28.3: Serum zinc (24c.3)

Serum and plasma zinc concen­trations can be used interchangeably because differ­ences between these two sources of circulating zinc are very small (English and Hambidge, 1988). Fasting serum zinc concen­trations are homeo­statically controlled within fairly narrow limits (about80–100µg/dL (12–15µmol/L). Only 0.1% of the total body zinc is present in the serum, whereas 3% and 63% are in the liver and skeletal muscle, respectively (Table 24c.1). Of the zinc in whole blood, from 12% to 22% is in the serum; the remainder is within the erythrocytes. Zinc is trans­ported in the serum bound principally to albumin (70%), so condi­tions that alter serum albumin levels will, in turn, affect serum zinc concen­trations. The remaining zinc in serum is tightly bound to α₂‑macroglobulin (18%), and the rest is bound to other proteins such as trans­ferrin and ceruloplasmin and to amino acids, primarily histidine and cysteine (Cousins, 1985).

Serum zinc concen­tration is considered the best available biomarker of zinc exposure, status, and risk of zinc defi­ciency at the popu­lation level(de Benoist et al., 2007; Lowe et al., 2009; Hess et al., 2007). Evidence for this recom­men­dation is derived from investigations on the following: (1) the effect of dietary zinc restrictions and repletion on serum zinc concen­trations; (2) the effects of zinc supple­mentation on serum zinc; (3) the association between serum zinc and clinical signs of zinc defi­ciency; and (4) a com­parison of initial serum zinc concen­trations between individuals who do or do not show a functional response to the correction of zinc defi­ciency (King et al., 2015).

In brief, in severe dietary zinc restriction (i.e., < 1mg/d) in previously healthy adults, there is a sharp decrease in serum zinc concen­trations, with values returning to baseline levels within 1–2 weeks of zinc supple­mentation Figure 24c.5.

Figure 24c.5. Changes in plasma zinc (mean±SD) at the end of basal, depletion, and repletion periods. The shaded area represents the normal range. From Baer and King(1984).

In more moderate dietary zinc restrictions (3–5mg/d),however, reductions in serum zinc only occur after prolonged periods of a restricted diet or when accom­panied by high phytate intakes (Gibson et al., 2008).

Supplemental zinc consistently and rapidly increases serum zinc in both children and adults regardless of initial serum zinc concen­trations (Hess et al., 2007; Wessells et al., 2020). Of interest is the finding that serum zinc responds less to additional zinc provided in food than to additional zinc supplied as a supple­ment administered between meals. It is possible that absorbed zinc may be meta­bolized differ­ently when consumed in food or as a supple­ment (King et al., 2015).

Low serum/plasma zinc has been related to clinical signs of zinc defi­ciency based on data from experimental zinc-depletion / repletion studies and case reports of acrodermatitis enteropathica. From these data, the sensitivity and specificity of a plasma zinc below 50µg/dL (7.65µmol/L) for detecting clinical signs of defi­ciency was 82% and 92%, respectively (Wessells et al., 2014).

Finally, if the initial serum zinc is very low, a functional response to zinc supple­mentation may occur. Functional responses may include increases in growth (including linear growth, weight gain, and accrual of fat-free mass) and decreased morbidity from diarrhea and respiratory infection. Of these responses, IZiNCG has selected linear growth outcom­e as the preferred functional response based on height- or length-for‑age (de Benoist et al., 2007).

Serum or plasma zinc has been used to assess the risk of zinc defi­ciency in several national surveys. The popu­lation is considered to have an elevated risk of zinc defi­ciency of public health importance when a particular popu­lation has more than 20% of individuals with a serum zinc concen­tration below the age and sex-specific cutoffs. In a review of national surveys in twenty low‑ or middle-incom­e countries the prevalence of low plasma zinc was more than 20% in 68% of the children and 93% of the women (Hess, 2017), com­pared to the US NHANES 2011–2014 survey where only 3.8% of children (6– < 10y) and 8.2% of females (≥ 10y) had low serum zinc concen­trations (Hennigar et al., 2018).

In general, national surveys (Gibson et al., 2011; Engle-Stone et al., 2014; Hennigar et al., 2018) have not shown a positive relationship between individual dietary zinc intakes and serum zinc concen­trations. Meta-analyses conducted on adults (Lowe et al., 2012), pregnant and lactating women (Moran et al., 2012a), and children (Moran et al., 2012b), showed only small changes in serum zinc when dietary intakes were doubled. These findings are not unexpected in view of the strong homeo­static mechanisms that prevent changes in serum zinc when zinc levels in the diet fluctuate. King (2018) has emphasized that doubling dietary zinc intake in adults only results in a 6% change in serum zinc concen­trations, a differ­ence within the margin of error in measuring serum zinc. Among some popu­lation groups in whom high-phytate cereals and legumes replace animal source foods as the major source of dietary zinc (e.g., vegetarians), significant inverse relationships between serum zinc and dietary phytate : zinc molar ratios have been observed (Donovan and Gibson, 1995; Cantoral et al., 2015).

For an individual, serum zinc concen­trations are not a reliable indicator of zinc status due to both the strong homeo­static control mechanisms and the many factors other than zinc intake that can independently influence serum zinc concen­trations (see below). This means that serum zinc concen­trations may be within the normal range despite the presence of mild zinc defi­ciency generating growth retardation, loss of appetite, or impaired immune function. Currently for individuals the recom­men­ded approach com­bines a thorough medical history of risk factors for zinc defi­ciency, a qualitative dietary assessment to identify an inadequate dietary intake, and a clinical assessment, with plasma or serum zinc providing only supple­mentary data (King et al., 2015).

In individuals with more severe zinc defi­ciency, with manifestations such as skin lesions, hair loss, diarrhea, delayed sexual maturation, impotence, hypogonadism in males, and eye lesions, serum zinc concen­trations are usually low. For example, markedly low serum zinc concen­trations are present in individuals with the genetic disorder acrodematitis enteropathica (Wessells et al., 2014), in patients receiving TPN unsupple­mented with zinc (Arakawa et al., 1976), and in those individuals who have developed clinical signs during experimentally-induced zinc defi­ciency (King et al., 2015).

24c.3.1 Factors affecting serum zinc

Age-related changes in serum zinc concen­trations (Table 24c.3) may be apparent, with levels increasing with age in some Pilch and Senti, 1984. Villalpando et al., 2003; Gibson et al., 2011), but not all (Hennigar et al., 2018) national surveys and in some smaller studies (Rükgauer et al., 1997). In the 2011–2014 US NHANES III survey, where serum zinc was deter­mined in males and females aged ≥ 6y, only the serum zinc concen­trations for the males (not the females) increased from the ages of6–30y, declining thereafter

Table 24c.3 Median ±SE serum zinc levels of males and females in the United States by age. Data from the US NHANES 2011–2014 (Hennigar et al., 2018) (For conversion to µmol/L, divide by 6.536)

Age (years)	Male serum zinc (μg/dL)	Female serum zinc (μg/dL)
6–8	80.8±1.4	77.7±0.9
9–13	81.7±1.2	81.1±1.6
14–18	85.5±2.3	81.3±1.9
19–30	84.9±1.4	77.9±1.3
31–50	84.4±0.7	79.6±0.7
51–70	82.4±1.9	79.8±1.5
≥71	78.7±2.3	80.5±1.5

Gender may influence serum zinc concen­trations. During infancy and early childhood, boys tend to have lower serum zinc levels than girls (Smit-Vanderkooy and Gibson, 1987; Cavan et al., 1993a; Gibson et al., 2011). In the US NHANES III2011–2014 survey males from 6y to adults ≥ 71y had higher serum zinc concen­trations than females (Table 24c.3), a relationship that may be related to differ­ences in body size and lean body mass.

Diurnal variation in serum zinc concen­trations has been reported. Levels are higher in serum in the morning, regardless of fasting status, com­pared to the afternoon Pilch and Senti, 1984. In the US NHANES III 2011–2014 survey, serum zinc concen­trations were 9% lower in the blood samples drawn in the afternoon (Hennigar et al., 2018),(Table 24c.4)

Fasting status also markedly affects serum zinc concen­trations. During a usual overnight fast of ≥ 8h, serum zinc levels increase, generally reaching a peak just before breakfast. Fluctuations in serum zinc also occur throughout the day, mainly in response to the consumption of meals. Serum zinc levels decline progressively for several hours after each meal, before rising prior to the next meal (King et al., 1994). Hence, in addition to the time of day, fasting status and meal status should also be carefully controlled during the collection of blood samples for serum / plasma zinc to avoid confounding the results. Alternatively, the time of day and the time elapsed since the previous meal can be recorded and taken into account statistically during analysis (Diana et al., 2017) (Table 24c.4).

Table 24c.4 Median serum zinc for different collection times from NHANES III. Data from the US NHANES 2011–2014 (Hennigar et al., 2018). (For conversion to µmol/L, divide by 6.536)

All participants Time/fasting status	n	Median serum zinc (µg/dL ± SE)
Morning, fasting	1423	86.8 ± 0.8
Morning, nonfasting	89	85.0 ± 3.1
Afternoon	1124	76.9 ± 0.9
Evening	489	73.8 ± 0.9

Hemolysis either in vivo or in vitro has the potential to increase serum zinc concen­trations because the concen­tration of zinc in erythrocytes is about 10–20 times higher than in serum or plasma. Hemolysis may be particularly important in cases of zinc defi­ciency, when red cell fragility increases (Bettger et al., 1978). Killilea et al. (2017) suggest that a 5% increase in plasma or serum zinc concen­tration occurs with each 1g hemoglobin/L in plasma or serum, a concen­tration that can be identified by chemical hemoglobin assays or by matching to a color scale. A threshold of 1g hemoglobin/L is recom­men­ded for measure­ments of plasma or serum zinc to avoid significant increases in zinc caused by hemolysis.

Storage of blood samples prior to separation affects serum zinc concen­trations. Long intervals prior to separation are associated with progressively increasing serum zinc concen­trations as zinc is released from platelets(English and Hambidge, 1988). To avoid this increase, serum or plasma should be separated rapidly from the red blood cells within 30–40 minutesafter collection. If this is not feasible, the blood should be held at 2–10°C for no more than 24 hours to limit movement of cellular zinc into the plasma or serum (IZiNCG Technical Brief No.2, 2007).

Prolonged use of a tourniquet may result in an increase in serum or plasma zinc levels. This occurs as a result of increased intravascular pressure caused by venous occlusion, which may cause movement of fluid into the interstitial space, thus increasing the zinc concen­tration(Juswigg et al., 1982). Hence, tourniquet occlusion should only be used for about one minute (King et al., 2015).

Acute and chronic infection and systemic inflam­mation leads to serum or plasma zinc values that are spuriously low because of the transfer of zinc from the blood to the liver. This redistribution of zinc is caused by hepatic synthesis of metallo­thionein activated by the release of cytokines during the acute phase response (Raiten et al., 2015). Both the stage and severity of the infection influence the change in plasma zinc, but these changes do not correspond to a change in zinc nutrition. To adjust for the presence of systemic inflam­mation, two inflam­matory biomarkers (C‑reactive protein (CRP) and α‑1‑acid glycoprotein (AGP)) should be measured along with serum or plasma zinc, allowing a regression-based correction to be applied (McDonald et al., 2020).

Evaluation of zinc intervention programs requires scheduling the end-line blood collection before the end of the intervention because serum zinc concen­trations rapidly return to baseline levels after the withdrawal of supple­mentation (King et al., 2015).

Figure 24c.6. Median serum zinc concen­trations during pregnancy at 1-mo intervals with the line of linear regression. From Brown et al., (2004)

Pregnancy is associated with decreases in serum zinc concen­trations irrespective of zinc intake Donangelo and King, 2012). The decline is evident at two months of gestation and at this early stage is attributed to hormonal changes, whereas the later decline is linked to plasma volume expansion (King, 2018). At the end of pregnancy, serum zinc concen­trations are about 20–25% lower than pre-pregnancy values.(Hennigar et al., 2018). In the US NHANES III 2011–2014 survey, the geometric mean ± SD serum zinc for pregnant women (n=34) was69.1±l2.0µg/dL vs. 79.1±1.9µg/dL (10.6±0.3 vs. 12.1±0.3µmol/L) for the age-matched non-pregnant women popu­lation; data by month of pregnancy are not available. The median serum zinc concen­trations by month of pregnancy, drawn from the NHANES II data is shown in Figure 24c.6.

Hypoalbuminemia (serum albumin ≤ 3.5g/dL) is associated with an increased risk of low serum zinc concen­trations because about60–70% of the serum zinc is bound to albumin (Cousins, 1989). condi­tions such as inflam­mation and infection, and disease states such as alcoholic cirrhosis and protein-energy malnutrition, result in a decline in serum zinc concen­trations concurrent with hypoalbuminemia which are probably independent of an individual’s zinc status (King, 2018).Not surprisingly, in the US NHANES III 2011–2014 survey, low serum albumen concen­trations were associated with an increased risk of low serum zinc (Hennigar et al., 2018).

Estrogen-containing preparations, such as oral contraceptive agents and other hormones, when used, may lead to markedly lower serum zinc oncentrations (Swanson and King, 1982), depending on the hormone dosage (Kamp et al., 2011). (Table 24c.5).

Table 24c.5 The effect of oral contraceptive (OC) use on serum zinc and alkaline phosphatase concen­trations. ^aAfter cor­rection for age. Data from Gibson et al. (2001).

Participants (n)	Mean serum zinc (µmol/L)	Mean serum alkaline phosphatase (U/L)
Non-users (202)	12.11	22.8
OC Users (128)	11.81	19.5
Probability of significant diff.^a	0.05	0.01

In the US NHANES III survey, however, there were no differ­ences in serum zinc concen­trations between oral contraceptive users and nonusers, perhaps because of the now decreased hormone dosage (Hennigar et al., 2018). The IZiNCG Technical Brief No.2 (2007) advises the collection of data on the use of oral contraceptive agents and other hormones in studies measuring serum zinc concen­trations.

Rapid synthesis of tissues during growth may lead to a fall in serum or plasma zinc. This effect has been reported in children during the anabolic phases of recovery from malnutrition (Golden and Golden, 1981) and during the rapid growth that occurs in preterm infants (Altigani et al., 1989). The decline in plasma zinc probably arises from the increased uptake of zinc from the exchangeable zinc pool in the plasma, induced by the increased demands for zinc for growth (Aggett and Comerford, 1995).

Weight loss or starvation releases zinc from muscle breakdown, resulting in transient increases in serum zinc concen­trations (Henry and Elmes, 1975; IZiNCG Technical Brief No.2, 2007).

Anemia has been associated with low serum zinc concen­trations in a few studies in children in both high-resource (Cole et al., 2010; Houghton et al., 2016) and low resource (Wieringa et al., 2016) settings, in pregnant women in rural Ethiopia (Gibson et al., 2008), and in females in the2011–2014 US NHANES III survey (Hennigar et al., 2018). Several mechanisms may explain the association of serum zinc and hemoglobin concen­trations (King, 2018): zinc-dependent enzymes are needed for heme synthesis, and zinc may also stabilize the erythrocyte membrane, thereby protecting the membrane from degradation during oxidative stress (O’Dell, 2000).

Malabsorp­tion syndromes and inflam­matory bowel diseases result in low serum or plasma zinc concen­trations arising from alterations in the integrity of the mucosal cells and a reduction in zinc absorp­tion (King and Cousins, 2014).

Environmental enteric dysfunction is known to impair zinc absorp­tion and increase intestinal endogenous zinc losses (Lindenmayer et al., 2014) that together result in low serum or plasma zinc concen­trations (Manary et al., 2010).

Sickle cell disease has been associated with low plasma zinc concen­trations in children and young adults (Abshire et al., 1988). This has been attributed to increased urinary zinc excretion, possibly mediated by increased zinc mobilization from bones due to recurrent bone ischemia (Schimmel et al., 2016).

High levels of supple­mental iron beyond those usually consumed in the diet may reduce serum zinc concen­trations due to a decrease in zinc absorp­tion. The mechanism is uncertain. Iron and zinc are thought to com­pete for intestinal absorp­tion via the shared divalent metal trans­porter 1 (DMT1) and/or another com­mon pathway in the apical membrane of the intestinal cell (Gunshin et al., 1997). The adverse effect on zinc absorp­tion, however, is less when lower amounts (≤ 10mg) of iron supple­ments are given (Fischer Walker et al., 2005) or when both the iron and zinc are provided in a com­plex food matrix (Olivares et al., 2012; Esamai et al., 2014).

24c.3.2 Interpretive criteria

Table 24c.4 (shown above) presents the medium serum zinc by fasting status and time of sampling for all NHANES III participants.

The International Zinc Consultative Group (IZiNCG) reanalyzed serum zinc data from NHANES II. In this reanalysis, data for participants with condi­tions significantly affecting serum zinc concen­trations were excluded. These were individuals: with low serum albumin (< 3.5g/dL); with an elevated white blood cell count (> 11.500 cells per µL); using oral contraceptive agents, hormones or steroids; or experiencing diarrhea. The IZiNCG also took age, gender, fasting status (i.e., > 8h since the last meal), and time of day of the sample collection into account in the reanalysis; details are given in (Hotz et al., 2003) and (IZiNCG, 2004).

From these reference data on healthy individuals, IZiNCG defined statistically derived reference limits based on the 2.5th percentile of serum zinc concen­tration for males and females aged < 10y and ≥ 10y (by fasting status) and time of sampling. These are given in (Table 24c.6) for children < 3y. IZiNCG suggests that the cutoffs for children < 10y also be applied to children < 3y, until appropriate reference data are available (IZiNCG, 2004).

Table 24c.6 Suggested lower reference limits (2.5th percentiles) (µmol/L) for serum zinc concen­tration based on age group, sex, fasting status, and time of day collection, derived from the NHANES II data. For conversion to µg/dL, divide by 0.1530. Adult reference limits based on data from subjects 20y and older only. NA, not available. Data from Brown et al. (2004).

Collection time	Children < 10y	Males ≥ 10y	Females ≥ 10y
AM fasting	NA	11.3	10.7
AM other	9.9	10.7	10.3
PM	8.7	9.3	9.0

Only a small number of pregnant women were examined during NHANES II (n=61). The 2.5th percentile values for the first and second plus third trimesters of pregnancy were calculated by IZiNCG and are 8.6 and 7.6µmol/L, respectively. No estimate for the 2.5th percentile serum zinc value for lactating women can be derived from the NHANES II survey results because of the limited sample size. Instead, at present the cutoff derived for nonpregnant women should be used for lactating women (IZiNCG, 2004). Further reference data are required to confirm and develop lower cutoffs for pregnant and lactating women and infants and children < 3y.

In addition to recording age group and sex, data on both fasting status (including the time interval since the last meal) and the time of day of the blood collection, should also be collected. The IZiNCG Technical Brief No.2 (2007) also recom­mends measuring an acute-phase protein indicative of infection or inflam­mation to assist in the interpretation of the serum zinc results. Ideally the two acute phase proteins recom­men­ded by WHO (2014) — C‑reactive protein (CRP) and α1‑acid glycoprotein (AGP) — should be measured as elevated levels could indicate underlying inflam­mation, which reduces serum/plasma zinc concen­trations (Raiten et al., 2015). Hence, before applying the appropriate IZiNCG reference limits, serum or plasma zinc concen­trations may need to be adjusted statistically for the time of day, the time interval since the previous meal (Arsenault et al., 2011), and inflam­mation (Macdonald et al., 2020), in order to provide a true estimate of the prevalence of low serum zinc concen­trations. To ascertain whether adjustment for inflam­mation is warranted, correlation and decile analysis between plasma / serum zinc concen­trations and CRP or AGP should be conducted. If the correlation is negative and significant, a visual inspection of the decile analysis should be undertaken to confirm whether there is an increasing prevalence of zinc defi­ciency by CRP or AGP decile (McDonald et al., 2020). If either of the latter is observed, then the regression correction approach recom­men­ded by the Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) Project should be used to adjust the serum / plasma zinc concen­trations affected by inflam­mation. In this approach the inflam­matory biomarkers (CRP and AGP) are treated as continuous variables so that greater corrections can be applied when the inflam­matory biomarkers indicate severe inflam­mation (Namaste et al., 2017; McDonald et al., 2020).

IZiNCG (Technical Brief No.2, 2007) has recom­men­ded that if more than 20% of the popu­lation (or popu­lation sub-group) has a serum zinc concen­tration (after appropriate adjustment) below the relevant reference limit (based on age, sex, time of day of the blood collection, and fasting status), the whole popu­lation (or sub-group) should be considered to have an elevated risk of defi­ciency of public health importance. This “trigger level”, however, can be modified depending on the available resources within a country to control zinc defi­ciency (King et al., 2015).

Not surprisingly, the cutoff associated with overt clinical signs of zinc defi­ciency (i.e. 50µg/L, 7.6µmol/L) (Wessells et al., 2014) is higher than the statistically derived IZiNCG reference limits presented in Table 24c.6 that are based on reference data from healthy individuals. As such, these IZiNCG reference limits provide a greater margin of safety for assessing an increased risk of zinc defi­ciency at the popu­lation level before the appearance of clinical signs (King et al., 2015).

24c.3.3 Measurement of serum zinc

Blood samples for serum or plasma zinc should be taken under carefully controlled, standardized condi­tions. Contamination from various sources such as preservatives, evacuated tubes, lubricants, anticoagulants, water, and rubber stoppers must be avoided. The reader is advised to consult the IZiNCG website for practical tips on the collection of blood samples for the measure­ment of serum or plasma zinc (IZiNCG Technical Brief No.2, 2007). For venipuncture blood samples, trace-element-free evacuated tubes with siliconized rather than rubber stoppers must always be used. Stainless steel or siliconized needles and Teflon or polypropylene catheters can all be used. For capillary blood samples, the use of polyethylene serum separators with polyethylene stoppers and olefin-oligomer is recom­men­ded (Iyengar, 1998). As noted earlier, fasting status, time of day of blood collection, time elapsed since the previous meal and since the blood sample was centrifuged, and presence of inflam­mation should all be recorded so that serum zinc values can be adjusted statistically as needed (Arsenault et al., 2011) prior to selecting the appropriate IZiNCG reference limits.

Box 24c.3. shows pre-collection considerations recom­men­ded by IZiNCG (2007) to minimize contamination during the collection of samples for serum / plasma zinc:

Box 24c.3 Pre-collection considerations to minimize contamination

• Obtain supplies of diposable polyethylene gloves, free of talc or other coatings, for those handling blood samples.
• Arrange for samples be processed in a laminar flow class 100 clean room, laminar flow hood, or otherwise in a clean, dust- and smoke-free laboratory.
• Check that stainless steel needles are available.
• Pre-screen anticoagulants, if required, for adventitious zinc contamination prior to use; zinc-free heparin is preferred.
• Pre-screen trace-element-free polyethylene evacuated tubes, stoppers, and serum separators for zinc contamination prior to use.
• Pre-screen polyethylene storage vials and transfer pipets for zinc contamination before use.
• Decontaminate all equipment to be used, with the exception of the prescreened disposable items. This can be achieved by careful washing (soak for 24h in ultrapure 10%–20% HCl or HNO₃ solution and then rinsing three or four times with distilled, deionized water)
• Cover or seal all materials and equipment to be used, prior to storage to avoid dust.

Once the blood has been collected, blood samples should be placed in a refrigerator or on ice and allowed to clot for 30–40 minutes, and then centrifuged to separate the serum or plasma. This approach minimizes any transfer of zinc from the cells to the serum or plasma. Alternatively, if centrifugation on site is not feasible, then the blood should be placed in a cool box or refrigerator immediately after collection where it can be held at 2–10°C for up to 24 hours prior to being centrifuged and the serum or plasma separated. After separation, the serum or plasma should be trans­ferred to a trace-element free screw-top vial for storage either in a refrigerator (for up to several days) or frozen until analyses. Any obviously111 hemolyzed samples (i.e., > 1g hemoglobin/L) should be discarded (Killilea et al., 2017; IZiNCG Technical Brief No.xx, 2020).

Some investigators prefer to use plasma because it is readily separated, less susceptible to platelet contamination, and not subject to contamination from a reaming instrument (Smith et al., 1985). For plasma samples, the recom­men­ded anticoagulant is zinc-free heparin. Before using any other anticoagulant it should be screened for contaminant zinc. For more details, see King et al. (2015).

The most frequently used method for the analysis of serum zinc is flame atomic absorp­tion spectrom­etry (AAS), using a direct method in which the samples are diluted by 5 to 10-fold in solvents such as 6% aqueous butanol or 10% aqueous propanol (King et al., 2015). Dilution of the serum reduces the viscosity of the sample, which minimizes both the matrix effect on the rate of aspiration and the tendency for the flame atomic absorp­tion burner-head to becom­e blocked. This may be especially a problem with plasma because of precipitates that form in these samples. Sometimes the serum or plasma samples are ashed at low temperatures prior to analysis by flame AAS. For very small samples, flameless AAS can be used. Increasingly, inductively coupled plasma (ICP) atomic emission spectrom­etry or ICP mass spectrom­etry are being used. For all methods, in‑house quality controls and standard reference materials (SRMs) such as bovine serum (SRM 1598) from the National Institute of Standards and Technology should be used. A coefficient of variation of < 5% is attainable for zinc using flame AAS. Use of a trichloroacetic acid deproteinization technique is not recom­men­ded.

Box 24c.4 shows the optimal procedures for the collection and analysis of serum or plasma samples to avoid adventitious trace element contamination. as recom­men­ded by IZiNCG (2004).

Box 24c.4 Optimal collection and analysis procedures

• Arrange for the participant to be seated.
• Clean the participant’s skin with alcohol at the site of the antecubital vein.
• Restrict occlusion of the participants’s arm with tourniquet to about 1 min.
• Draw blood, using a stainless steel needle, and collect into trace-element-free evacuated blood collection tubes without anticoagulant for processing serum.
• Place blood sample immediately in a refrigerator or on ice and allow to clot for 30–40min.
• Centrifuge blood sample at 2000–3000g for 10–15 min.
• Separate the serum using a zinc-free tranfer pipet.
• Discard any obviously hemolyzed serum samples.
• Consider retaining frozen red-cell residue for other determinations.
• Store serum samples frozen unless they are to be analyzed immediately.
• For analysis, dilute the sample quantitatively 5 or10-fold in solvents such as 6% aqueous butanol or 10% aqueous propanol.
• As soon as possible, determine sample zinc concen­tration using AAS or an alternative with appropriate standard dilutions, along with in-house quality controls and standard reference materials.

Practical tips for implementing these procedures are provided in IZiNCG Technical Brief No.2 (2012). Serum or plasma samples for zinc analyses can be refrigerated (4°C) for 2–3 weeks prior to analysis. For longer storage periods, samples should be frozen at −25°C or below. To prevent dehydration during long-term storage, especially if frost-free freezers are used, serum or plasma samples should be stored together with ice cubes in sealed plastic bags.