An Epidemiological Case Control Study Design of Musculoskeletal/ Joint disorders in Reindeer Herders in Mongolia


The reindeer herders of Mongolia known as the Tsaatan community are a diminishing community. Yet they have many medical problems including osteoarthritis which has many causative factors. The socio-economic situation of the community is suffering due to the illness. Reindeer are also slowly disappearing from the globe. This study seeks to identify the possible causative factors and epidemiology of the illness and how the community may be saved from future problems of the illness. The case study control design is being used here to discover the nature of the illness and whether the population is struck to despair. The problem of osteoarthritis is found in all the reindeer herders of the world existing in Sweden, Alaska and North America, and Russia.


Musculoskeletal and Joint Disease

Illnesses of the musculoskeletal system and joints have been a major occupational hazard in Russia and Mongolia and among the reindeer herding communities. An osteoarthritis is a group of degenerative illnesses characterized by painful mechanical movements of joints. Clinical symptoms include painful joints, stiffness of the joints, tenderness, inflammation, and creaking of joints. The degradation occurs in the cartilage and adjacent subchondral bone.

The cushioning effect of the cartilage between the two bones and the smooth gliding surface for normal joint functioning is lost (Rose and Kaye, 1983, p. 471). Relief of pain is the main objective of arthritis treatment. Medication or non-drug therapy is instituted and if there is no response, surgery is advised (Rose and Kaye, 1983, p. 472). Musculoskeletal illnesses involve the muscles, ligaments, and tendons. Pain fatigue and sleep disturbances are the major symptoms. The pain may be just an achy sensation or pain.

Among various etiologies like trauma, degenerative, immune-mediated (rheumatoid arthritis), congenital disorders, nutritional and metabolic (diabetes mellitus), infections are the main causes being studied in this project. Infections could be due to zoonotic pathogens leading to brucellosis, tick-transmitted illnesses like borreliosis, rickettsiosis, anaplasmosis, and osteomyelitis caused by various bacteria or viruses. Brucellosis is being considered in this study.

New Emerging Diseases

Tick transmitted diseases are re-emerging due to climate change, increasing pollution, alterations of habitat, increased surveillance, faster detection, more outdoor

professions and travel and recreation leading people to move far and wide. Human granulocytic anaplasmosis caused by Anaplasma phagocytophilum is one such organism which has adaptations and pathogenesis (Dumler, 2005, p. 1828). Anaplasmosis produces infection in humans but maintains a persistent subclinical state in animals.

Brucellosis is a zoonosis and occupational hazard which is important in Mongolia among the herders and reindeer. Serological surveillance is necessary for the control of the illness (Erdenebaatar, 2004, p. 571. Domestic and wild animals are infected equally. Economic loss in the form of reproductive losses has occurred in Mongolia in nomadic animal breeding. The placenta, fetus, fetal fluids and vaginal discharges from an infected animal are infective (CFSPH, 2007, p.2).

Venereal transmission is common except for some varieties of Brucella. Transmission can occur through fomites, water and food. All domestic animals are therefore vaccinated. Reindeer, Rangifer Tarandus, has a tendency to be infected by Brucella suis, type 4 (Biovar 4) (Rausch and Huntley, 1978, p. 129). Only some animals in experiments showed musculoskeletal symptoms. Most of them did not.

Humans become infected by ingesting organisms or contamination of mucous membranes and abraded skin. Infection can occur when humans come into contact with “animal abortion products, ingestion of unpasteurized dairy products from cows, small ruminants or camels; ingestion of undercooked meat, bone marrow or other uncooked meat products; contact with laboratory cultures and tissue samples; and accidental injection of live brucellosis vaccines” (CFSPH, 2007, p.2.). Brucellosis is a self limiting disease in humans but may be long drawn out.

Study rationale

The Tsaatan reindeer herder community of Mongolia has many illnesses which affect their socio-economic status and physical conditions. Efforts have been made to find out whether the occupation as such has influenced the physical health of the community by previous researchers in other herding communities in Sweden, the Sami herders, those in Alaska and North America. Very little research has been done in the Tsaatan community in Mongolia. This study is concentrating on this community and hoping to find a relationship between brucellosis exposure and osteoarthritis.

Research Hypothesis

The study aims to study the osteoarthritis seen in the reindeer herders among the Tsaatan community. Brucellosis, an emerging new infection, is also being investigated to decide whether it contributed to the musculoskeletal illness.

  • Null Hypothesis: Brucellosis has no role in the development of Osteoarthritis.
  • Alternative Hypothesis: Osteoarthritis of the Tsaatan community has occurred due to the Brucellosis.

Conceptual model

Conceptual model

Reference Population

220 individuals in 40-45 families of herders and around 600 reindeer are present here near the Ulaan Baatar area of Mongolia. 85 of the individuals are not associated with reindeer herding. These constitute the reference population. The Tsaatan communities here are facing various problems which include decreased reproduction, increasing musculoskeletal problems, and general unthriftiness in their reindeer. The reindeer are losing their previous health and having limb lameness in different combinations of limbs affected and joint effusion mainly in the carpals. Human beings too have similar problems. This study is considering the Tsaatan people.

Setting and sampling

76% of the 85 Sami herders in the Daerga study of 2003 were found to have work related musculoskeletal pains in 2001 in at least one region while 30 did not participate as they were not associated with reindeer herding. Two years later the ones with pain reported more pains. Based on this finding, the sample size in this study has been calculated after five persons dropped out after the study had started. The sample size of 40 each was selected for the two groups excluding the children. The groups had 28 men and 22 women each and the control group was formed of an individual comparable in age and sex to the one in the group of cases.

Study population

The sampling is from the Mongolian Tsaatan community, close to Ulaan Baatar, in a community health centre which caters to this community for its health requirements. The study population of 80 individuals has been given the questionnaires which decide about their osteoarthritis or musculoskeletal illness and the possibility of brucellosis in the course of the past one year on occasions of their visit for health care.

Their willingness was solicited through a letter and a filled and signed consent form. The reindeer herders who had no possibility of brucellosis or arthritis were also given the questionnaires so that they could be included in the control group. The completed ones have been collected. The questionnaire answers would determine the group to which the participant belongs. 40 participants have been selected for the group with illness and a matching 40 in age and gender have been selected as control group. 28 men and 22 women are in each group.


This study will be investigating the osteoarthritis which affects the reindeer herders in the Tsaatan community. The case control study design has been selected for keeping the study small and short in duration. The persons with previous illness would be identified and their previous exposure to brucellosis will be investigated. Cases are selected as to whether they have been affected by the osteoarthritis. The visual analogue scale was used to identify the cases. The people having a pain with a result of 5 and more on the scale were placed in the illness group and those with less than 1 were in the control group.


Exposure was defined as the state of having osteoarthritis following brucellosis infection.

Community field workers would be collecting the data or filled questionnaires (See Appendix A & B on Pg.14 & 16, Kuorinka, 1987). The people who collect will not be told the objective of the study. The questionnaires would be randomly distributed to 80 willing persons of the herder community in the year prior to the study. They would be filled by the herders with the help of the community workers in case of not fully understanding the language. The answered questionnaires would be divided into two groups according to whether the individuals had osteoarthritis or not. Each person in the ill group would be matched by a person in the control group of same sex and age.

Unit of analysis

The unit of analysis is the individual belonging to the herder community having arthritis and the control group would have the people without illness.

Sample Size

Sample size was calculated using Epi info according to an unmatched design, a level of significance of alpha=0.05 and power=0.80.

Sample Size

Prevalence of the exposure was estimated to be 12% in cases and 2% in control group, and an odds ratio of 3 was assumed to be biologically and programmatically meaningful.(i.e., subjects with high Brucellosis have three times the odds of being identified as an Osteoarthritis case).These settings yielded a sample size of 28 per group, to which we added 10% (3 subjects) for attrition as well as 3 subjects per covariate ( 3 covariates) for additional power in logistic regression analysis, yielding a final sample size of 40 per group.

Data Quality assurance

Only those who were herders continuously for the past two years were selected. Anyone who changed occupation or stopped working altogether was excluded. The women were mostly helping part time only in the occupation while four women were totally involved. They also did house work. The community was given health examination for the past year which included “anthropometrics, BMI, blood lipids, blood pressure, aerobic capacity, glucose load test, lung functions tests and bone mass measurements” (Daerga et al, 2003, p. 343). The base data was collected earlier during the one year periodic comprehensive medical examinations.

Statistical Methods

Analyses will be conducted using SAS version 9.0 with a level of significance of alpha=0.05. Differences in exposure outcomes and covariates among cases and controls will be assessed using chi-square test for our categorical variables. Unmatched (?)

conditional logistic regression analysis will be used to assess the relationship between Brucellosis (predictor) and Osteoarthritis cases. Adjusted Odds ratio will be derived from models including the listed covariates. Specifically, we will compute an odds ratio comparing the odds of exposure in osteoarthritis cases as compared to odds of exposure in controls.

Physical and psychosocial risk factor among the reindeer herders shown as Frequency values with Odds Ratio, separately for the ill group and control group.

Table 1.

Risk factors Case group 40 Control group-40 P value &
Frequency OR Frequency OR
Age ns
20-40> <0.001
20-40< <0.001
20< <40 <0.001
Sex <0.001
Male <0.05
Female ns
Climate Change ns
cold < 0.05
warm ns


A higher number of pain regions were complained of by the illness group. The number of painful days were significant. Pain intensity increased for the illness group. A history of infections was present in almost all the cases of osteoarthritis. The questions dealing with the psychosocial factors have been found to have reliability and validity previously. Self reported data on the risk factors were further complemented by interviews which provided added information.


Comparisons were made between the number of pain regions, intensity and duration and the types of exposure. Differences between dichotomous variables were explored using the chi-square test. One way ANOVA compared numerical variables. Possible associations between the musculoskeletal illness and the exposures using Pearson’s correlation coefficient were found. The null hypothesis would be rejected

as it cannot be finally concluded that brucellosis has no role in the development of osteoarthritis. The alternative hypothesis also would be rejected as other factors like the use of heavy vehicles and the difficulties experienced in the work-related atmosphere

raises the question of whether the brucellosis alone is the cause of the osteoarthritis.


Conditional logistic regression modeling will be used second to assess the magnitude and nature (additive, multiplicative, etc) of interaction. All potential interaction terms will be included in the preliminary maximum model and those terms found to be significant predictors of osteoarthritis will be retained.

Confounding Bias

Confounding will be assessed by logistic regression modeling. Estimated odds ratios will be calculated using both a crude model with only the predictor and response as well as an adjusted model that includes all covariates. If the difference between the estimated odds ratio derived from the crude model and the adjust model is greater than ?%, we will conclude that confounding is present and will accept the adjusted odds ratio as a more valid estimate of the true relationship between Brucellosis and osteoarthritis.


The possible bias in this study is Recall bias. In this study, if herders tend to forget whether they had infection or not during the time they fill out questionnaire then the exposure odds of subjects will not be representative for the source of population.


The study is being implemented in the right setting and in a population which is still rearing reindeer in a nomadic fashion. The researchers and community have a chance of familiarizing with each other for a better extraction of information. The confidence of the community members is built up from one year back and services are implemented for the health upkeep of the community. A similar study has been done in other reindeer herder populations but not in the Tsaatan population.


One limitation is that osteoarthritis has many causes all of which are not being considered here. An authoritative conclusion is thereby not possible. The reindeer herds are actually disappearing due to loss of forests and appropriate homeground for the reindeer. This study may thereby be limited in its usefulness for future considerations as the herder population would also diminish in numbers.


Brännström I,Weinehall L, Persson LÅ,Wall S. (1993). “Changing social pattern of risk factors for cardiovascular disease in a Swedish community intervention programme. Int J Epidemiol 1993; 22: 1026-1037.

CFSPH , Centre for Food Security and Public health , (2007). “Brucellosis”. Institute for International Cooperation in Animal Biologics. Web.

Daerga, L. et al. (2003). “Work-related musculoskeletal pain among reindeer herding Sami in Sweden – a pilot study on causes and prevention” Circumpolar Health, Nuuk.

Dumler, J.S. et al. (2005). “Human Granulocytic Anaplasmosis and Anaplasma phagocytophilum”. Emerging Infectious Diseases. Web.

Erdenebaatar, J. et al. (2004). “Epidemiological and serological survey of Brucellosis in Mongolia by EISA using sarcosine extracts”. Microbiological Immunology, Vol. 48 No. 8, Pg 571-577.

Karasek R, Theorell T. Healthy Work. (1990). “Stress, productivity and the reconstruction of working life”. New York: Basic Books Inc. Publishers.

Kuorinka I, Jonsson B, Kilbom Å et al. Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms.Appl Ergon 1987; 18: 233-237.

Rausch, R.L. and Huntley, B.E. (1978). “Brucellosis in reindeer, Ra~zgifer tnrandus L., inoculated experimentally with Brucella suis, type 4”. Can. Journal of Microbiology, Vol. 24, P.129-135. US Public Health Service.

Rose, L., & Kaye, D. (1983). “Osteoarthritis”. Internal medicine for dentistry. St. Louis, MO: C.V. Mosby.