promoting the sustainable and responsible use of cobalt in all forms


Does cobalt cause cancer in humans?

Based on studies in humans, there is inadequate evidence to show any relationship between cobalt exposure and cancer in humans by any route of exposure (i.e. by oral intake, inhalation, or skin contact).

The literature contains mixed results from studies in workers in industries where cobalt exposure occurs: While some studies reported an excess risk of lung cancer[1], no definitive conclusion could be made concerning the potential carcinogenic effect of cobalt exposure because of several significant scientific concerns, such as low number of observed workers, unknown smoking habits and co-exposures to other chemicals. Furthermore, studies[2] which involved a longer follow-up period of a previously studied worker population, or adjustment for co-exposures, did not find any relationship between cobalt exposure and cancer. In addition, when observing patients with failed hip implants (metal-upon-metal wear of the implant which led to elevated levels of cobalt in the blood), no relationship was found between cobalt levels and cancer[3],[4]. Based on this body of evidence, the US National Toxicology Program (NTP) has stated in the final Report on Carcinogens (RoC, 2016) monograph on ‘cobalt and cobalt compounds that release cobalt ions in vivo’ that “the data from the human studies is inadequate[5] to evaluate the relationship between cobalt exposure and human cancer”[3].

Further epidemiology studies are due to be published in 2017. One is a follow up to a cohort that has been studied for cobalt-related asthma in 1980[6] and 2010[7], and the second is the largest study to date in the hard metal industry (in this case, tungsten carbide mixed with cobalt; used in the production of machine tools) with over 35,000 subjects.

Does cobalt cause cancer in animals?

Cobalt is an inhalation carcinogen in rats and mice.

Two chronic lifetime inhalation studies have determined that cobalt, at high exposure levels, can cause lung cancer in rodents. These studies were conducted by the US NTP using rats and mice to investigate inhalation exposure to cobalt sulphate heptahydrate and cobalt metal powder[8]. In both cases, lung cancer was attributed to cobalt exposure, as the number of tumours increased with increasing cobalt exposure levels.

Due to excessive exposure levels (cobalt metal study), and rat-specific sensitivities in both NTP studies, exposure-related effects (effects that occurred in the exposed animals, but were not cobalt-specific) were seen in addition to lung cancer. The adrenal tumours seen in addition to the lung tumours in rats are secondary effects caused by lung impairment and an overall lack of oxygen. It is important to note that comparable levels of cobalt in the lung cannot generally be achieved by inhalation exposure in humans, since the NTP used very high exposure levels combined with small particle sizes in order to maximise exposure to the rodent lung. Another effect that was not cobalt specific was leukaemia, observed in female rats only. Leukaemia occurs at a high spontaneous rate in the type of rats used in these studies (i.e. Fischer rats)[9]. There were also kidney and pancreas tumours observed only in one species (rat), in one sex (male) in the Co metal NTP study, however these tumours did not occur to a significant degree and were seen in a rat strain (i.e. F344/NTac) suffering from various health impairments for which an NTP historical database is not available.

All the available evidence points to a threshold mode of action for cobalt-related lung cancer in animals.

In the 14th Report on Carcinogens (RoC, 2016), the NTP stated that “cobalt and cobalt compounds that release cobalt ions in vivo are reasonably anticipated to be human carcinogens based on sufficient evidence from studies in experimental animals and supporting mechanistic data”.

All the available evidence points to a threshold mode of action for cobalt-related lung cancer.

Cobalt-related lung cancer is a multi-factorial (many causes) disease, involving a variety of mechanisms that, when the exposure threshold is exceeded, contribute in a pathway towards cancer. Cobalt is a substance for which the ‘mechanisms of action’ (defined biological pathways activated by a substance) are documented in the scientific literature. Cobalt acts predominantly by the following mechanisms:

This describes cobalt as leading to cancer through an ‘indirect’ pathway, which does not involve direct interaction with DNA. These mechanisms of action are known to have biological thresholds, meaning they already occur to some degree at a background level in the body and need to surpass a certain level before an adverse effect is seen. In practical terms, this means that a threshold of exposure to cobalt needs to be exceeded before the pathway leading to cancer is initiated.

There is no evidence that cobalt can cause cancer in animals through the oral or dermal routes of exposure.

There is a small amount of cobalt in a normal human and animal diet. Indeed, inorganic cobalt salts and cobalt as part of vitamin B12 are added to animal diets for general health and well-being.

There have been no oral cancer studies investigating cobalt exposure in animals. In a 90-day repeated-dose oral toxicity study in rats with a very soluble form of cobalt, no tumours were discovered. Instead, what was observed was the ‘hallmark’ of cobalt exposure, an increase in red blood cells due to a cobalt-induced hypoxia-like response.

In a study in which the skin surface barrier was bypassed (i.e. subcutaneous injection of a very soluble cobalt substance) tumours were seen in the injection sites[10]. Subcutaneous injection is not a route of exposure relevant to human health. Humans are primarily exposed to cobalt orally, dermally and through inhalation. In a laboratory human skin study conducted by the CI, the most soluble form of cobalt was applied to the skin surface in both small and large amounts to investigate skin penetration. In both cases, results showed that less than 1% of the applied cobalt dose was found to have penetrated into the skin or through the skin (absorbed into the blood).

Carcinogenicity hazard classifications

International Agency for Research on Cancer (IARC)

The IARC is an international advisory body that is part of the World Health Organisation (WHO). IARC evaluates both epidemiological and laboratory evidence when looking into the causes of human cancer.

The IARC has evaluated four groups of cobalt compounds[11] (‘cobalt and cobalt compounds’, cobalt metal with tungsten carbide (not covered by the CI), cobalt metal without tungsten carbide and cobalt sulphate with other soluble cobalt salts), and classifies them as follows:

Agent CAS No. IARC Classification
Cobalt and cobalt compounds (1990) Various 2B
“Possibly carcinogenic to humans”
Cobalt metal with tungsten carbide - 2A
“Probably carcinogenic to humans”
Cobalt metal without tungsten carbide (2006) 7440-48-4 2B
“Possibly carcinogenic to humans”
Cobalt sulphate and other soluble cobalt (II) salts (2006) Various 2B
“Possibly carcinogenic to humans”

UN Globally Harmonized System of Classification and Labelling (UN GHS)

Based on the results of the US NTP inhalation carcinogenicity studies described above, cobalt sulphate and cobalt metal, and any analogous cobalt compounds exhibiting the same in vivo behaviour, meet the criteria for classification under UN GHS as carcinogenic by inhalation (Carc. 1B; H350i). Since 2008, the European Union (EU) has adopted a harmonised classification under the Dangerous Substances Directive (67/548/EEC) and its successor, the Classification, Labelling and Packaging (CLP) Regulation, for carcinogenicity of cobalt sulphate, cobalt dichloride, cobalt carbonate, cobalt di(acetate) and cobalt dinitrate:

Substance EC No. EU CLP Classification
Cobalt sulphate 233-334-2 Harmonised Classification Carc.1B; H350i (inhalation route only)
‘Presumed to have carcinogenic potential for humans’
Cobalt dichloride 231-589-4
Cobalt carbonate 208-169-4
Cobalt di(acetate) 200-755-8
Cobalt dinitrate 233-402-1

In addition to the harmonised classification for the five cobalt soluble salts, cobalt metal has been self-classified by the cobalt industry since December 2013, and cobalt monoxide since November 2016, according to the UN GHS criteria for carcinogenicity:

Substance EC No. UN GHS Self-Classification
Cobalt metal (all physical forms) 231-158-0 Self-Classification Carc.1B; H350i (inhalation route only)
‘Presumed to have carcinogenic potential for humans’
Cobalt monoxide 215-154-6

Please see the self-classification statements on cobalt metal and cobalt monoxide for further information.


This summary is intended to provide general information about the topic under consideration. It does not constitute a complete or comprehensive analysis, and reflects the state of knowledge and information at the time of its preparation. This summary should not be relied upon to treat or address health, environmental, or other conditions.

[1] Mur et al 1987, Tüchsen et al 1996, Grimsrud et al 2005, Moulin et al 1998, Wild et al 2000
[2] Moulin et al 1993, Moulin et al 2000, Grimsrud et al 2005
[4] Christian et al 2014, Atrey et al 2016, Ayoub et al 2016
[5] Inadequate = there is low or very low confidence in the body of evidence for an association between exposure to the substance and the health outcome(s) or no data are available (definition from “Draft NTP Approach for Consideration at August 28-29, 2012 BSC Working Group Meeting”)
[6] Roto et al 1980
[7] Sauni et al 2010
[8] NTP TR 471 (cobalt sulfate) and NTP TR 581 (cobalt metal)
[9] Suckow, Weisbroth et al. 2006
[10] Shabaan, A.A., Marks, V., Lancaster, M.C., Dufeu, G.N. Fibrosarcomas induced by cobalt chloride (CoCl2) in rats. Lab Anim 11(1): 43-46. 1976.
[11]World Health Organisation International Agency for Research on Cancer. Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 52. Chlorinated Drinking-Water; Chlorinated By-Products, Some other Halogenated Compounds, Cobalt and Cobalt Compounds. France, 1990.
World Health Organisation International Agency for Research on Cancer. Monographs on the Evaluation of Carcinogenic Risks to Humans, Volume 86. Cobalt in Hard Metals and Cobalt Sulfate, Gallium arsenide, Indium Phosphide and Vanadium Pentoxide.France, 2006.

Photo credit: Eric Snyder, Huntsman Cancer Institute at the Univ. of Utah