Background: Human plasma osmolality is tightly regulated in the 275 – 295 mOsm/kg range and breast milk osmolality (Mosm) oscillates through a slightly higher range. Studies in dairy cows demonstrate an influence of systemic hydration on bovine Mosm.

Objective: To describe hydration status in lactating women and explore any relation with Mosm.

Methods: 31 low- and middle-income women gave a full-breast expression after 90 min of nursing abstinence along with a concurrent spot sample of urine, 15 of whom repeated ~7 days later. Milk volume (Mvol) was measured and aliquots of urine and milk stored from 1 to 12 wks. Osmolality was determined in both fluids using a Vogel, Löser 815 osmometer expressed as mOsm/kg. Data were analyzed using SPSS version 20.

Results: The Mvol ranged from 2-100 mL, and Mosm from 288-448 mOsm/kg, whereas Uosm ranged from 93-1678 mOsm/kg. Nevertheless, the Spearmann rank-order correlation coefficient was r=0.214 (p=0.153) [n=46]; Mosm was not associated with maternal age r=0.243 (p=0.104), nor with lactation age r= -0.280 (p=0.060). Pearson correlation for repeat Mosm was r=0.204 (p=0.233) and for Uosm was r=0.642 (p=0.005) [n=15].

Conclusions: Lactating women have higher Uosm than other sectors in Western Guatemala. Mosm is not associated with systemic hydration as reflected by Uosm. There is a high association of systemic hydration status on repeated measurement.

Supported by The European Hydration Institute, Madrid, Spain; Fundación Iberoamericana de Nutricion, Granada, Spain; and Hildegard Grunow Foundation, Munich, Germany

Background: Diarrhea remains the second greatest cause of child morbidity and mortality in Guatemala, yet adherence to the WHO recommended treatment of Zinc (Zn) & ORS is low.

Objective: We evaluated the effectiveness of health-facility-level co-packaging of Zn and ORS to improve health provider prescription practices and caregivers‘ adherence to the diarrhea treatment (2 days ORS; 10 days Zn) for children 2-59 months of age in rural Guatemala.

Methods: Zn & ORS co-packaging development was guided by social marketing and then evaluated in a community-randomized intervention trial. The intervention group (IG) received Zn & ORS in a graphic co-pack with instructions and provider messages for counseling in diarrhea treatment. The control group (CG) received normative care of Zn and ORS without co-packaging or messages. Home-monitoring of adherence was conducted at 5 and 10 days post-prescription in 20 health posts in San Marcos Province.

Results: Health providers in the IG were more likely to dispense both medications than those of the CG (aOR: 2.3; 95%CI: 1.0, 5.4). IG mothers (n=123) were more likely to give the full 10 days of zinc (aOR:1.7; 95%CI; 1.0, 2.8) than CG mothers (n=138), and IG mothers provided 1 more day of Zn (p<0.01).

Conclusions: The co-packaging intervention improved both the prescription practices and adherence to zinc. This innovation could improve diarrhea treatment in Guatemala and holds potential utility in other situations of combined therapies.

Research Support: Grand Challenges Canada & Micronutrient Initiative

Background: In diverse areas, culturally-sensitive tools may better assess the food environment associated with diet.

Objective: To determine the availability of healthy foods in Latino corner stores (bodegas) to elucidate its relationship with culture.

Methods: Three bodegas were selected from low-income urban areas in Somerville, MA by field enumeration. The reference Nutrition Environment Measures Survey for Stores (NEMS-S) and a Guatemala NEMS-S version were used to assess healthy food vs. regular food availability in bodegas. Both surveys were conducted in each bodega. Frequency and proportion of items were determined. Chi-square tests assessed differences in availability by survey method across stores.

Results: Bodegas have limited healthy foods as measured by the reference NEMS-S. A greater availability of healthy foods in bodegas is indicated by the Guatemala NEMS-S. All stores had 100% of the time, regular: milk, baked goods, soda, juice, chips, bread; diet soda; and 100% juice. Often, stores had skim milk, low-sugar cereal, and wheat bread; but rarely, frozen dinners and low fat baked goods. Of the three vegetables added to the Guatemala NEMS-S, at least one was available in two of the three stores; and of the six fruits added at least two were available in each store. No significant differences in availability by store were found.

Conclusions: The Guatemala NEMS-S indicates a greater availability of healthy foods in bodegas than the reference NEMS-S. Therefore, culture may impact food availability, as stores may cater towards specific ethnic populations. The existence of culturally-specific stores needs consideration when creating healthy food policies.

Funding: Tufts Undergraduate Research Fund

Background: Diarrheal is one of the major causes of mortality and morbidity among Guatemalan children. As Zn+ORS in diarrhea therapy has potential to save lives and reduce illness in children, it is recommended by the WHO. In Guatemala, little is known about Zn+ORS adherence and the influencing factors among caregivers.

Objective: To understand caregivers‘ knowledge, attitudes and beliefs that influence adherence to Zn+ORS for diarrhea treatment in children ≤5y and how these were impacted by a Zn+ORS co-packaging intervention.

Methods: We report the qualitative findings from a larger community-based randomized intervention trial, in which Zn+ORS was co-packaged at the facility and provider counseling for diarrhea treatment was strengthened. In-depth interviews were conducted with 40 caregivers of children <5y in 20 health posts in Western Guatemala.

Results: Mothers perceived the co-packaging of Zn+ORS as an added value and a motivation to adhere to treatment. Enablers to adherence included receiving respectful and treatment specific counseling in their mother tongue, treatment without-cost, a good-graphical explanation of dose and preparation, and home visits to follow-up on child´s health improvement by health care representatives.

Conclusion: The co-packaging of Zn+ORS and respectful counseling both increased the perceived value of regular diarrhea treatment. The intervention positively influenced mothers‘ adherence to the prescription and created demand for treatment.

Financed by: Grand Challenges Canada & Micronutrient Initiative

Background: To overcome potential liabilities of conventional infantometer measurement of birth-length, we have explored in the past the use of standardized cell-phone camera photographic images, first with inanimate 2-D and 3-D models and later with newborn infants, all to standardized conditions.

Objective: To compare newborn birth-length estimates from: A) trained anthropometrist measurements; B) photographic digital camera images; and C) routine hospital records.

Methods: 29 infants born in the Departmental Hospital of Sololá in western Guatemala, were measured once routinely at delivery by the maternity nurse, and again, at 0-2 days, by a trained investigator on a Seca 2010 infantometer. Immediately thereafter, a sagittal photo was taken with a Nikon Coolpix S-3300 camera. Alignment of head and neck and to 90o flexion of the ankle was combined with inked landmarks on the knee and hip joints and a 5-cm reference-object on the surface of the thorax. The sum of the head+trunk plus upper- and lower-leg segments was measured in mm on a paper print-out, and converted to actual length in cm in reference to the object.

Results: We compared the investigator’s “gold-standard” length measurement (A) 47.0±2.2 cma, with the photographic image estimate (B) 44.8±3.6 cmb, and the hospital record (C) 50.3±2.2 cmc by repeated-measures ANOVA (p<0.001) (n=25). The Pearson correlation coefficient for A vs B was r=0.603 (p=<0.001), and for A vs C was r=0.642 (p=<0.001). Note: superscripts indicate ANOVA post-hoc differences.

Conclusion: The 2-cm underestimation may be amenable to a correction factor application, and the photographic method could be a valid and comfortable technique for assessing birth-length in newborns.

Background: Measures of the community nutrition environment are important to understanding the food environment, which affects individual diet.

Objective: To develop a formal protocol to improve the reliability of using direct observation to map the food environment in Guatemala.

Methods: Four teams of two raters were asked to walk two urban thoroughfares in urban Quetzaltenango, Guatemala. Each team consisted of one rater from Guatemala and one from the U.S. All teams read the mapping protocol and explored the area prior to mapping. Data was collected in a step-wise fashion. The first team collected data and revised the mapping protocol that was then utilized by the next team. The validity of direct observation in mapping a community food environment in Quetzaltenango was determined by comparing the inter-team concordance on each subsequent use of the protocol; and compared to data previously collected without a protocol.

Results: The inter-team reliability improved with the use of a formal protocol and mixed-ethnicity teams. On one street, the inter-team reliability improved to 76% using the revised protocol from 67% with the original protocol from 60% without a protocol. On the other street in which 40% of the food establishments are mobile vendors, the improvement in inter-team reliability was modest.

Conclusions: In Guatemala, street-mapping correspondence saturates at a certain point with or without considering the ambulatory aspects of carts and street vendors. A formal protocol and mixed teams that include a rater native to the country of interest improve the use of direct observation as a reliable low-cost method to map the community food environment.

Background: SCM is an asymptomatic inflammatory condition of the lactating breast where a breast milk ratio of Na/K>0.6 is the current indicator. Our objective was to determine maternal and infant factors associated with an increased or decreased likelihood of SCM.

Methods: Breast milk samples were collected from 105 lactating Mam-Mayan mothers with infants < 6 mo. Inductively Coupled Plasma Mass Spectrometry measured Na and K. A structured in-depth questionnaire for independent factors including socio-demographic characteristics, maternal and infant health status, breastfeeding (category, initiation, feeding frequency), cultural practices and beliefs, and maternal physical activity that might be associated with SCM was administered to mothers; anthropometry of both mother and infant was easured.

Results: Fourteen-percent had a breast milk Na/K ratio >0.6. SCM was associated with maternal age (OR=1.1, p=0.011), parity (OR=1.3, p=0.011), walking (OR=1.7, p=0.008), breastfeeding frequency (OR=1.1, p=0.024), and infant weight-for-age (WAZ) score (OR=0.57, p=0.049).

Conclusion: Results suggest that interventions to decrease the prevalence of SCM should be targeted to older mothers with high parity. Interestingly, breastfeeding frequency and walking were associated with SCM whereas cultural practices and beliefs were not in our population. Our findings were consistent with previous research where maternal SCM was associated with decreased infant weight.

Funding: McGill University International Mobility Travel Award

Background: Little is known about the impact of SCM on the immunological and mineral concentrations of human breast milk. Our objective was to determine if minerals and cytokines differed in mothers with and without SCM during 3 stages of lactation.

Methods: Transitional milk (TM: 5-17d, n=21), early mature milk (EMM: 18-46d, n=32) and mature milk (MM:109-187d, n=59) samples were collected from Mam-Mayan women. Inductively Coupled Plasma Mass Spectrometry was used to analyze the concentration of 13 minerals (Na, K, Ca, Mg, Mn, Zn, Cu, Cr, Sr, Se, Rb, Fe, P) and immunoassay with Luminex was used to determine the concentration of 4 cytokines (IL-1β, IL-6, IL-8, TNF-α).

Results: Prevalences of SCM, using Na/K > 0.6, were 26.3% in TM, 15.6% in EMM and 8.9% in MM. Na, K, P, Cu, Fe, Rb, Zn and IL6 were higher in TM and EMM whereas Mg was higher in MM, as was IL-8. SCM was associated with changes in P and Se and with the presence of 3 cytokines (IL-6, IL-8, and TNF-α) in TM only. Regression analyses for each mineral showed that cytokines were associated with higher milk mineral concentrations: IL-1β with P, Fe and Mn; IL-6 with Na, K, Ca and Cu; IL-8 with Zn; and TNF-α with Na, Mn and Se.

Conclusion: Milk minerals and cytokines concentrations vary by lactation stages. The cytokines are associated with changes in milk mineral concentrations. Our finding suggests that IL-6 is associated with elevated Na/K ratio.

Background: Freezing and thawing of milk alters a number of assay values and a practical perspective for osmolality is needed.

Objectives: To determine the consistency of osmolality values in human milk (Mosm) after varying frozen storage time and thawing protocols.

Methods: 14 lactating women gave a 1st full­breast expression after 90 min of nursing abstinence and 11 returned within a wk for a 2nd extraction. 3 aliquots were stored at ­20oC: 1st brought to 21°C and measured within 1 wk; 2nd with same procedure ~10­11 wks later; and 3rd warmed to 37oC in a water bath and then cooled to 21oC also at this later time. Single milk collections from 16 additional women were analyzed at the final time­point by both thawing routines. Mosm was determined on a Vogel Löser 815 osmometer and expressed as mOsm/kg.

Results: For the subgroup of 25 samples measured twice after standard thawing (early and late), the respective Mosm medians were 288 mOsm/kg and 309 mOsm/kg (p=<0.001, Wilcoxon test). The Pearson r­value was 0.772 (p=<0.001). This is consistent with the typical osmolality of ~300 mOsm/kg. The overall descriptive Mosm median for the 46 milk specimens measured after simple thawing was 308 mOsm/kg compared to 304 mOsm/kg; this 1.3% difference was significant (p=0.011). The Pearson correlation coefficient was r=0.847 (p=<0.001). 85% of values were within a band of ±5% of the median.

Conclusion: Minor variations in storage and handling produce significant, but not practically meaningful, differences for osmolality research with human milk.

Objective: To assess vitamin D status and the influence of risk factors for low levels of this vitamin among Kekchi-Mayan and Garifuna adolescents from Guatemala.

Methods: We recruited 86 adolescents, boys and girls, 43 from each ethnic group, in 2 communities from the province of Izabal: Kekchis from Rio Dulce, and Garifunas, of African descent, from Livingston. We obtained data on demographics plus sun protection and exposure practices. Blood samples were collected to assess 25(OH)D concentrations as nanograms per milliliter (ng/ml). Descriptive statistics were generated and regression models were tested to explain differences between groups.

Results: Mean 25(OH)D levels were 26 and 30 ng/ml (p=0.01), for Kekchis and Garifunas, respectively. 21% of Kekchis vs. 5% of Garifunas were identified as vitamin D deficient. About 51% of both ethic groups had insufficient vitamin D, while those with normal values included 28% of Kekchis and 43% of Garifunas. We found that age, being Garifuna, being male, and >2 hrs spent outside were strong predictors of Vitamin D concentrations. However, with multivariate analysis only time >2 hrs spent outside and being male were predictive of Vitamin D concentrations.

Conclusions: Despite residing in an optimal geographic location to receive adequate sunlight exposure, nearly 65% of adolescents from Kekchi and Garifuna ethnicities were either Vitamin D insufficient or deficient. This finding of low vitamin D levels in adolescents is relevant as low 25(OH)D during adolescent years has been implicated in a number of health conditions during adulthood.