Concentrations and deficiencies of minerals in cattle submitted to a diagnostic laboratory in Saskatchewan from 2003–2012: A retrospective study

Abstract

Trace mineral analyses of samples submitted to Prairie Diagnostic Services laboratory from Saskatchewan cattle between 2003 and 2012 were examined, with the objective of describing trends and reporting concentrations and deficiencies of minerals. Deficiencies were observed with copper, iron, manganese, magnesium, zinc, and cobalt. Deficiency was most commonly seen in copper, followed by iron, manganese, and magnesium accounting for 47.2%, 15.1%, 13.0%, and 10.8% of deficiencies, respectively. Deficiency in cobalt was least common followed by zinc accounting for 4.2% and 9.7% of deficiencies, respectively. The following minerals were also analyzed: barium, beryllium, bismuth, cadmium, chromium, antimony, tin, molybdenum, strontium, thallium, and vanadium. Submissions from 1434 animals were reviewed and a diagnosis of mineral deficiency was made for 509 animals with 92 of these having multiple deficiencies. There were significant differences in the number of deficient animals by year (P = 0.001), age group (P = 0.01), but not month (P = 0.109) or soil type (P = 0.172).

Introduction

Mineral deficiencies are a pervasive challenge for livestock producers and veterinarians. The non-specific or vague clinical manifestations often allow these problems to go undiagnosed until losses are encountered. Lower concentrations of trace minerals and lack of mineral or concentrate supplementation are associated with increased pathologies at the herd level compared with healthy dairy and beef herds (1). Subclinical deficiencies impair reproductive and immune system function and are often manifested as infertility or a predisposition to disease (2–5). Barriers to a diagnosis are related to not only the subtle nature of clinical signs but also to the cost and invasiveness of sampling for certain minerals. For example, the liver is the preferred sample for diagnosis of copper (Cu) deficiency because in the blood it is maintained at a constant concentration through depletion of the liver stores (6). Published reference ranges are scarce, or often inconsistent, and frequently have overlapping values for what is considered deficient, marginal, normal, or toxic (6,7).

Essential elements are those which can reliably produce clinical symptoms when they are deficient. Non-essential elements play important biological roles often as cofactors; however, when they are deficient other elements can substitute for them. Few non-essential elements with no identifiable roles are found in tissues (8,9). Understanding the variation in non-essential trace minerals is still important in providing a comparison to real world exposure, and with new and better understanding of the biological roles of some elements, their classification as non-essential trace minerals may change (10). In Saskatchewan, the following trace minerals have been evaluated: Cu, iron (Fe), manganese (Mn), selenium (Se), and zinc (Zn) with Cu deficiency being recognized since the 1950s and in 1989 the magnitude of hypocuprosis was described as extensive (11–13).

The objectives of this study were to characterize the trends in trace mineral deficiencies in Saskatchewan cattle samples submitted to Prairie Diagnostic Services from 2003 to 2012 and to report the ranges of non-essential trace minerals in these samples.

Materials and methods

The trace mineral analyses used in the study were obtained from submissions to Prairie Diagnostic Services, Saskatoon, Saskatchewan, from 2003 to 2012. A computer search for all submitted samples identified 1434 animals that were tested. A submission occasionally contained multiple samples each belonging to a different animal. For each submission, the clinical history was reviewed, and the purpose of the submission was reported. Mineral concentrations were evaluated on a wet-weight basis for serum, plasma, liver, and kidney using inductively coupled plasma mass spectrometry (14). Published reference ranges for the trace minerals were used to designate deficient and normal concentrations for cobalt (Co), Cu, Fe, Mn, magnesium (Mg), nickel (Ni), and Zn (7). Selenium was not included in the study as this major biologically significant and economically important element will be considered in association with vitamin E in a subsequent manuscript. Tissue reference ranges for liver and kidney were used for Co, Cu, Fe, Mg, and Zn but limited to kidney for Ni and to liver for Mn. Reference values were also available for serum for Cu, Fe, Mg, and Zn. Each mineral status was classified as either deficient or normal. Occasionally, reported reference ranges for “deficient” and “normal” overlapped and a mineral was classified as deficient if its concentration was lower than the lowest value reported in the reference range. When a marginal range was also reported for a mineral and its concentration overlapped with the reported deficient range, deficiency was assigned when the mineral concentration was lower than or equal to the highest value reported in the deficient range. The limits of detection for Co, Cu, Fe, Mn, Mg, Ni, and Zn were 1.35, 36.0, 0.90, 3.00, 2.65, 17.0, and 360 ng/kg on a wet matter basis, respectively. The geometric mean and range of Mn concentrations in fetal livers were reported with other elements (see below) as they are usually 60% to 80% of maternal values, and no reference ranges have been established for this age group (15).

The clinical history, location, age, date, and year of submission were considered. Locations were assigned soil types based on the chernozemic soil classification of the rural municipality (16). If more than one soil type was present in a rural municipality, which occurred infrequently, the predominant soil classification was assigned. The ages included fetal, newborn < 7 d postpartum, calf 7 to 364 d, and adult ≥ 365 d. Copper was the only element for which there was a deficiency category for fetal liver, and therefore was the only element with a different deficiency designation based on age (7). Liver concentrations for the following elements were also reported: barium (Ba), beryllium (Be), bismuth (Bi), cadmium (Cd), chromium (Cr), molybdenum (Mo), antimony (Sb), tin (Sn), strontium (Sr), titanium (Ti), and vanadium (V).

IBM SPSS Statistics (version 24.0; IBM, Armonk, New York, USA) was used to perform Pearson’s Chi-square test for independence to determine if there were any significant associations between the binomial outcome of deficiency and year, month, age, and soil type of submissions. The contribution of categorical variables from significant Chi-square tests was further evaluated by a Z-test of column proportion with Bonferroni correction for multiple comparisons.

Results

From the 1434 cases submitted for trace mineral analyses there were 509 cases considered deficient with 92 cases containing multiple deficiencies (range: 2 to 5 minerals), resulting in a total of 638 deficiencies. The purpose for the submissions is listed in Table 1, with abortions, deaths, downer cows, neonatal losses, calf scours, weak calves, and pneumonia or cough comprising about 50% of the submissions. The most common deficiency was Cu (47.2%), followed by Fe (15.1%), Mn (13.0%), and Mg (10.8%). Zinc and Co deficiencies represented 13.9% of the total with no deficiencies reported for Ni. Most submissions having a deficient status were received between February and April (Table 2). The largest number of deficiencies occurred in March, but the highest percentage was in February, during which 42.8% of the submissions had a mineral deficiency. The observed cases of deficiencies recorded each month were not significantly different from expected values (P = 0.109).

Table 1.

Reason for sample submission for trace mineral analysis.

	N	Submissions
Abortion	260	198
Deaths	145	138
Downer	79	65
Neonatal losses	62	55
Calf scours	61	43
Weak calves	55	43
Pneumonia/cough	57	39
Neurological	46	38
Loss of body condition	55	35
Stillbirth	23	23
Non-specific	27	19
Herd survey/healthy	91	18
Lameness	19	14
Diarrhea	18	12
Open cows	47	10
Toxicitya	16	10
Premature calves	10	8
Vaccine reaction	40	7
Coat	13	7
Loss of production (milk)	9	5
Recheck	16	4
“Poor doer” calf	6	4
Hind-end weakness	4	4
Anestrus	25	3
Pica	8	1
LDA/metritis/mastitis	7	1
Miscellaneous (each < 3 cases)	37	35
Unknown	198

N — refers to the number of individual clinical histories evaluated (a submission may have samples from different animals in the same herd).

^aCauses of toxicity include lead, monensin, mycotoxin, nitrate, copper, and mineral mix.

Table 2.

Mineral deficiencies by month of submission of samples from Saskatchewan cattle between 2003 and 2012 from the Prairie Diagnostic Services database.

Mineral	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sept	Oct	Nov	Dec	Total
Copper	24	40	49	58	31	20	14	8	14	7	15	21	301
Iron	7	6	27	14	7	11	2	1	5	2	5	9	96
Manganese	3	14	20	25	8	2	5	1	—	—	—	5	83
Magnesium	8	14	21	10	3	4	—	1	—	—	—	8	69
Zinc	4	10	12	10	5	7	3	1	1	3	—	6	62
Cobalt	3	15	5	2	—	1	1	—	—	—	—	—	27
Nickel	—	—	—	—	—	—	—	—	—	—	—	—	0
Total number of mineral deficiencies	49	99	134	119	54	45	25	12	20	12	20	49	638
N	127	182	292	261	158	73	50	26	62	45	64	94	1434
Number of animals deficient	38	78	117	96	46	29	18	9	17	12	20	29	509
Percentage of animals deficient	29.9a	42.8	40.1	36.8	29.1	39.7	36.0	34.6	27.4	26.7	31.3	30.9	35.5

N — total number of samples tested. Only 1 sample was tested from each animal.

Reference values for deficiencies adapted from (7) are: copper < 3, 10, 20, and 0.5 mg/kg for kidney, liver, fetal liver, and plasma or serum, respectively; iron < 20, 30, and 1.3 mg/kg for kidney, liver, and plasma or serum, respectively; manganese liver values < 1 mg/kg; zinc < 18, 25, and 0.5 mg/kg for kidney, liver and serum, respectively; magnesium < 50, 100, and 11 mg/kg for kidney, liver, and serum or plasma respectively; cobalt < 0.014 and 0.005 mg/kg for kidney and liver, respectively; nickel < 0.01 mg/kg for kidney (wet weight). Pearson’s Chi-square test for the number of deficiencies by month was P = 0.109.

^aPercentage of deficient animals = number of animals deficient/N.

The greatest number of deficiencies occurred in 2009, 142 cases, but the highest proportion of diagnosed deficiencies was in 2010 with 47.1% of cases considered deficient (Table 3). The difference in deficiencies by year was significant (P = 0.001); with 2003 having a significantly lower number of deficiencies (P = 0.002) than 2010. There were no significant differences among the other years. There was a significant difference from the expected values for deficiencies based on the age of the animal (P = 0.01). Adult cattle had the greatest number and percentage of deficiencies (247 cases and 42.0% of submissions, Table 4). The number of deficiencies for adult cattle was significantly greater than for calves (P = 0.01) using a Z-test with Bonferroni correction.

Table 3.

Mineral deficiencies by year identified in Saskatchewan cattle between 2003 and 2012 from the Prairie Diagnostic Services database.

Mineral	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	Total
Copper	4	9	6	32	15	37	53	43	73	29	301
Iron	4	5	2	3	4	15	8	18	32	5	96
Manganese	1	2	1	8	8	13	26	7	12	5	83
Magnesium	—	6	—	2	2	4	24	2	15	14	69
Zinc	5	5	1	4	3	10	12	8	6	8	62
Cobalt	—	—	—	1	—	2	19	1	1	3	27
Nickel	—	—	—	—	—	—	—	—	—	—	0
Total number of mineral deficiencies	14	27	10	50	32	81	142	79	139	64	638
N	56	47	29	148	85	148	334	136	293	158	1434
Number of animals deficient	12	10	9	42	28	62	112	64	119	51	509
Percentage of animals deficient	21.4	21.3	31.0	28.4	32.9	41.9	33.5	47.1	40.6	32.3	35.5

N — total number of samples tested. Only 1 sample was tested from each animal.

Reference values for deficiencies adapted from (7) are: copper < 3, 10, 20, and 0.5 mg/kg for kidney, liver, fetal liver, and plasma or serum, respectively; iron < 20, 30, and 1.3 mg/kg for kidney, liver, and plasma or serum, respectively; manganese liver values < 1 mg/kg; zinc < 18, 25, and 0.5 mg/kg for kidney, liver and serum, respectively; magnesium < 50, 100, and 11 mg/kg for kidney, liver, and serum or plasma, respectively; cobalt < 0.014 and 0.005 mg/kg for kidney and liver, respectively; nickel < 0.01 mg/kg for kidney (wet weight). Pearson’s Chi-square test for the occurrence of deficiencies by year P = 0.001 and P = 0.002 for the Z-test of column proportions for 2003 when compared to 2010, with Bonferroni correction.

Table 4.

Mineral deficiencies by age group identified in Saskatchewan cattle between 2003 and 2012 from the Prairie Diagnostic Services database.

Mineral	Fetusa	Newborn	Calf	Adult	Total
Copper	42	8	62	135	247
Manganese	—	37	23	20	80
Iron	7	7	14	47	75
Magnesium	46	2	4	15	67
Zinc	12	5	11	28	56
Cobalt	21	1	3	2	27
Nickel	—	—	—	—	0
Total number of mineral deficiencies	128	60	117	247	552
N	270	140	325	448	1181
Number of animal deficiencies	95	53	98	188	437
Percentage of animal deficiencies	35.2	37.9	30.2	42.0	37.0

N — total number of samples tested. Only 1 sample was tested from each animal.

Reference values for deficiencies adapted from (7) are: copper < 3, 10, 20, and 0.5 mg/kg for kidney, liver, fetal liver, and plasma or serum, respectively; iron < 20, 30, and 1.3 mg/kg for kidney, liver, and plasma or serum, respectively; manganese liver values < 1 mg/kg; zinc < 18, 25, and 0.5 mg/kg for kidney, liver and serum, respectively; magnesium < 50, 100, and 11 mg/kg for kidney, liver, and serum or plasma, respectively; cobalt < 0.014 and 0.005 mg/kg for kidney and liver, respectively; nickel < 0.01 mg/kg for kidney (wet weight). Pearson’s Chi-square test for the number of deficiencies by age group P = 0.01, Z-test of column proportions indicate that adults are significantly different from calves, P = 0.009 with Bonferroni correction.

^aFetus <0 days, newborn < 7 days, calf 7 to 364 days, adult > 364 days.

The highest proportion of mineral deficiencies by soil type occurred in dark gray soil zones, 42.9%, with the greatest number of deficiencies from the dark brown soil zone (Table 5). The observed values were not significantly different from the expected values based on the soil zone of the submission (P = 0.17). There was also no significant difference in the number of Cu deficiencies in relation to soil type (P = 0.53).

Table 5.

Mineral deficiencies by chernozemic soil zone identified in Saskatchewan cattle between 2003–2012 from the Prairie Diagnostic Services database.

Mineral	Dark gray	Black	Dark brown	Brown	Gray	Total
Copper	21	93	103	61	23	301
Iron	9	26	43	10	8	96
Manganese	9	27	19	25	3	83
Magnesium	9	16	20	18	6	69
Zinc	9	17	13	19	4	62
Cobalt	2	9	6	8	2	27
Nickel	—	—	—	—	—	0
Total number of mineral deficiencies	59	188	204	141	46	638
N	77	413	449	378	117	1434
Number of animals deficient	33	151	169	120	36	509
Percentage of animal deficiencies	42.9	36.6	37.6	31.7	30.8	35.5

N — total number of samples tested. Only 1 sample was tested from each animal.

Reference values for deficiencies adapted from (7) are: copper < 3, 10, 20, and 0.5 mg/kg for kidney, liver, fetal liver, and plasma or serum, respectively; iron < 20, 30, and 1.3 mg/kg for kidney, liver, and plasma or serum, respectively; manganese liver values < 1 mg/kg; zinc < 18, 25, and 0.5 mg/kg for kidney, liver and serum, respectively; magnesium < 50, 100, and 11 mg/kg for kidney, liver, and serum or plasma, respectively; cobalt < 0.014 and 0.005 mg/kg for kidney and liver, respectively; nickel < 0.01 mg/kg for kidney (wet weight). Pearson’s Chi-square test for the occurrence of deficiencies by soil type was not significant, P = 0.172 and P = 0.053 for Cu deficiencies.

Geometric mean, 95% confidence interval (CI), standard deviation (SD), and range of liver values for antimony, barium, beryllium, bismuth, cadmium, chromium, fetal manganese, molybdenum, strontium, tin, thallium, and vanadium are summarized in Table 6.

Table 6.

Reported trace liver mineral values for Ba, Be, Bi, Cd, Cr, Mn, Mo, Sb, Sn, Sr, Tl, and V measured in μg/kg wet weight in Saskatchewan cattle identified between 2003 and 2012 from the Prairie Diagnostic Services database.

Mineral	Geometric mean	Standard deviation	Range		95% CI		N
			min	max
Barium	42.1	122	0.82	1046	32.4	54.6	78
Beryllium	1.03	13.65	0.02	60	0.645	1.64	54
Bismuth	0.418	2.22	0.001	10.7	0.214	0.818	26
Cadmium	20.2	94.4	0.14	490	13.7	29.7	88
Chromium	76.8	121	9.72	747	64.1	92	90
Manganesea	709	416	70	2100	656	766	237
Molybdenum	427	442	6	3470	403	452	774
Antimony	2.25	33.3	0.06	268	1.57	3.23	73
Tin	7.92	17.7	0.27	106	6.18	10.1	71
Strontium	82.3	77.9	4.9	454	71.1	95.2	89
Thallium	0.36	1.25	0.058	8.09	0.281	0.472	69
Vanadium	44.1	116	0.26	600	30.5	63.8	71

^aFetal liver concentrations.

Discussion

The observed monthly differences in mineral deficiencies likely reflect the seasonal changes in the beef industry and most common clinical manifestations of deficiency, such as abortion given the observed trends. Submissions are highest in March and April with 134 and 119 cases, respectively, and the greatest proportion of deficiencies occur in February, March, and June. Late-term abortions or weaker calves likely prompt testing within the herd during February and March. Whereas early-term abortions or increased number of open cows would be overlooked between November and January. An alternative to increased recognition is increased manifestation of deficiencies due to the quality of stored feeds or increased energy demands, such as inclement weather or advanced gestation during these months (17). The greater percentage in June likely reflects individual testing of animals showing distinct clinical disease.

For wild deer, the severity of winters, specifically the temperature in early winter, is associated with poorer conditions and increased mortalities (18). Despite this association between body condition, physiologic adaptations, and ambient temperature, there is no published information on the effects of ambient temperature on trace mineral status (19). A summary of the severity of Saskatchewan winters was reviewed, but increased deficiencies did not coincide with severe winters (20). The significantly greater number of deficiencies in 2010 compared to 2003 could also not be explained by crop quality, which was poor in 2002 to 2003 coincident with a smaller proportion of deficiencies; whereas, in 2010 the crop quality was high with a higher proportion of deficiencies (21). Our data set did not include the source of feedstuffs and therefore limited conclusions can be made. Year 2003 was the year bovine spongiform encephalopathy was first diagnosed in Canada and may have had a negative impact on diagnostic submissions (22). The diagnostic submissions were low in the following years and increased after 2005, when the United States lifted the trade ban (22), but there were no significant differences between these years. The highest number of submissions occurred in 2009 with 336 followed by 2011 with 293. Economic factors over this period that may have influenced the submission of diagnostic samples include an economic recession, country of origin labelling, strong Canadian dollar, and increasing feed prices (23). A limitation of the study was in dichotomizing the outcome as normal or deficient instead of treating each element separately. There could be interactions amongst these minerals that could affect the significance of the association. This method was chosen to prevent repeated statistical testing on the data set, which would inflate the type 1 error. Chi-square analysis was chosen over logistic regression as the purpose of this study was to describe trends and not to predict deficiency. Logistic regression may offer more complexity but essentially showed similar results when it was performed.

The relationship of soil type and mineral content has been demonstrated in Saskatchewan for Cu and Mo but not Fe, Zn, or Mn (12,24). This is inconsistent with our findings, in which there is no significant difference in Cu deficiencies associated with the soil types. The greatest percentage of deficiencies occurred in the dark gray soil zone which also had the largest percentage of Cu deficiencies. These dark gray and muskeg soil zones have lower concentrations of Cu and higher concentrations of Mo, which antagonizes Cu absorption, and may explain this difference, although it was not significant (24). Reported liver values from Saskatchewan cattle from the dark gray and muskeg soil zones were higher in Fe and lower in Cu content (13,24). This is explained by Fe accumulation in the liver with decreased Cu concentrations, which are low in the aforementioned soil zones (24). These soil classifications are chernozemic and are based on the organic content of the surface horizon and the mineral content in the organic layer which is released from the clay components (25–27). The lack of association could be attributed to the soil classification system used. Alternatively, the type of soil, source of feedstuffs, use of mineral supplements, use of fertilizers, or management strategies may confound our results.

Barium has not been shown to be involved in any biochemical pathways in the body and is of toxicological concern. The free ion and barium chloride are highly toxic, whereas the environmental forms, barium carbonate and barium sulphate, are much less so (7). The primary sources for the toxic forms are associated with industrial chemicals (28). Liver concentrations are considered elevated between 10 to 21 400 μg/kg (14). The geometric mean for laboratory samples of Saskatchewan cattle is within this range at 42.1 μg/kg (Table 6). This is unexpected considering Saskatchewan does not contain any deposits of barite which is the primary environmental source of barium (29). The main industrial chemical sources are insecticides, pesticides, rodent poisons, lubricants, and rubber (29).

The geometric mean Cd liver concentrations from laboratory samples of Saskatchewan cattle was 20.2 μg/kg (Table 6), which falls within the normal reference range of 20 to 1000 μg/kg (7). Previous reports of Cd liver concentrations in Canadian cattle did not indicate a mean as most samples were below the limit of detection, 40 to 300 μg/kg, and our current mean falls below this range as well (30). The main sources of Cd as well as V are industrial contamination with exposure occurring by ingestion of contaminated food or water (12). The geometric mean and range for V, 44.1 μg/kg (range: 0.26 to 600 μg/kg) in liver, was well above the normal range for liver, 6 to 7 μg/kg (7). The only published comparison is from a test herd in South Africa that was divided into high and low environmental exposure groups, median of 1340 μg/kg (range: 330 to 11 500 μg/kg) (31). The geometric mean would, from our study, be above normal but below toxic levels (7). Vanadium accumulates in the liver and signs of toxicity include diarrhea, inappetence, dehydration, emaciation, dry hair coat, and inability to rise (7). Tolerance can develop to greater exposure and these increased values may not correlate well with clinical manifestations (7). Mineral concentrations that are higher than the normal reference range for Ba and V could indicate an increase in exposure, alternatively it could reflect elemental interactions, or sampling bias in our study. The geometric mean concentrations for Mo, Sn, Sr, Cr, and Tl 427, 7.92, 82.3, 76.8, and 0.36 μg/kg, were comparable to the previously reported normal reference ranges of 140 to 14 000, < 600, 190, 40 to 3800, and < 50 μg/kg, respectively (7). The geometric mean concentration for fetal liver Mn was 709 μg/kg. This is comparable to a previous study in which the average liver value of aborted fetuses from Saskatchewan cattle was 870 μg/kg (14). There were no comparisons in the literature or reference ranges listed for cattle for Be, Bi, and Sb. The minerals in Table 6, excluding Mn, exhibited large ranges and standard deviations. This is expected for these minerals as they reflect environmental exposure and generally have a log normal distribution (7). This was accounted for in the use of geometric means, confidence interval, and standard deviation (8). It is important to note that the percentage of deficiencies reported in this study do not represent deficiencies in the actual population but are drawn from the cases submitted for mineral analysis during the evaluated period.

The variable reference ranges and inconsistent reporting make interpretation of mineral concentrations and comparisons challenging, limitations that have also been encountered by other researchers (32). For non-essential minerals, there is limited published data and further investigation is warranted, particularly for Ba and V in Saskatchewan cattle. Additionally, a better understanding of the kinetics of mineral distribution and use, particularly in relation to weather extremes, would be beneficial. Further investigation of the trends in mineral deficiencies and their economic impact should include consideration of breed, mineral supplementation, feedstuff, and use of fertilizers.