**5. General considerations**

The studies discussed in the previous sections highlight the importance of specific metals as contaminants and additives in cosmetic products. To determine any particular relationships between formulations, a multi-variate meta-analysis was carried out using Spearman correlation and Principal Component Analysis, taking into account the maximum levels obtained for the various cosmetic formulations. Pearson correlation statistics (**Table 4**) reveal a relationship between all five metals (r > 0.466). Two latent factors had an eigenvalue greater than 1, which together explained 80.54% of the total variance. The factor loadings demonstrated the different groups of variables (**Figure 1**). Factor 1, displayed on the horizontal axis, weighed heavily on Pb, Cd, Ni and As with lipsticks, eye shadows, face paints, make-up foundation and skin


#### **Table 4.**

*Spearman correlation matrix for the five metals.*

#### **Figure 1.**

*Observations plot for the formulations. Legend: Lp = lipsticks; Es = eye shadows; Ep = eyepencils; Ma = mascaras; Mf = foundation; Fp = face paint; Fc = face cream; Tp = toothpaste; SLc = skin-lightening creams; Sb = sunblock; Hp = hair products; Hd = hair dyes; Bl = body lotion; Cl = cleansers; Lo = lotions; and Tc = tonic creams.*

lightening creams exhibiting high levels of these metals. These formulations were discriminately different from the rest. On the other hand, F2, displayed on the vertical axis, weighed heavily on Hg with skin lightening creams having superior quantities of this metal for the other formulations. This multi-variate analysis consolidates the findings from previous studies. The findings of such research works are at the disposition of authorities and policy makers for the formulation of high-quality cosmetic products.
