completed the study. Of these, 612 (77.4%) had at least one chronic disease (Fig 1). Hypertension, dyslipidemia and depression were the most prevalent conditions, affecting 35.8% (n = 283), 34.4% (n = 272) and 16.9% (n = 134), respectively, of the farmers. Pulmonary emphysema, hepatic cirrhosis, infertility, Parkinson's, stroke, infarction, and Alzheimer's were reported by less than 1% of the sample. When the affected systems were evaluated, we found that 42.7% (n = 338) of the changes referred to endocrine, nutritional or metabolic diseases, followed by the circulatory system (37.4%, n = 296) and mental, behavioral or neurodevelopmental disorders (16.9%, n = 134).

Fig 1

Prevalence of chronic conditions expressed alone and according to organic system/ICD-11 domain affected in rural workers from Espírito Santo, Brazil.

Multimorbidity (≥ 2 CD) was found in 328 farmers (41.5%), and complex multimorbidity in 132 (16.7%) of the sample.

In the bivariate analyses (Table 1), the sociodemographic variables associated to both multimorbidity (≥ 2 DC) and complex multimorbidity were the age group and socioeconomic class. Sex (p = 0.005), marital status (p = 0.012) and schooling (p = 0.001) were only associated with multimorbidity (≥ 2 CD).

Table 1

Prevalence of multimorbidity (≥ 2 CD) and complex multimorbidity according to sociodemographic characteristics of farmers from Espírito Santo, Brazil.

Variable	Sample	Multimorbidity (≥ 2 CD)			Complex Multimorbidity
	n (%)	%	IC 95%a	p-valueb	%	IC 95%a	p-valueb
Sex
Male	413 (52.3)	36.8	(33.4–40.2)	0.005c	15.5	(13.0–18.0)	0.339
Female	377 (47.7)	46.7	(43.0–50.0)		18.0	(15.3–20.7)
Age Group
Up to 29 years	213 (27.0)	23.5	(20.5–26.5)	0.000c	9.4	(7.4–11.4)	0.000c
30 to 39 years	231 (29.2)	33.8	(30.5–37.1)		12.1	(9.8–14.4)
40 or more	346 (43.8)	57.8	(54.4–61.2)		24.3	(21.3–27.3)
Race / Color
White	702 (88.9)	41.2	(37.8–44.6)	0.572	16.2	(13.6–18.8)	0.318
Non-White	88 (11.1)	44.3	(40.8–47.8)		20.5	(17.7–23.3)
Marital status
Not married	59 (7.5)	27.1	(24.0–30.2)	0.012c	13.6	(11.2–16.0)	0.246
Married/Living with partner	678 (85.8)	41.7	(38.3–45.0)		16.4	(13.8–19.0)
Separated/Divorced/Widowed	53 (6.7)	54.7	(51.2–58.2)		24.5	(21.5–27.5)
Schooling
Less than 4 years	533 (67.5)	46.2	(42.7–49.7)	0.001c	18.0	(15.3–20.7)	0.367
4 to 8 years	173 (21.9)	33.5	(30.2–36.8)		13.9	(11.5–16.3)
More than 8 years	84 (10.6)	28.6	(25.4–31.8)		14.3	(11.9–16.7)
Socioeconomic class
Class A or B	58 (7.3)	31.0	(27.8–34.2)	0.033c	6.9	(5.1–8.7)	0.050c
Class C	395 (50.0)	39.0	(35.6–42.4)		15.9	(13.4–18.4)
Class D or E	337 (42.7)	46.3	(42.8–49.8)		19.3	(16.5–22.1)
Land ownership
Owner	609 (77.1)	41.4	(38.0–44.8)	0.884	15.6	(13.1–18.1)	0.125
Non-Owner	181 (22.9)	42.0	(38.6–45.4)		20.4	(17.6–23.2)

a Confidence Interval.

b Chi-square test.

C Statistically significant value (p <0.05).

With regard to the occupational characteristics related to the use of pesticides, lifestyle and clinical condition, we verified that alcohol consumption, medical diagnosis of pesticide intoxication, waist circumference, and health self-assessment were associated with both outcomes (Table 2). Smoking was only associated with multimorbidity (≥ 2 CD).

Table 2

Prevalence of multimorbidity (≥ 2 CD) and complex multimorbidity according to occupational characteristics related to the use of pesticides, lifestyle and clinical condition of farmers from Espírito Santo, Brazil.

Variables	Sample	Multimorbidity (≥ 2 CD)			Complex Multimorbidity
	n	%	IC 95%a	p-valueb	%	IC 95%a	p-valueb
Type of occupational contact with pesticide
Direct	550 (69.6)	40.7	(37.3–44.1)	0.494	15.3	(12.8–17.8)	0.101
Indirect/Non-Contact	240 (30.4)	43.3	(39.8–46.8)		20.0	(17.2–22.8)
Total number of pesticides used
None	240 (32.0)	43.3	(39.8–46.8)	0.502	20.0	(17.1–22.9)	0.268
1 to 5 types of pesticides	223 (29.7)	42.6	(39.1–46.1)		15.7	(13.1–18.3)
More than 5 pesticides	287 (38.3)	38.7	(35.2–42.2)		15.0	(12.4–17.6)
Use of PPE
Do not use PPE/Incomplete PPE	380 (49.2)	42.6	(39.1–46.1)	0.194	16.6	(14.0–19.2)	0.073
Complete PPE	152 (19.7)	34.9	(31.5–38.3)		11.2	(11.2–13.4)
Without direct contact	240 (31.1)	43.3	(39.8–46.8)		20.0	(17.2–22.8)
Frequency of contact with pesticide
Daily/Weekly	453 (61.4)	40.8	(37.3–44.3)	0.717	15.2	(12.6–17.8)	0.235
Monthly/Yearly	206 (27.9)	43.7	(40.1–47.3)		17.5	(14.8–20.2)
Without contact	79 (10.7)	44.3	(40.7–47.9)		22.8	(19.8–25.8)
Smoking
Non-smoker	665 (84.2)	39.8	(36.4–43.2)	0.028c	15.9	(13.4–18.4)	0.181
Smoker or ex-smoker	125 (15.8)	50.4	(46.9–53.9)		20.8	(18.0–23.6)
Practices physical activity
No	669 (84.7)	42.9	(39.4–46.4)	0.064	17.2	(14.6–19.8)	0.394
Yes	121 (15.3)	33.9	(30.6–37.2)		14.0	(11.6–16.4)
Alcohol consumption
Does not consume	444 (56.2)	46.8	(43.3–50.3)	0.001c	21.2	(18.3–24.1)	0.000c
Consumes	346 (43.8)	34.7	(31.4–38.0)		11.0	(8.0–13.2)
Medical diagnosis of poisoning by pesticides
Yes	59 (7.5)	57.6	(54.1–61.1)	0.010c	32.2	(28.9–35.5)	0.001c
No	729 (92.5)	40.3	(36.9–43.7)		15.5	(13.0–18.0)
Waist circumference
Without metabolic risk	384 (48.7)	26.0	(22.9–29.1)	0.000c	9.4	(7.4–11.4)	0.000c
Increased metabolic risk	405 (51.3)	56.3	(52.8–59.8)		23.7	(20.7–26.7)
Health self-assessment
Good/ Very good	459 (58.1)	32.2	(28.9–35.5)	0.000c	10.5	(8.4–12.6)	0.000c
Fair/Poor	331 (41.9)	54.4	(50.9–57.9)		25.4	(22.4–28.4)

a Confidence Interval.

b Chi-square test.

C Statistically significant value (p <0.05).

After a logistic regression analysis (Table 3), it was verified that being 40 years of age or older (OR 3.33, 95% CI 2.06–5.39), previous medical diagnosis of pesticide poisoning (OR 1.89, 95% CI 1.03–3.44), high waist circumference (OR 2.82, 95% CI 1.98–4.02), and fair or poor health self-assessment (OR 2.10, 95% CI 1.52–2.91) significantly increased the chances of multimorbidity (≥ 2 DC).

Table 3

Association between multimorbidity (≥ 2 DC), complex multimorbidity and socio-demographic characteristics, occupational contact with pesticides, lifestyle and clinical condition in farmers from Espírito Santo, Brazil.

	Multimorbidity (≥ 2 DC)				Complex Multimorbidity
Variables	p-valuea	OR adjustedb	LLc	ULd	p-valuea	OR adjustedb	LLc	ULd
Sex
Male		1
Female	0.854	1.037	0.702	1.533
Age Group
Up to 29 years		1				1
30 to 39 years	0.131	1.438	0.897	2.305	0.682	1.141	.606	2.149
40 or more	0.000e	3.336	2.065	5.390	0.004e	2.250	1.295	3.909
Marital status
Not married		1
Married / Living with partner	0.999	1.000	0.511	1.957
Separated / Divorced / Widowed	0.951	1.028	0.418	2.527
Schooling
More than 8 years	0.643	1.162	0.617	2.188
4 to 8 years	0.971	1.011	0.559	1.829
Less than 4 years		1
Socioeconomic class
Class A or B		1				1
Class C	0.652	1.165	0.600	2.263	0.174	2.115	0.719	6.224
Class D or E	0.256	1.488	0.749	2.956	0.117	2.378	0.805	7.025
Smoking
Non-smoker		1
Smoker or ex-smoker	0.070	1.534	0.965	2.438
Alcohol consumption
Does not consume		1				1
Consumes	0.314	0.835	0.588	1.186	0.069	0.666	0.429	1.032
Medical diagnosis of poisoning by pesticides
No		1				1
Yes	0.038e	1.891	1.037	3.449	0.005e	2.474	1.319	4.638
Waist Perimeter
Without metabolic risk		1				1
Increased metabolic risk	0.000e	2.829	1.986	4.029	0.001e	2.142	1.370	3.349
Health Self-Assessment
Good/ Very good		1				1
Fair/Poor	0.000e	2.107	1.524	2.913	0.000e	2.248	1.493	3.384

a Binary Logistic Regression. Enter Method.

b Odds Ratio.

C Lower Limit– 95% Confidence interval.

d Upper Limit– 95% Confidence interval.

e Statistically significant value (p <0.05).

Hosmer-Lemeshow = 0.795 (Multimorbidity ≥ 2 DC) and 0.701 (Complex Multimorbidity)

The same associations were found for the diagnosis of complex multimorbidity.

Discussion

This is the first Brazilian study to estimate the prevalence of multimorbidity in rural workers and to use the complex multimorbidity criterion to determine this outcome. The representative sample, stratified and randomly selected, allows us to extrapolate the results to the target population.

Agriculture is often described as an occupation that promotes health, being associated with the image of a healthy lifestyle with exposure to nature, outdoors, physical effort, and a diet based on natural foods [3]. However, the results reflect a different reality. Eight out of 10 farmers had at least one chronic disease, more than 40% had two or more, and around 17% had three or more, affecting at least three or more organic systems or ICD-11 domains.

Among the chronic conditions analyzed, there was a predominance of arterial hypertension and dyslipidemia, similar to other multimorbidity studies performed in Brazil [20] and in countries such as Portugal [21] and Australia [22]. These two morbidities were also more frequent in disease pattern studies conducted for the population of the United States [23] and New York State [24]. In a systematic review involving studies from 16 European countries [9] hypertension also occupied a prominent position, as well as countries such as China, Finland, Ghana, Russia, South Africa [25] and in four Greater Mekong countries [26]. The prevalence of these diseases is also observed when evaluating multimorbidity studies with the elderly [27, 28]. These values, however, are above the estimate for the Brazilian population through wide-ranging studies such as VIGITEL (24.1%) [29] and National Household Sample Survey—PNAD (20.9%) [30].

By analyzing the presence of chronic diseases according to the organic system or affected area, it was found that the most frequent ones were endocrine, nutritional or metabolic diseases, due to the high rates of dyslipidemia, diabetes and thyroid disorders, and the circulatory system, due to arterial hypertension. In Spain, a research project with more than one million patients also identified a predominance of alterations in these two systems, especially in individuals over 45 years old [31]. In Ethiopia, however, musculoskeletal system diseases were the most prevalent, affecting about 20% of the sample [32]. In an Australian study, there were 32.4% alterations involving the circulatory system, 32.1% of musculoskeletal and connective tissue, and 30.7% of endocrine, nutritional and metabolic alterations [33]. These results corroborate the three globally most common multimorbidity groups, composed of “metabolic disorders”, including diabetes, obesity and hypertension, “mental-articular disorders”, including arthritis and depression, and “cardio-respiratory” including angina, asthma and chronic obstructive pulmonary disease [25].

The high number of farmers with chronic conditions involving mental, behavioral or neurodevelopmental disorders, especially due to the high prevalence of depression in these workers, is worth highlighting. Depressive disorders were also among the most frequent in the study by Prazeres and Santiago [21] and the study by Harrison et al. [33], in which mental/psychological changes were found in 26.7% of the sample.

The prevalence of multimorbidity presented by rural workers was higher than estimated for the Brazilian population through the World Health Survey (13.4%, 95% CI 12.4–14.5) [34] and the National Health Survey [12], in which the expected multimorbidity was 18.6% (95% CI 17.2–20.0%) in rural areas and 22.8% (95% CI 22–23.5%) in Brazilian urban areas. It was also higher than the prevalence found in developed countries, such as Portugal (38.3%) [35], Spain (20%) [36], Canada (12,9%) [37], Denmark (22%) [38], and Belgium (22.7%) [39], and, in middle-income countries, where 12.6% (Mexico), 19.4% (Russia), and 10.4% (South Africa) of the 40–49 year-old population reported two or more chronic diseases [40]. In a study involving six countries in South America and the Caribbean, the self-reported multimorbidity ranged from 12.4% in Colombia to 25.1% in Jamaica [41]. It is estimated that between 16% and 57% of adults in developed countries suffer from more than one chronic condition [42]. In European health systems, the estimated prevalence of multimorbidity was 33% in 2015 [43]. A systematic review by Nguyen et al. [44] involving only community studies found a combined global prevalence of multimorbidity of 33.1%. Among the 37 representative studies of developed countries involved in this review, the lowest prevalence of multimorbidity was identified in Hong Kong (3.5%) and the highest in Russia (70%). Among developing countries, the lowest percentage identified was in 26 Indian villages (1%) and the highest prevalence was in China (90%) [44]. We emphasize that the methodological differences, especially those related to the target population and the diagnostic criterion of multimorbidity, limit the comparison and interpretation of the results.

With respect to the complex multimorbidity, few studies are available in international literature using this methodology. An Australian study estimated that 25.7% of the population had two or more chronic diseases and 12.1% showed complex multimorbidity [33]. This methodology shows itself as a more discriminatory measure, and among farmers, reduced the prevalence of multimorbidity compared to the criterion of two or more diseases. Harrison et al. [11] argue that counting affected body systems instead of evaluating individual chronic conditions has the advantage of more carefully identifying patients who need more complex care, as well as the number and types of specialized health services that are necessary for such assistance, thus being a more useful and effective way of planning actions and investments in health [11].

The only sociodemographic factor that remained associated with multimorbidity, regardless of the form of evaluation of this outcome, was age. This association is well documented in the literature. In Canada, the prevalence of multiple diseases increased from 12.5% in the younger age group (18–24 years) to 63.8% in the more advanced ones (≥ 65 years) [45]. The onset of chronic diseases with increasing age seems to be related to the physiological imbalance and general senescence in multiple organs that aging causes [46]. This influence can be seen in comparison with the significant increase in multimorbidity in studies conducted with the elderly. Nunes et al. [47], analyzing a representative national sample of the non-institutionalized population, identified a prevalence of 82.4% of multimorbid individuals (CI 95% 78.5–85.7%) among older adults aged 80 years or older. In Southern Brazil, the estimate was 93.4% in the city of Pelotas [48] and 81.3% in Bagé [49]. In Canada, the overall prevalence of multimorbidity in the older age group (≥ 85 years) was 58.6% higher when compared to younger age groups [50]. A marked difference was also found in the study by Puth et al. [51] in Germany, where the prevalence of this condition increased from 7% in individuals aged 18–29 to 77.5% in those aged 80 and older.

Although there is a large amount of evidence that the occurrence of multimorbidity is higher in females and at low socioeconomic and educational levels [52], the association with socioeconomic variables is very heterogeneous between studies [53]. As with our results, the lack of association with income [26], education and gender [5] has also been documented. Several factors may be related to these results. Among them, we can mention the homogeneity of the rural population investigated in relation to income (92.7% belonged to lower socioeconomic classes—C, D or E), education (89.4% had fewer than 8 years of schooling) and marital status (85.8% were married or living with a partner), compared with the urban population, which generally has more heterogeneous strata, and is therefore more differentiated. This homogeneity of the analyzed population may have compromised the identification of statistically significant differences between strata.

Regarding gender, considering that in rural areas there is limited access to health services [54], there may have been under-reporting in the diagnosis of chronic diseases, especially among women, who generally use health services more often than men. This may have led to a reduction in self-reported disease among women and consequently the absence of statistical association between genders. In addition, the frequency of some diseases is known to vary by gender [55]. In this sense, the methodological differences regarding the type and quantity of diseases to be considered in each study for classification of multimorbidity have a direct influence on the results found by each author [56]. As an example, we can mention the study by Pengpid and Peltzer [26] that identified a higher prevalence of multimorbidity in men due to the inclusion of smoking and alcoholism among the evaluated chronic conditions. The methodology used for disease identification may also have contributed to the difference between the results [56, 57]. In the study by Guerra et al. [58], for example, gender was not associated with multimorbidity measured from administrative data, but was associated with self-reported multimorbidity, regardless of the cutoff point adopted. For this reason, the fact that we used both self-reported data as well as biochemical and hemodynamic data may justify the differences in association found when compared to other studies, which mostly use self-reported data.

In addition to methodological differences, the lack of association with some sociodemographic variables may be due to the presence of factors that contribute more closely to the development of multimorbidity, such as waist circumference, previous pesticide poisoning or other factors not intrinsic to agricultural activity within the scope of this study. Further studies involving farmers are needed to better understand the risk factors present in daily agricultural work, thus facilitating the comparison of results.

This study, however, strengthens the evidence of the association between the accumulation of visceral fat and the occurrence of chronic diseases. In addition to reflecting the level of central adiposity, high waist circumference is also directly related to excess body fat, and is considered a major risk factor for the early development of various morbidities, including hypertension, diabetes, dyslipidemias, and cancers [59]. Corroborating these results, a cohort conducted in the United Kingdom concluded that overweight participants were 25% more likely to have at least one of 11 assessed health conditions compared to normal weight subjects. In obese patients, the odds increased to 54%, 81% and 124% for categories I, II and III, respectively [60]. Similarly, different disease patterns identified in the Brazilian population were also associated with obesity [54]. In low- and middle-income countries, the prevalence of multimorbidity increased 5.78 fold in obese individuals when compared to those of normal weight [61].

In addition to the negative influence of age and waist circumference on the occurrence of multiple diseases, previous poisoning by pesticides also seems to be related to this condition, increasing by 1.89 and 2.47 the chance of occurrence of multimorbidity and complex multimorbidity among workers in rural areas, respectively. We highlight that there are several harmful health effects that have been related to the use of pesticides, among them mental disorders, respiratory, and autoimmune diseases [62]. Farmers who have reported being poisoned with pesticides may be more chronically exposed to these products and, therefore, more likely to show the cumulative deleterious effects of this exposure. The comparison of this result becomes limited, since other similar studies in literature were not found. It should be emphasized that the association between the outcome and the variables of exposure, intensity and frequency of contact with pesticides may not have been evidenced, due to the limitations of cross-sectional studies, when compared to cohort studies, to evaluate the oscillations in occupational exposure years.

Another factor associated with the higher prevalence of multimorbidity was the health self-assessment. Fair or poor health perception doubled the chances of occurrence of multimorbidity (CD ≥ 2) (OR 2.10, 95% CI 1.52–2.91) or complex multimorbidity (OR 2.24, 95% CI 1.49–3.38) among farmers. In European countries, the increased number of chronic diseases was also associated with a higher probability of reporting poor/fair health self-perception (OR = 2.13, 95% CI 2.03–2.24) [9]. The same association was found in countries in South America and the Caribbean [41], in the rural population of South Africa [63], and in Myanmar [57].

Thus, we verified that the rural population analyzed showed alarming rates, not only of a single chronic condition, but of multiple conditions. The occurrence of multiple diseases has been associated with aging, being overweight, the way farmers perceive their health, and occupational exposure to agrochemicals. It is also worth noting that, despite the fact that it was not within the scope of this research, it is known that factors such as the difficulty of access to health services and specialized treatments, which are common in rural communities, further increase the vulnerability of these workers to the development of multimorbidity.

Considering the serious economic, social and health implications of the presence of multiple chronic diseases in people of working age [60], it is necessary to re-examine the focus of the health system, which currently does not seem well suited to the new medical and social reality that the multimorbidity presents. As strategies and public policies must ensure holistic care, implementing actions that consider the particularities and vulnerabilities of this community, as well as stimulating self-care, controlling modifiable risk factors and adopting healthy behaviors [32]. Also, the training of health teams to attend multimorbid patients is essential, as well as the elaboration of clinical protocols for multiple diseases, and, especially, effective allocation of financial resources [64]. In this sense, although there is a great value in measuring the occurrence of chronic conditions in an individualized way, complex multimorbidity, through the measurement of the patterns of the bodily systems affected by chronic conditions, seems to be a good tool to screen, select and align services and prioritize resources more effectively to patients with greater need [11].

Among the limitations of this study, we emphasize that diseases identified through self-report may be subject to the under-reporting of diagnosis or memory bias. Despite the extensive list of diseases included, some chronic conditions may not have been identified. The lack of standardization on the way to evaluate multimorbidity and the unavailability in the literature of articles on this theme involving farmers limited the comparison of the results. Because this is a cross-sectional study, reverse causality cannot be disregarded in the data interpretation. It should be noted that the prevalence of multimorbidity may be under-reported since patients with more severe conditions may not have participated in the study.

Despite these limitations, it is worth noting the unprecedented nature of the study in relation to the involved target population, the adoption of the complex multimorbidity criterion and the included variables, such as waist circumference and pesticide intoxication, in articles of this theme. We sought in this study to involve a representative sample to allow extrapolation of the results to farmers with similar profile. In addition, in order to minimize the errors of underdiagnosis, the identification of diseases occurred, both through self-report, as well as through laboratory and hemodynamic measures.

Conclusions

We identified a high prevalence of multimorbidity and complex multimorbidity among the evaluated farmers. Factors associated with these outcomes in this population were increased age, high waist circumference, history of pesticide intoxication, and poor or fair health self-assessment. Considering the serious physical, functional, psychological and economic implications of multimorbidity, it is fundamentally important to plan economic, social and health policies aimed at controlling, monitoring and treating this condition in this professional category. In addition, new research is needed to evaluate in more detail the impacts that the risk factors identified in this study may have on the health of rural workers, especially those resulting from being overweight and from occupational exposure to agrochemicals, both of which are associated in this study with the presence of multiple diseases.

(XLSX)

Click here for additional data file.

Database variable codes.

(DOCX)

Click here for additional data file.

9 Sep 2019

PONE-D-19-17072

Multimorbidity and complex multimorbidity in Brazilian rural workers

PLOS ONE

Dear Dr. Salaroli,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

In particular, please address the point made by Reviewer #1 on the lack of socio-demographic correlates. I suspect the reviewer is expecting to see multimorbidity occurring mainly in those with lower socio-economic status, and also those who are older. Please comment on whether there are special circumstances for the cohort studied that mask these factors. Please also provide more details on the sampling strategy and statistical procedures, in case this might be the result of a simple mistake.

We would appreciate receiving your revised manuscript by Oct 24 2019 11:59PM. When you are ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter.

To enhance the reproducibility of your results, we recommend that if applicable you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). This letter should be uploaded as separate file and labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. This file should be uploaded as separate file and labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. This file should be uploaded as separate file and labeled 'Manuscript'.

Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

We look forward to receiving your revised manuscript.

Kind regards,

Siew Ann Cheong, Ph.D.

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

1. Please include additional information regarding the semi-structured questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you developed a questionnaire as part of this study and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

2. Please provide further details regarding how participants were recruited.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In the present manuscript Petarli and colleagues aimed to estimate the prevalence of multimorbidity and complex multimorbidity in rural workers and their association with sociodemomographic, occupational, and clinical correlates. The article is quite interesting as it provides estimates based on a population with quite specific lifestyle and occupational exposure to specific agents (e.g. pesticides), which might require targeted interventions/public health policies. What I found quite important to highlight is that in the multivariable model, in contrast to previous literature, almost no socio-demographic correlates (with the exception of age, which is a well-established strong predictor of multimorbidity) remained significant, while all clinical ones were found associated with the likelihood of having multimorbidity and I would suggest authors to elaborate more on this (and possible explanations).

Other points:

- Introduction and discussion are sufficiently well-written but to make comparisons even more relevant the cited literature should be enriched with even more recent published papers on the epidmiology of multimorbidity (especially in western countries)

- Further details on the sampling strategy should be provided. Authors provide a SS calculation to detect selected outcomes (which should also be expanded with additional information to allow reader to better understand and repeat calculations) but I doubt that this survey was originally based on this SS calculation, please explain and elaborate more on the sampling in general. Additionally, authors do no mention the use of survey weights: does this mean that non-respondent problem was encountered?

- The authors adopt an empirical approach for the inclusion of covariates in the multirvariable models which is somehow in contrast with the current tendency to select covariates mostly by literature research. Please justify this approach

Reviewer #2: General comments: Authors have highlighted a very pertinent issues on multimorbidity and its complexities. The overall article content is clear; however,it needs to be reviewed and copy-edited for English language to make language of the article clear, correct before re-submitting.

Specific comments:

1. Abstract conclusion is not clear and need rephrasing for better clarity.

2. line 100-105: The sentences are confusing and not clear. It would be better to rewrite the sentences with more clarity. It is not clear what authors want to convey by saying "lack of standardization regarding the most effective way to diagnose this condition."

3. The study is based on data from "Condition of health and associated factors: a study in farmers of Espírito Santo - AgroSaúdES". As a part of better clarity it would be appropriate to provide a brief description of the parent study: how the sampling was done, how data were collected and what information collected in the parent study.

4. Line 120-121: Need rephrasing!

5. Sample size calculation: Authors have mentioned that they included 806 sample for their study which would provide 96.9% power to detect the outcome. This is not clear. Authors should provide a detail description of the sample size calculation in a separate section or within the section of study population.

6. It would be better to provide a more detail information on how and what information were collected under Anthropometric, hemodynamic and biochemical data.

7. line 157: what is the full form of RSI/WMSD

8. line no.170- The classification of Blood pressure level and not the "pressure level "

9. Line 197: lifestyle (smoking, physical activity, alcohol consumption)- How these lifestyle factors were elicited and examined?

10. line 206-208: Authors have categorised waist circumference based on WHO classification and termed it as without cardiovascular risk and increased cardiovascular risk. However, this WHO classification relates to metabolic risk rather than CVD risk.

11. Is there any other anthropometric measurements such as height and weight (BMI) were collected in the survey and if collected why authors have not included in the study.

12. Line 368: What are the types of diseases when you mentioned the word "disease" in this sentence?

13. References:

a. Line 450: link not working

b. References should be checked for uniformity in the formating style and journal names.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Raffaele Palladino

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files to be viewed.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org. Please note that Supporting Information files do not need this step.

26 Sep 2019

Dear Editor,

We appreciate all the contributions made by the reviewers. We reviewed the entire manuscript and followed all suggestions. For ease of identification, corrections have been marked in red in the article and we have answered questions one by one in this letter.

Regarding the lack of association between sociodemographic factors and multimorbidity (except age), I inform you that, as requested, we insert possible explanations for these results. We emphasize, however, that although less frequent, there are similar results already documented in the literature [16] [165].

We also report that the article was reviewed by a research expert at Proof Reading Service (https://www.proof-reading-service.com/en/). The certificate provided by the company has been attached to the PLOS ONE platform.

Sincerely Regards,

The authors.

Journal Requirements:

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

http://www.journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and http://www.journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

ANSWER: We review and follow the PLOS ONE's style.

1. Please include additional information regarding the semi-structured questionnaire used in the study and ensure that you have provided sufficient details that others could replicate the analyses. For instance, if you and it is not under a copyright more restrictive than CC-BY, please include a copy, in both the original language and English, as Supporting Information.

ANSWER: Further details on the methodological processes were included.

2. Please provide further details regarding how participants were recruited.

ANSWER: The methodology has been rewritten in more detail.

Reviewer's Responses to Questions

Comments to the Author

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In the present manuscript Petarli and colleagues aimed to estimate the prevalence of multimorbidity and complex multimorbidity in rural workers and their association with sociodemomographic, occupational, and clinical correlates. The article is quite interesting as it provides estimates based on a population with quite specific lifestyle and occupational exposure to specific agents (e.g. pesticides), which might require targeted interventions/public health policies. What I found quite important to highlight is that in the multivariable model, in contrast to previous literature, almost no socio-demographic correlates (with the exception of age, which is a well-established strong predictor of multimorbidity) remained significant, while all clinical ones were found associated with the likelihood of having multimorbidity and I would suggest authors to elaborate more on this (and possible explanations).

ANSWER: Although there is a large amount of evidence that the occurrence of multimorbidity is higher in females and at low socioeconomic and educational levels [1], the association with socioeconomic variables is very heterogeneous between studies [2]. As with our results, the lack of association with income [3], education and gender [4] has also been documented. Several factors may be related to these results. Among them, we can mention the homogeneity of the rural population investigated in relation to income (92.7% belonged to lower socioeconomic classes - C, D or E), education (89.4% had fewer than 8 years of schooling) and marital status (85.8% were married or living with a partner), compared with the urban population, which generally has more heterogeneous strata, and is therefore more differentiated. This homogeneity of the analyzed population may have compromised the identification of statistically significant differences between strata.

Regarding gender, considering that in rural areas there is limited access to health services [5], there may have been under-reporting in the diagnosis of chronic diseases, especially among women, who generally use health services more often than men. This may have led to a reduction in self-reported disease among women and consequently the absence of statistical association between genders. In addition, the frequency of some diseases is known to vary by gender [6]. In this sense, the methodological differences regarding the type and quantity of diseases to be considered in each study for classification of multimorbidity have a direct influence on the results found by each author [7]. As an example, we can mention the study by Pengpid and Peltzer [3] that identified a higher prevalence of multimorbidity in men due to the inclusion of smoking and alcoholism among the evaluated chronic conditions. The methodology used for disease identification may also have contributed to the difference between the results [7,8]. In the study by Guerra et al. [9], for example, gender was not associated with multimorbidity measured from administrative data, but was associated with self-reported multimorbidity, regardless of the cutoff point adopted. For this reason, the fact that we used both self-reported data as well as biochemical and hemodynamic data may justify the differences in association found when compared to other studies, which mostly use self-reported data.

In addition to methodological differences, the lack of association with some sociodemographic variables may be due to the presence of factors that contribute more closely to the development of multimorbidity, such as waist circumference, previous pesticide poisoning or other factors not intrinsic to agricultural activity within the scope of this study. Further studies involving farmers are needed to better understand the risk factors present in daily agricultural work, thus facilitating the comparison of results.

All of these considerations were included in the discussion of the article in lines 421 - 456.

Other points:

- Introduction and discussion are sufficiently well-written but to make comparisons even more relevant the cited literature should be enriched with even more recent published papers on the epidmiology of multimorbidity (especially in western countries)

ANSWER: As requested, more recent articles were inserted for comparison purposes. However, it should be noted that a large part of the studies on this subject is conducted with the elderly and our target audience were adult individuals. For this reason, we were careful in choosing the articles to be inserted. It is also worth mentioning that, unfortunately, there are few articles that address multimorbidity in farmers, which made the comparison of results difficult and limited.

- Further details on the sampling strategy should be provided. Authors provide a SS calculation to detect selected outcomes (which should also be expanded with additional information to allow reader to better understand and repeat calculations) but I doubt that this survey was originally based on this SS calculation, please explain and elaborate more on the sampling in general. Additionally, authors do no mention the use of survey weights: does this mean that non-respondent problem was encountered?

ANSWER: For ease of understanding, the sampling session has been rewritten to include more detail on the sample size calculation of the original study and this article.

- The authors adopt an empirical approach for the inclusion of covariates in the multirvariable models which is somehow in contrast with the current tendency to select covariates mostly by literature research. Please justify this approach

ANSWER: The authors chose to insert the variables according to the statistical associations in the chi-square test because they consider this methodology more appropriate given the scarcity of data available in the literature on multimorbidity and its risk factors in a population as specific as farmers. The articles available in the literature, in general, have as target audience very different populations of rural workers, besides presenting methodological differences that may impact the found associations. It is also worth noting that complex multimorbidity has been poorly investigated, and the authors have not identified any studies with this approach in farmers, which limits the theoretical basis on the factors associated with this condition.

Reviewer #2: General comments: Authors have highlighted a very pertinent issues on multimorbidity and its complexities. The overall article content is clear; however,it needs to be reviewed and copy-edited for English language to make language of the article clear, correct before re-submitting.

ANSWER: The article has been reviewed by a specialist in the research area of Proof Reading Service (https://www.proof-reading-service.com/pt/). The certificate provided by the company was attached to the PLOS ONE platform.

Specific comments:

1. Abstract conclusion is not clear and need rephrasing for better clarity.

ANSWER: The Abstract conclusion was rephrased for better clarity.

2. line 100-105: The sentences are confusing and not clear. It would be better to rewrite the sentences with more clarity. It is not clear what authors want to convey by saying "lack of standardization regarding the most effective way to diagnose this condition."

ANSWER: This paragraph was rephrased for better clarity.

3. The study is based on data from "Condition of health and associated factors: a study in farmers of Espírito Santo - AgroSaúdES". As a part of better clarity it would be appropriate to provide a brief description of the parent study: how the sampling was done, how data were collected and what information collected in the parent study.

ANSWER: The methodology has been rewritten to include more detail about the calculation of the original project sample size, how to collect data, and what data was collected as requested by the reviewers. However, to prevent the article from becoming too long, more detail was added especially to the data that was used in this article, since the original project was very large and collected a diverse range of information.

4. Line 120-121: Need rephrasing!

ANSWER: This paragraph was rephrased for better clarity.

5. Sample size calculation: Authors have mentioned that they included 806 sample for their study which would provide 96.9% power to detect the outcome. This is not clear. Authors should provide a detail description of the sample size calculation in a separate section or within the section of study population.

ANSWER: For the analytical developments proposed in this paper the minimum required sample size was recalculated based on an estimated prevalence of multimorbidity in rural populations of 18.6% [10], 3% error, 95% confidence interval, resulting in a minimum required sample of 594 individuals. However, to improve sample representativeness and statistical relevance, we used data from all farmers who participated in the original project. This information was entered in the “sample size calculation” session to facilitate readers' understanding.

6. It would be better to provide a more detail information on how and what information were collected under Anthropometric, hemodynamic and biochemical data.

ANSWER: More details on anthropometric, hemodynamic and biochemical data collection were included in the "data collection" session and in the "variables used in the present study" section. However, to prevent the article from becoming too long, more detail was added to the data that was used in this article, since the original project was very large and collected a diverse range of information.

7. line 157: what is the full form of RSI/WMSD

ANSWER: Full form of RSI/WMSD is Repetitive Strain Injuries/Work Related Musculoskeletal Disorders. This full form was inserted in text.

8. line no.170- The classification of Blood pressure level and not the "pressure level "

ANSWER: Done.

9. Line 197: lifestyle (smoking, physical activity, alcohol consumption)- How these lifestyle factors were elicited and examined?

ANSWER: Further details were added on how lifestyle variables were collected and analyzed in the “Variables selected for this study” session.

10. line 206-208: Authors have categorised waist circumference based on WHO classification and termed it as without cardiovascular risk and increased cardiovascular risk. However, this WHO classification relates to metabolic risk rather than CVD risk.

ANSWER: The term “cardiovascular risk” has been replaced in the text and tables by “metabolic risk” as recommended.

11. Is there any other anthropometric measurements such as height and weight (BMI) were collected in the survey and if collected why authors have not included in the study.

ANSWER: Yes, weight and height (BMI) measurements were collected, but the authors chose to use waist circumference data as they consider that the location of body adiposity is a higher risk factor for disease onset compared to being overweight itself assessed through the IMC [11] In addition, waist circumference is less susceptible to muscle mass interference than BMI and therefore less subject to bias. The authors did not maintain both measures to avoid the effect of collinearity between variables.

12. Line 368: What are the types of diseases when you mentioned the word "disease" in this sentence?

ANSWER: It refers to at least one of the 11 health conditions investigated in the article cited. For the sake of clarity, this passage has been rewritten.

13. References:

a. Line 450: link not working

ANSWER: The link has been changed.

b. References should be checked for uniformity in the formating style and journal names.

ANSWER: References were checked and corrected as suggested by the reviewers.

REFERENCES

1. Violan C, Foguet-Boreu Q, Flores-Mateo G, Salisbury C, Blom J, Freitag M, et al. Prevalence, determinants and patterns of multimorbidity in primary care: a systematic review of observational studies. PLoS One. 2014;9(7):e102149. https://doi.org/10.1371/journal.pone.0102149 PMID: 25048354.

2. Pathirana TI, Jackson CA. Socioeconomic status and multimorbidity: a systematic review and meta‐analysis. Aust N Z J Public Health. 2018;42(2):186-194. https://doi.org/10.1111/1753-6405.12762 PMID: 29442409.

3. Pengpid S, Peltzer K. Multimorbidity in Chronic Conditions: Public Primary Care Patients in Four Greater Mekong Countries. Int J Environ Res Public Health. 2017;14(9). pii: E1019. https://doi.org/10.3390/ijerph14091019 PMID: 28878150.

4. Ba NV, Minh HV, Quang LB, Chuyen NV, Ha BTT, Dai TQ, et al. Prevalence and correlates of multimorbidity among adults in border areas of the Central Highland Region of Vietnam, 2017. J Comorb. 2019; 9: 2235042X19853382. https://doi.org/10.1177/2235042X19853382 PMID: 31192142.

5. Carvalho JN, Cancela MC, Souza DLB. Lifestyle factors and high body mass index are associated with different multimorbidity clusters in the Brazilian population. PLoS One. 2018;20:13(11):e0207649. https://doi.org/10.1371/journal.pone.0207649 PMID: 30458026

6. Zhang R, Lu Y, Shi L, Zhang S, Chang F. Prevalence and patterns of multimorbidity among the elderly in China: a cross-sectional study using national survey data. BMJ Open 2019;9:e024268 https://doi.org/10.1136/bmjopen-2018-024268

7. Abad-Díez JM, Calderón-Larrañaga A, Poncel-Falcó A, Poblador-Plou B, Calderón-Meza JM, Sicras-Mainar A, et al. Age and gender differences in the prevalence and patterns of multimorbidity in the older population. BMC Geriatr. 2014;14:75. https://doi.org/10.1186/1471-2318-14-75 PMID: 24934411.

8. Aye SKK, Hlaing HH, Htay SS, Cumming R. Multimorbidity and health seeking behaviours among older people in Myanmar: A community survey. PLoS One. 2019;14(7):e0219543. https://doi.org/10.1371/journal.pone.0219543 PMID: 31295287.

9. Guerra SG, Berbiche D, Vasiliadis HM. Measuring multimorbidity in older adults: comparing different data sources. BMC Geriatr. 2019; 19(1):166. https://doi.org/10.1186/s12877-019-1173-4 PMID: 31200651.

10. Nunes BP, Filho ADPC, Pati S, Teixeira DSC, Flores TR, Camargo-Figuera FA, et al. Contextual and individual inequalities of multimorbidity in Brazilian adults: a cross-sectional national-based study. BMJ Open. 2017;7(6):e015885. https://doi.org/10.1136/bmjopen-2017-015885

11. Uranga RM, Keller, JN. The Complex Interactions Between Obesity, Metabolism and the Brain. Front Neurosci. 2019; 13: 513. https://doi.org/10.3389/fnins.2019.00513 PMID: 31178685.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

5 Nov 2019

Multimorbidity and complex multimorbidity in Brazilian rural workers

PONE-D-19-17072R1

Dear Dr. Salaroli,

We are pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it complies with all outstanding technical requirements.

Within one week, you will receive an e-mail containing information on the amendments required prior to publication. When all required modifications have been addressed, you will receive a formal acceptance letter and your manuscript will proceed to our production department and be scheduled for publication.

Shortly after the formal acceptance letter is sent, an invoice for payment will follow. To ensure an efficient production and billing process, please log into Editorial Manager at https://www.editorialmanager.com/pone/, click the "Update My Information" link at the top of the page, and update your user information. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, you must inform our press team as soon as possible and no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

With kind regards,

Siew Ann Cheong, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Authors have addressed all my comments. A very minor comment is that p value = 0.000 should be replaced with p value<0.0001

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Raffaele Palladino

Reviewer #2: No

7 Nov 2019

PONE-D-19-17072R1

Multimorbidity and complex multimorbidity in Brazilian rural workers

Dear Dr. Salaroli:

I am pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact. If they will be preparing press materials for this manuscript, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

For any other questions or concerns, please email plosone@plos.org.

Thank you for submitting your work to PLOS ONE.

With kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Siew Ann Cheong

Academic Editor

PLOS ONE