Sage Journals: Discover world-class research

Abstract

Background:

Adolescence presents an opportunity to influence diet, which impacts present and future health outcomes, yet adolescent diets globally are poorly understood.

Objective:

We generate evidence on adolescent diets globally and explore patterns and trends by subpopulation.

Methods:

We estimated mean frequency of consumption and prevalence of less-than-daily fruit and vegetable consumption, at-least-daily carbonated beverage consumption, and at-least-weekly fast-food consumption among school-going adolescents aged primarily 12 to 17 years from the Global School-based Student Health Surveys in Africa, Asia, Oceania, and Latin America between 2008 and 2015. Random-effects meta-analysis was used to pool estimates globally and by subgroup.

Results:

On average, adolescents consumed fruit 1.43 (95% confidence interval [CI] 1.26–1.60) times per day, vegetables 1.75 (1.58–1.92) times per day, carbonated soft drinks 0.99 (0.77–1.22) times per day, and fast food 1.05 (0.78–1.32) times per week. Overall, 34.5% (95% CI 29.4–39.7) consumed fruit less than once per day, 20.6% (15.8–25.9) consumed vegetables less than once per day, 42.8% (35.2–50.7) drank carbonated soft drinks at least once per day, and 46.1% (38.6–53.7) consumed fast food at least once per week. Mean daily frequency of fruit consumption was particularly low in South and East Asia (1.30 [1.02–1.58]); carbonated soft drink consumption high in Latin America (1.54 [1.31–1.78]), high-income countries (1.66 [1.29–2.03]), and modern food system typologies (1.44 [0.75–2.12]); and mean weekly fast food consumption high in mixed food system typologies (1.29 [0.88–1.71]).

Conclusions:

School-going adolescents infrequently consume fruits and vegetables and frequently consume carbonated soft drinks, but there is wide variability by subpopulation.

Keywords

adolescents youth children dietary intake fruits and vegetables carbonated soft drinks soda fast foods nutrition GSHS

Introduction

Adolescence is a window of opportunity to influence eating patterns and nutritional status, which can have a lasting impact on future health outcomes.^1
-3 Yet we are failing to support optimal nutrition among a great proportion of the largest generation of adolescents in history (1.8 billion).^4,5 Adolescent diets are typically characterized by low intakes of fruits and vegetables and high intakes of energy-dense, nutrient-poor foods, including sweet and salty items, sugar-sweetened beverages, and fast foods.^5,6 Anemia and micronutrient deficiencies are widespread among adolescent boys and girls, having profound impacts on their quality of life and risk of premature death as well as the health of future generations.^7,8 At the same time, overweight and obesity are increasing among adolescents in most regions of the world.^9
-11 Obesity during childhood and adolescence is associated with increased risk of noncommunicable diseases including hypertension, atherosclerosis, nonalcoholic fatty liver disease, and metabolic syndrome.^12
-14

Dietary quality plays a key role in these forms of malnutrition among adolescents. Fruits and vegetables are rich in micronutrients, fiber, and beneficial bioactive compounds.¹⁵ High intakes of fruits and vegetables are associated with increased blood vitamin concentrations among European adolescents,¹⁶ and low intakes are a major contributor to the global burden of disease, particularly noncommunicable diseases.¹⁷ Sugar-sweetened carbonated soft drinks generally contain trivial amounts of nutrients and high amounts of sugar. Consumption of sugar-sweetened beverages, including carbonated soft drinks, is strongly associated with obesity and type 2 diabetes in adolescents.¹⁸ Most fast foods are energy-dense and nutrient-poor, high in refined grains, sodium, and sugar and low in fruits and vegetables.^19
-21 Fast-food consumption among adolescents is associated with obesity and metabolic markers of diabetes and cardiovascular disease.^22,23

While there is growing global evidence for adult dietary patterns,²⁴ there is limited evidence on adolescent diets. A recent systematic review provided an overview of the scientific literature on dietary patterns of adolescent girls in low- and middle-income countries, but this study contained mostly non-nationally representative studies and excluded boys.⁶ A meta-analysis of the prevalence of risk factors for cardiovascular disease included pooled estimates of combined low fruit and vegetable intake, defined as consuming fruits or vegetables less than 5 times per day, among school-going adolescents aged 12 to 15 years between 2003 and 2011.²⁵ But this analysis did not include estimates of carbonated soft drinks or fast food; excluded adolescents aged 10 to 11 years and 16 to 19 years; did not analyze fruit and vegetable consumption separately; and used only a dichotomous outcome that disposes useful information. An analysis of the 2010 Health Behavior in School-aged Children (HBSC) study provided insight into the frequency of fruit and vegetable consumption in 33 mostly high-income countries in North America and Europe, but these findings are not generalizable to low- and middle-income countries.²⁶ To our knowledge, there have been no other global dietary assessments or reviews of adolescent diets.

There has been a recent surge in interest in adolescent nutrition, which has generated evidence reviews,^27
-29 new global guidelines,³⁰ international meetings,^31
-33 donor commitments, and political will to institute new policies and programs specifically for adolescents.³⁴ But the evidence base on adolescent diets does not currently support informed decision-making. To strengthen efforts to generate global evidence on adolescent diets and nutrition, we estimated mean frequency of consumption as well as prevalence of less-than-daily fruit and vegetable consumption, at least daily carbonated beverage consumption, and at least weekly fast-food consumption among adolescents from 72 countries. We use random-effects meta-analysis to pool estimates globally and by region, income group, and food system typology (FST)³⁵ and test for differences between the subgroups.

Methods

Overview

Global School-based Student Health Surveys (GSHSs) are nationally representative self-administered questionnaires of school-going adolescent boys and girls of all ages (primarily 12-17 years). Respondents were asked “during the past 30 days, how many times per day do you usually eat” fruit, vegetables, and carbonated soft drinks (excluding diet soft drinks). Adolescents were also asked “during the past 7 days, on how many days did you eat food from a fast food restaurant?” Country-specific examples for all foods were provided for clarity. The GSHSs have been conducted largely in middle-income countries in Asia, Africa, and South and Central America but include countries of all income groups. At the time of writing, the most recent surveys with data available were from 2015. We excluded studies before 2008. If more than 1 survey was conducted in the same country, only the most recent survey was included. This allowed us to analyze 72 studies for frequency of fruit and vegetable consumption and 64 for frequency of carbonated soft drink and fast-food consumption.

All World Health Organization (WHO) regions except Africa had 50% or more countries sampled out of the total for each region: 10 of 47 in Africa, 22 of 35 in the Americas, 15 of 21 in the Eastern Mediterranean, 6 (5 for carbonated soft drinks and fast food) of 11 in South and East Asia, and 19 (16 for carbonated soft drinks and fast food) of 27 in the Western Pacific. Middle-income countries were oversampled, but there were at least 6 in each income group: 8 (6 for carbonated soft drinks and fast food) in low-income, 29 (26 for carbonated soft drinks and fast food) in lower middle-income, 18 (17 for carbonated soft drinks and fast food) in upper middle-income, and 10 in high-income countries. Data were also analyzed by broad FST, where countries were classified into traditional, mixed, or modern by the High-Level Panel of Experts on nutrition and food systems.³⁵ In making these classifications, the panel considered characteristics from the food supply chain, which included production (availability); storage and distribution; processing and packaging; and retail and markets, as well as from the food environment, which included availability and physical access (proximity); economic access (affordability); promotion, advertising, and information; and food quality and safety.³⁵ Mixed FSTs were oversampled, but there were at least 5 in each group: 8 (7 for carbonated soft drinks and fast food) in traditional, 29 (24 for carbonated soft drinks and fast food) in mixed, and 5 in modern FSTs.

Data Analysis

We calculated mean frequency and prevalence estimates with corresponding 95% CIs for each country, using sample weights to account for stratification and clustering in the survey design. All prevalence variables were coded to represent a less healthy behavior: consuming fruit less than once per day, vegetables less than once per day, carbonated soft drinks at least once per day, and fast food at least once per week. We dichotomized outcomes as such to allow for contrasting comparisons between countries and groups and to present general consumption patterns. There are no clear cutoffs for frequency of consumption of these foods and health outcomes among adolescents, since studies finding associations between health outcomes and foods consumed are primarily based on quantities consumed.

Random-effects meta-analysis was conducted to pool estimates overall, since the true frequency of consumption was expected to vary between studies.³⁶ Individual studies were weighted according to their national population totals of adolescents aged 13 to 17 years in 2015³⁷ rather than their sample sizes, which are not reflective of the relative size of different national adolescent populations. Variance in the raw prevalences was stabilized using a double arcsine transformation and back-transformed by the inverse of the double arcsine transformation using the harmonic mean of the sample sizes.^38,39 I² was calculated to present the percentage of variability due to study heterogeneity. F tests (allowing a variable τ²), which are similar to the Knapp and Hartung method⁴⁰ and generally result in more conservative P values than χ² tests, were used to test for overall differences between groups (region, income, and FST) and subgroups (sex).^36,41 Any outcome variables with an F test P value below 0.05 were subsequently tested for pairwise comparisons between specific groups using the Holm adjustment⁴² to account for type I errors when testing multiple hypotheses simultaneously. Univariate meta-regression⁴³ was conducted to detect trends for World Bank income group and FST: We fit bivariate linear and cubic spline models to the data and performed a Wald-type test for linear trend.⁴¹ All analyses were performed using R (version 3.5.1).

Results

Of the 94 countries screened, 72 studies contained data on frequency of fruit (N = 254 331) and vegetable (N = 254 135) consumption and 64 on frequency of carbonated soft drink (N = 241 784) and fast-food consumption (N = 241 991). Twenty-two were excluded since they were conducted before 2008. Table 1 shows the summary characteristics for each study. On average, adolescents consumed fruit 1.43 (95% confidence interval [CI] 1.26-1.60) times per day, vegetables 1.75 (1.58-1.92) times per day, carbonated soft drinks 0.99 (0.77-1.22) times per day, and fast food 1.05 (0.78-1.32) times per week (Figure 1). Overall, 34.5% (95% CI 29.4-39.7) consumed fruit less than once per day, 20.6% (15.8-25.9) consumed vegetables less than once per day, 42.8% (35.2-50.7) consumed carbonated soft drinks at least once per day, and 46.1% (38.6-53.7) consumed fast food at least once per week (Figure 2). Heterogeneity of consumption between studies was very high—I² was at least 98.5% for mean and 99.3% for prevalence for all foods (Figures 1 and 2). Prevalence of less-than-daily fruit and vegetable consumption ranged from 17.7% (15.2-20.3) and 8.2% (6.7-9.8), respectively, in Algeria in 2011 to 61.7% (59.2-64.2) and 62.9% (60.5-65.2) in Maldives in 2014 (Supplemental Figures 1 and 2). At least daily carbonated soft drink consumption ranged from 22.2% (19.2-25.3) in Kiribati in 2011 to 80.4% (77.3-83.3) in Suriname in 2009 (Supplemental Figure 3). At least weekly fast-food consumption ranged from 20.9% (17.6-24.4) in Pakistan in 2009 to 85.8% (83.3-88.0) in Qatar in 2011, both countries in the same region (Supplemental Figure 4). F tests revealed no significant differences in consumption between boys and girls overall or by region, income, or FST (data not shown). However, there were notable differences by sex in individual countries. For example, in Pakistan, a higher proportion of boys consumed fruit less than once per day (40.3% [33.3-47.4]) than girls (20.4% [16.5-24.6]), while a lower proportion of boys than girls drank carbonated soft drinks at least once per day (28.1% [22.9-33.7] versus 49.2% [44.1-54.4], respectively; Supplemental Figures 1 and 3).

Table 1.

Summary Characteristics of Country Data Sets Included in the Meta-Analysis.^a

Region, Country	World Bank Income Group	Food Systems Typology	Year	Age, years	% male	Fruit, N	Veg, N	Carb Soft Drinks, N	Fast Food, N	Population (13- to 17-Year Olds)
Africa
Benin	Low-income	Mixed	2009	15·2 ± 1·1	65·14	2689	2686	2688	2685	1 200 989
Algeria	Upper middle-income	Mixed	2011	14·1 ± 1·4	48·03	4509	4512	4512	4521	2 874 960
Mozambique	Low-income	Traditional	2015	15·9 ± 1·9	53·33	1871	1884	1893	1885	3 263 493
Mauritania	Lower middle-income	—	2010	14·8 ± 1·2	47·42	2030	1998	2029	2033	451 987
Mauritius	Upper middle-income	—	2011	14·0 ± 1·1	45·04	3280	3292	3279	3267	94 317
Malawi	Low-income	Traditional	2009	13·9 ± 0·9	46·55	2334	2348	—	—	2 092 354
Namibia	Upper middle-income	—	2013	15·8 ± 1·8	47·29	4483	4478	4480	4522	268 718
Swaziland	Lower middle-income	—	2013	16·0 ± 1·7	47·64	3650	3663	3653	3631	149 895
Seychelles	Upper middle-income	—	2015	13·8 ± 1·5	47·34	2527	2520	2500	2516	5690
Tanzania	Low-income	Traditional	2014	14·3 ± 1·8	47·94	3760	3768	3772	3772	6 104 521
Americas
Anguilla	Low-income	—	2009	14·2 ± 1·4	47·08	937	940	946	941	1279
Argentina	Upper middle-income	Modern	2012	14·5 ± 1·1	47·75	28106	28021	28045	28173	3 484 891
Antigua and Barbuda	Upper middle-income	—	2009	13·9 ± 0·9	46·65	1257	1255	1261	1250	8594
Bahamas	High-income	—	2011	14·2 ± 0·9	45·17	1592	1606	1608	1610	27 915
Belize	Lower middle-income	—	2011	13·8 ± 1·5	47·19	2106	2104	2083	2111	39023
Bolivia	Lower middle-income	Mixed	2012	14·4 ± 1·2	50·69	3668	3660	3677	3658	1 107 796
Chile	High-income	Mixed	2013	14·6 ± 1·8	50·07	2039	2025	2040	2033	1 297 258
Costa Rica	Upper middle-income	Modern	2009	14·2 ± 1·1	48·22	2662	2673	2677	2670	382 447
Curacao	—	—	2009	15·3 ± 2·0	44·92	2742	2721	2742	2747	13 856
Dominica	Upper middle-income	Mixed	2009	13·9 ± 1·4	43·59	1628	1629	1627	1636	996 508
Grenada	Upper middle-income	Mixed	2013	14·0 ± 1·3	44·94	1526	1525	—	—	9343
Guatemala	Lower middle-income	Mixed	2015	14·4 ± 1·2	49·42	4332	4268	4348	4333	1 893 705
Guyana	Lower middle-income	Mixed	2010	14·4 ± 1·1	43·78	2353	2379	2368	2373	83 525
Honduras	Lower middle-income	Mixed	2012	13·9 ± 1·3	48	1772	1756	1762	1753	996 855
Jamaica	Upper middle-income	Mixed	2010	14·6 ± 1·1	48·78	1609	1588	1610	1607	266 052
Saint Kitts and Nevis	High-income	—	2011	14·3 ± 1·0	44·08	1728	1704	1731	1728	4747
Peru	Upper middle-income	Mixed	2010	14·5 ± 1·0	48·95	2881	2877	2878	2878	2 787 606
El Salvador	Lower middle-income	Mixed	2013	14·3 ± 1·0	53·8	1900	1886	1907	1901	656 905
Suriname	Upper middle-income	—	2009	14·7 ± 1·3	51·06	1692	1694	1681	1694	49 609
Trinidad and Tobago	High-income	—	2011	13·6 ± 1·4	54·6	2801	2795	2767	2802	86 652
Uruguay	High-income	Modern	2012	14·4 ± 1·0	46·47	3513	3497	3514	3509	254 494
British Virgin Islands	—	—	2009	14·0 ± 1·5	44·53	1651	1660	1649	1634	2655
Eastern Mediterranean
Afghanistan	Low-income	—	2014	15·1 ± 1·7	42·06	2505	2463	2507	2527	4 246 258
United Arab Emirates	High-income	Mixed	2010	14·1 ± 1·1	42·12	2568	2562	2552	2572	315 440
Egypt	Lower middle-income	Mixed	2011	13·5 ± 1·1	46·68	2519	2533	2528	2523	8 381 230
Iraq	Upper middle-income	—	2012	14·3 ± 1·3	56·49	2029	2020	2007	2016	3 964 241
Kuwait	High-income	Mixed	2015	15·1 ± 1·6	46·51	3559	3601	3591	3609	221 111
Lebanon	Upper middle-income	—	2011	13·9 ± 1·2	46·58	2282	2280	2272	2272	538 627
Morocco	Lower middle-income	Mixed	2010	14·0 ± 1·3	52·44	2906	2873	2899	2910	2 962 116
Oman	High-income	Modern	2015	15·4 ± 1·6	47·68	3387	3453	3430	3449	225 806
Pakistan	Lower middle-income	Traditional	2009	14·2 ± 0·8	75·17	5171	5176	5159	5175	19 364 073
Palestine	Lower middle-income	—	2010	13·9 ± 1·0	46·23	14450	14381	14479	14444	538 230
Qatar	High-income	Mixed	2011	13·3 ± 1·2	45·9	1912	1921	1985	1962	85 951
Sudan	Lower middle-income	Mixed	2012	14·8 ± 1·1	38·66	2173	2184	2172	2169	4 481 158
Syria	Lower middle-income	Traditional	2010	13·7 ± 1·1	40·07	3075	3094	3083	3072	2 311 460
Tunisia	Lower middle-income	Mixed	2008	13·8 ± 1·2	48·69	2848	2844	—	—	802 095
Yemen	Low-income	Mixed	2008	14·0 ± 1·4	61·08	1103	1118	—	—	3 101 491
South and East Asia
Bangladesh	Lower middle-income	Mixed	2014	14·2 ± 1·0	40	2951	2950	2929	2949	16 369 000
Indonesia	Lower middle-income	Traditional	2015	14·0 ± 1·6	45·81	11020	11087	11080	11077	23 011 064
Sri Lanka	Lower middle-income	Mixed	2008	13·7 ± 1·0	43·69	2595	2599	—	—	1 626 932
Maldives	Upper middle-income	Mixed	2014	15·5 ± 1·5	42·4	3450	3418	3450	3445	28 238
Thailand	Upper middle-income	Mixed	2015	14·5 ± 1·7	43·12	5844	5767	5869	5838	4 667 675
Timor-Leste	Lower middle-income	—	2015	15·4 ± 1·9	46·4	3595	3653	3625	3631	159 228
Western Pacific
Brunei	High-income	—	2014	14·6 ± 1·4	46·7	2586	2584	2574	2579	35 314
Cook Islands	—	—	2015	15·4 ± 1·4	49·14	699	701	696	700	1512
Fiji	Lower middle-income	Mixed	2010	14·3 ± 0·9	42·86	1657	1667	—	—	77 334
Cambodia	Low-income	Traditional	2013	15·6 ± 1·8	47·21	3798	3792	3799	3786	1 515 106
Kiribati	Lower middle-income	—	2011	14·3 ± 1·1	43·59	1567	1568	1574	1575	11 422
Laos	Lower middle-income	Traditional	2015	15·6 ± 1·2	45·6	3674	3665	3671	3669	713 552
Mongolia	Lower middle-income	—	2013	14·5 ± 1·7	46·85	5365	5349	5379	5352	218 224
Malaysia	Upper middle-income	Modern	2012	14·9 ± 1·4	50·01	25439	25476	25461	25472	2 775 984
Niue	—	—	2010	14·2 ± 1·7	58·87	141	141	141	141	141
Nauru	—	—	2011	14·3 ± 1·5	41·88	561	558	—	—	1130
Philippines	Lower middle-income	Mixed	2015	14·6 ± 1·5	45·56	8702	8750	8738	8753	10 318 214
Solomon Islands	Lower middle-income	—	2011	14·6 ± 1·3	51·88	1408	1388	1403	1390	66 463
Tokelau	—	—	2014	14·1 ± 2·1	48·15	136	140	—	—	130
Tonga	Lower middle-income	—	2010	14·2 ± 1·1	45·89	2198	2175	2192	2207	12 075
Tuvalu	Upper middle-income	Mixed	2013	14·0 ± 1·5	48·5	916	929	936	928	961
Vietnam	Lower middle-income	—	2011	15·6 ± 1·3	46·87	3317	3308	3319	3317	6 627 981
Vanuatu	Lower middle-income	Mixed	2011	13·8 ± 1·5	44·06	1107	1089	1110	1111	26 838
Wallis and Futuna	—	Mixed	2015	14·6 ± 1·8	48·14	1108	1107	1104	1102	1030
Samoa	Lower middle-income	—	2011	14·0 ± 1·0	41·35	2382	2359	2343	2396	21 492

Abbreviations: carb, carbonated; veg, vegetables; SD, standard deviation.

^aValues for age are means ± SD. Dashes indicate no data.

Figure 1.

Mean frequencies of consumption by World Health Organization (WHO) region. (A) Mean fruit consumption (n = 254 331), (B) mean vegetable consumption (n = 254 135), (C) mean carbonated soft drink consumption (n = 241 784), and (D) mean fast-food consumption (n = 241 991). Error bars show 95% confidence intervals (CIs). I² represents the percentage of variability due to individual study heterogeneity. The F statistic tests for overall differences between regions.

Figure 2.

Prevalence of less-than-daily, at-least-daily, or less-than-weekly consumption by World Health Organization (WHO) region. (A) Less-than-daily fruit consumption (n = 254 331), (B) less-than-daily vegetable consumption (n = 254 135), (C) at-least-daily carbonated soft drink consumption (n = 241 784), and (D) at-least-weekly fast food consumption (n = 24 1991). Error bars show 95% confidence intervals (CIs). I² represents the percentage of variability due to individual study heterogeneity. The F statistic tests for overall differences between regions.

Regional Heterogeneity

There was significant heterogeneity between regions for mean frequency of consumption of fruit (F = 2.7, P = .0363) and carbonated soft drinks (F = 8, P < .0001; Figure 1). At least daily carbonated soft drink consumption also varied significantly by region (F = 14.3, P < .0001; Figure 2). There were no significant differences between regions for frequency of vegetable or fast-food consumption (Figures 1 and 2). Mean daily frequency of fruit consumption was highest in Africa (1.72 [1.45-1.99]) and lowest in South and East Asia (1.30 [1.02-1.58]), where frequency of consumption was significantly lower than in Africa (P = .0316) and the Americas (1.55 [1.35-1.75], P = .0364; Figure 1 and Supplemental Table 1). Mean daily frequency of carbonated soft drink consumption was lowest in the Western Pacific (0.86 [0.40-1.32]) and Eastern Mediterranean (0.86 [0.57-1.16]) and by far the highest in the Americas (1.54 [1.31-1.78]), where frequency of consumption was significantly higher than in the Western Pacific (P = .0009), Eastern Mediterranean (P = .0001), South and East Asia (0.96 [0.59−1.32], P = .0007), and Africa (1.20 [0.79−1.62], P = .027; Figure 1 and Supplemental Table 1). A much higher proportion from the Americas (62.6% [57.0-68.0]) drank carbonated soft drinks at least once per day than in the Western Pacific (34.8% [20.9-50.2], P < .0001), South and East Asia (37.8% [24.6-52], P < .0001), Eastern Mediterranean (42.3% [33.3-51.6], P < .0001), and Africa (53.1% [37.1-68.8], P = .0437; Figure 2 and Supplemental Table 1).

Heterogeneity Between Income Groups

There was significant heterogeneity between income groups for mean frequency of consumption of vegetables (F = 3.7, P = .0165) and carbonated soft drinks (F = 6.8, P = .0006; Figure 3). At least daily carbonated soft drink consumption (F = 4.2, P = .0098) also varied significantly by region (Figure 4). There were no significant differences between income groups for frequency of fruit or fast-food consumption (Figures 3 and 4). Mean daily frequency of carbonated soft drink consumption was highest in high-income countries (1.66 [1.29-2.03]), where it was significantly higher than in low-income (0.98 [0.71-1.25], P = .0006) and lower middle-income (0.90 [0.63-1.17], P = .0021) countries (Figure 3 and Supplemental Table 1). Mean daily frequency of carbonated soft drink consumption was also higher in upper middle-income countries (1.34 [1.03-1.64]) than low-income (P = .023) countries (Figure 3 and Supplemental Table 1). At least daily carbonated soft drink consumption was higher in high-income (61.4% [49.4-72.7]) than low-income (43.9% [32.3-55.9], P = .038) and lower middle-income (39.1% [29.2−49.4], P = .0189) countries (Figure 4 and Supplemental Table 1).

Figure 3.

Mean frequencies of consumption by World Bank income group. (A) Mean fruit consumption (n = 247 293), (B) mean vegetable consumption (n = 247 107), (C) mean carbonated soft drink consumption (n = 235 452), and (D) mean fast-food consumption (n = 235 667). Error bars show 95% confidence intervals (CIs). I² represents the percentage of variability due to individual study heterogeneity. The F statistic tests for overall differences between regions.

Figure 4.

Prevalence of less-than-daily, at-least-daily, or less-than-weekly consumption by World Bank income group. (A) Less-than-daily fruit consumption (n = 247 293), (B) less-than-daily vegetable consumption (n = 247 107), (C) at-least-daily carbonated soft drink consumption (n = 235 452), and (D) at-least-weekly fast-food consumption (n = 235 667). Error bars show 95% confidence intervals (CIs). I² represents the percentage of variability due to individual study heterogeneity. The F statistic tests for overall differences between regions.

Mean weekly frequency of fast-food consumption was lowest in lower middle-income (0.94 [0.61-1.26]) and highest in upper middle-income (1.46 [1.05-1.86]) countries (Figure 3). At least weekly fast-food consumption was higher in high-income (48.9% [32.3-65.6]) than lower middle-income (43.6% [34.1-53.3], P = .0700) countries but with limited certainty (Figure 4). There were no significant pairwise differences between income groups for frequency of vegetable consumption (Supplemental Table 1). Meta-regression of income groups revealed that prevalence of at least daily carbonated soft drink consumption was positively and linearly associated with income group (P = .0205; Supplemental Figures 9 and 11). No significant linear correlations were found for frequency of fruit, vegetable, or fast-food consumption by income group (Supplemental Figures 9 and 11).

Heterogeneity Between Food System Typologies

There was significant heterogeneity between FSTs for mean frequency of consumption of carbonated soft drinks (F = 7.5, P = .002) and fast food (F = 6, P = .0058; Figure 5). At least daily carbonated soft drink (F = 3.9, P = .03) and at least weekly fast-food (F = 4.2, P = .0241) consumption also varied significantly by FST (Figure 6). There were no significant differences between FSTs for frequency of fruit or vegetable consumption (Figures 5 and 6). Mean daily frequency of carbonated soft drink consumption in traditional FSTs (0.72 [0.45-0.99]) was considerably lower than in mixed FSTs (1.21 [0.95-1.46], P = .0007) as well as modern FSTs (1.44 [0.75-2.12], P = .06) although with much less statistical significance (Figure 5 and Supplemental Table 1). At least daily consumption of carbonated soft drinks was also considerably lower in traditional (35.0% [24.5-46.3]) than mixed (49.9% [40.2-59.6], P = .0185) FSTs (Figure 6 and Supplemental Table 1). Interestingly, mean weekly frequency of fast-food consumption was much higher in mixed (1.29 [0.88-1.71]) than both traditional (0.80 [0.37-1.24], P = .0107) and modern (0.71 [0.43-0.99], P = .0055) FSTs (Figure 5 and Supplemental Table 1). Correspondingly, at least weekly fast-food consumption was higher in mixed (52.6% [42.8-62.4]) than traditional (39.0% [23.3-56.0], P = .0421) as well as modern (39.8% [29.7-50.2], P = .07) FSTs (Figure 6 and Supplemental Table 1). Meta-regression revealed a positive linear relationship between FST and mean daily consumption of carbonated soft drinks (P = .0187) as well as prevalence of at least daily carbonated soft drink consumption (P = .08) although with less statistical significance (Supplemental Figures 10 and 12). No significant linear correlations were found for frequency of fruit, vegetable, or fast-food consumption by FST (Supplemental Figures 10 and 12).

Figure 5.

Mean frequencies of consumption by food system typology. (A) Mean fruit consumption (n = 176 634), (B) mean vegetable consumption (n = 176 637), (C) mean carbonated soft drink consumption (n = 164 864), and (D) mean fast food consumption (n = 164 958). Error bars show 95% confidence intervals (CIs). I² represents the percentage of variability due to individual study heterogeneity. The F statistic tests for overall differences between regions.

Figure 6.

Prevalence of less-than-daily, at-least-daily, or less-than-weekly consumption by food system typology. (A) Less-than-daily fruit consumption (n = 176 634), (B) less-than-daily vegetable consumption (n = 176 637), (C) at-least-daily carbonated soft drink consumption (n = 164 864), and (D) at-least-weekly fast food consumption (n = 164 958). Error bars show 95% confidence intervals (CIs). I² represents the percentage of variability due to individual study heterogeneity. The F statistic tests for overall differences between regions.

Discussion

Among school-going adolescents surveyed in the GSHS, frequency of fruit, vegetable, carbonated soft drink, and fast-food consumption varied substantially by country. Fruit consumption was particularly infrequent in South and East Asia, but we found no clear patterns or trends by income or FST. There were no significant trends or patterns in vegetable consumption by region, income, or FST. Carbonated soft drinks were most frequently consumed in the Americas (followed by Africa), high-income countries (followed by upper middle-income countries), and modern FSTs (followed closely by mixed FSTs). Carbonated soft drinks were the only food with a clear trend of increasing frequency of consumption by increasing income and FST. Fast food was far more frequently consumed in countries with mixed than both traditional and modern FSTs, but there were no clear patterns by region or income.

Given the low overall frequencies of fruit and vegetable consumption, intake in all countries was likely far below the minimum amount recommended by the WHO (at least 400 g). Regardless of large variations in consumption of carbonated soft drinks, they are commonplace among most adolescents, consumed once per day on average globally. A recent review of sugar consumption in Europe and North America found that adolescents consumed more added sugars than any other age-group.⁴⁴ Frequency of fast-food consumption appeared to be less common, but the GSHS only asked about fast-food consumption from restaurants, and there was likely substantial consumption from informal sources, such as street foods, which were not represented.⁴⁵ Street foods are increasingly common in Latin America, Asia, and Africa and makeup a significant portion of the urban food supply.^45,46 When fast foods were considered as meals purchased outside the home, 93% of adolescent girls from low- and middle-income countries globally were found to consume them at least weekly.⁶

A meta-analysis of GSHS conducted between 2003 and 2011 found that 74% of adolescents aged 12 to 15 years consumed fruits and vegetables less than 5 times per day.²⁵ They found no clear patterns by region, but this may have been because the outcome was dichotomized or because fruit and vegetable consumption was combined. While adolescents in South and East Asia appeared to have less frequent combined intakes of fruits and vegetables than other regions, our findings of mean daily consumption in South and East Asia suggest vegetable consumption is high and fruit consumption low relative to other regions. An analysis of HBSC surveys, which includes mostly high-income countries from North America and Europe, did not report pooled estimates, but less-than-daily fruit consumption ranged from 51% in Denmark and French Belgium to 85% in Greenland, while less-than-daily vegetable consumption ranged from 45% in Flemish Belgium to 80% in Estonia. Fruit and vegetable consumption were much more frequent in the high-income countries from our analysis of GSHS than in the HBSC. However, the questions were asked differently between surveys, and the HBSC included younger children in general (11, 13, and 15 years) than the GSHS. Additionally, as the focus of the GSHS is on middle-income countries, our analysis of high-income countries only included 5 countries from the Americas, 4 from the Eastern Mediterranean, and 1 from the Western Pacific.

A recent systematic review of diets and eating practices of adolescent girls aged 10 to 19 years in low- and middle-income countries also found higher prevalences of less-than-daily fruit (56%) and vegetable (64%) consumption than in our analysis of the GSHS, but it did not include boys.⁶ Additionally, while the review was global, certain countries and regions were highly over- or underrepresented due to limited available studies. Moreover, the analysis was based on data from unstandardized food frequency questionnaires, and most studies were not nationally representative and often only conducted in urban or rural areas.

Our study has important limitations. Subgroups within region, income, and FST are not equally represented: Africa, low-income countries, and modern FSTs are underrepresented. Moreover, no studies were available for Europe or North America. While a majority of the countries included had a median net secondary school enrollment rate greater than 50%,⁴⁷ we caution generalizing findings to adolescents out of school who may have difference dietary patterns. Additionally, surveys conducted in 2008 may not be representative of current dietary habits, especially among adolescents from low- and middle-income countries that are rapidly developing. Further, the GSHS surveys collected self-estimated frequency of foods consumed in the previous 30 days, so we were unable to estimate quantities consumed. Recall of usual frequency among adolescents is less accurate and valid than 24-hour recall, particularly for foods uncommonly consumed and among younger adolescents.^48,49 Finally, FSTs were not based on the year the survey was conducted, thus some countries may have been classified differently during the survey year.

To our knowledge, our study is the first to provide global estimates of frequency of fruit, vegetable, carbonated soft drink, and fast-food consumption among adolescent boys and girls. We analyzed 72 standardized and nationally representative surveys, accounting for weighting in the sampling design. Our random-effects meta-analysis accounted for the large heterogeneity between countries, and we statistically tested for trends and/or differences between regions, income groups, FSTs, and sex, adjusting for multiplicity of testing. We also weighted estimates according to the national populations of adolescents, which vary substantially between countries, improving the generalizability of our findings.

The GSHSs provide valuable insight into the frequency of fruit, vegetable, carbonated soft drink, and fast-food consumption among school-going adolescents. Ideally, dietary questions in the GSHS and HBSC surveys would be unified to allow for better global estimates and representation of regions, since their geographic scopes are complementary. Of course, modifying these surveys substantially may limit the ability to analyze trends over time, so this trade-off should be considered. If possible, future surveys should collect information on other foods that likely promote health (eg, fish, eggs, legumes, and dairy) or are associated with disease (eg, unhealthily processed foods such as candy, chips, and cookies) in adolescents. Research is needed that analyzes the relationships between adolescent diets and health to provide a better understanding of the consequences of specific diets in different contexts. Population surveys where information on the determinants of health, including diets, is collected along with health outcomes should be analyzed, adjusting for non-dietary health determinants. Knowing the quantities of specific foods consumed would enhance this process. There is also an urgent need for research that identifies the determinants of adolescent diets, particularly in low- and middle-income countries, so that effective policies and interventions can be designed.

In conclusion, school-going adolescents generally consume unhealthy diets, low in fruits and vegetables and high in carbonated soft drinks. The specific dietary inadequacies—be they low frequencies of fruit and vegetable consumption or high frequencies of carbonated soft drinks and fast food consumption—depend on country, region, income group, and FST, but there are few generalizable trends. Global, regional, and national policies and programs addressing adolescent diets should be tailored to the particular dietary needs of the corresponding populations. For example, taxing sugar-sweetened beverages or other unhealthy foods in Latin America and high-income countries has potential to reduce intake among adolescents.⁵⁰ Low-quality diets among adolescents, in addition to low physical activity, are likely contributing to the coexistence of undernutrition, overweight/obesity, and noncommunicable diseases, which can have harmful immediate, long-term, and intergenerational consequences. Adolescents are unlikely to be motivated by a desire for health in the distant future, and creative interventions that harness widely shared adolescent values are needed to successfully improve diets.⁵¹

Supplemental Material

Supplemental Material, Appendix_(6) - Global Patterns of Adolescent Fruit, Vegetable, Carbonated Soft Drink, and Fast-Food Consumption: A Meta-Analysis of Global School-Based Student Health Surveys

Supplemental Material, Appendix_(6) for Global Patterns of Adolescent Fruit, Vegetable, Carbonated Soft Drink, and Fast-Food Consumption: A Meta-Analysis of Global School-Based Student Health Surveys by Ty Beal, Saul S. Morris and Alison Tumilowicz in Food and Nutrition Bulletin

Footnotes

Authors’ Note

TB and SSM designed the study. TB did the statistical analysis and wrote the first draft. All authors interpreted the data and reviewed and contributed to the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: Funding for the study came from the Ministry of Foreign Affairs of the Netherlands.

ORCID iD

Ty Beal

Alison Tumilowicz

Supplemental Material

Supplemental material for this article is available online.

References

Black

Victora

Walker

, et al. Maternal and child undernutrition and overweight in low-income and middle-income countries. Lancet. 2013;382(9890):427–451. doi: 10.1016/S0140-6736(13)60937-X.

Georgiadis

Penny

. Child undernutrition: opportunities beyond the first 1000 days. Lancet Public Health. 2017;2(9):e399. doi: 10.1016/S2468-2667(17)30154-8.

Sawyer

Afifi

Bearinger

, et al. Adolescence: a foundation for future health. Lancet. 2012;379(9826):1630–1640. doi: 10.1016/S0140-6736(12)60072-5.

Patton

Sawyer

Santelli

, et al. Our future: a Lancet commission on adolescent health and wellbeing. Lancet. 2016;387(10036):2423–2478. doi: 10.1016/S0140-6736(16)00579 -1.

Akseer

Al-Gashm

Mehta

Mokdad

Bhutta

. Global and regional trends in the nutritional status of young people: a critical and neglected age group. Ann N Y Acad Sci. 2017;1393(1):3–20. doi: 10.1111/nyas.13336.

Keats

Rappaport

Jain

Shah

Bhutta

. Diet and Eating Practices among Adolescent Girls in Low- and Middle-Income Countries: A Systematic Review. Arlington, VA: Strengthening Partnerships, Results, and Innovations in Nutrition Globally (SPRING) project; 2018. https://www.spring-nutrition.org/publications/reports/diet-and-eating-practices-among-adolescent-girls-low-and-middle-income. Accessed May 1, 2018.

Bundy

de Silva

Horton

Jamison

Patton

, eds. Disease Control Priorities. Washington, DC: World Bank; 2017.

World Health Organization. Global Accelerated Action for the Health of Adolescents (AA-HA!): Guidance to Support Country Implementation. Geneva, Switzerland: World Health Organization; 2017.

Candler

Costa

Heys

Costello

Viner

. Prevalence of thinness in adolescent girls in low- and middle-income countries and associations with wealth, food security, and inequality. J Adolescent Health. 2017;60(4):447–454.e1. doi: 10.1016/j.jadohealth.2016.11.003.

10.

NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128·9 million children, adolescents, and adults. Lancet. 2017;390(10113):2627–2642. doi:10.1016/S0140-6736(17)32129-3

11.

Fleming

Robinson

, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014;384(9945):766–781. doi: 10.1016/S0140-6736(14)60460-8

12.

Wang

Freire

Knable

, et al. Childhood and adolescent obesity and long-term cognitive consequences during aging. J Comp Neurol. 2015;523(5):757–768. doi: 10.1002/cne.23708.

13.

Daniels

. Complications of obesity in children and adolescents. Int J Obesity. 2009;33(S1):S60–S65. doi: 10.1038/ijo.2009.20.

14.

Calcaterra

Klersy

Muratori

, et al. Prevalence of metabolic syndrome (MS) in children and adolescents with varying degrees of obesity. Clin Endocrinol (Oxf). 2008;68(6):868–872. doi: 10.1111/j.1365-2265.2007.03115.x.

15.

Liu

. Health-promoting components of fruits and vegetables in the diet. Adv Nutr. 2013;4(3):384S–392S. doi: 10.3945/an.112.003517.

16.

Mielgo-Ayuso

Valtueña

Huybrechts

, et al. Fruit and vegetables consumption is associated with higher vitamin intake and blood vitamin status among European adolescents. Eur J Clin Nutr. 2017;71(4):458–467. doi: 10.1038/ejcn.2016.232.

17.

Lim

Vos

Flaxman

, et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380(9859):2224–2260. doi: 10.1016/S0140-6736(12)61766-8.

18.

. Resolved: there is sufficient scientific evidence that decreasing sugar-sweetened beverage consumption will reduce the prevalence of obesity and obesity-related diseases. Obes Rev. 2013;14(8):606–619. doi: 10.1111/obr.12040.

19.

Kirkpatrick

Reedy

Kahle

Harris

Ohri-Vachaspati

Krebs-Smith

. Fast-food menu offerings vary in dietary quality, but are consistently poor. Public Health Nutr. 2014;17(4):924–931. doi: 10.1017/S1368980012005563.

20.

O’Donnell

Hoerr

Mendoza

Tsuei Goh

. Nutrient quality of fast food kids meals. Am J Clin Nutr. 2008;88(5):1388–1395. doi: 10.3945/ajcn.2008.26197.

21.

Dunford

Webster

Barzi

Neal

. Nutrient content of products served by leading Australian fast food chains. Appetite. 2010;55(3):484–489. doi: 10.1016/j.appet.2010.08.015.

22.

Marlatt

Farbakhsh

Dengel

Lytle

. Breakfast and fast food consumption are associated with selected biomarkers in adolescents. Prev Med Rep. 2016;3:49–52. doi: 10.1016/j.pmedr.2015.11.014.

23.

Sahoo

Choudhury

Sofi

Kumar

Bhadoria

. Childhood obesity: causes and consequences. J Family Med Prim Care. 2015;4(2):187–192. doi: 10.4103/2249-4863.154628.

24.

Micha

Khatibzadeh

Shi

Andrews

Engell

Mozaffarian

. Global, regional and national consumption of major food groups in 1990 and 2010: a systematic analysis including 266 country-specific nutrition surveys worldwide. BMJ Open. 2015;5(9):e008705. doi: 10.1136/bmjopen-2015-008705.

25.

Caleyachetty

Echouffo-Tcheugui

Tait

Schilsky

Forrester

Kengne

. Prevalence of behavioural risk factors for cardiovascular disease in adolescents in low-income and middle-income countries: an individual participant data meta-analysis. Lancet Diabetes Endo. 2015;3(7):535–544. doi: 10.1016/S2213-8587(15)00076-5.

26.

Vereecken

Pedersen

Ojala

, et al. Fruit and vegetable consumption trends among adolescents from 2002 to 2010 in 33 countries. Eur J Public Health. 2015;25(suppl 2):16–19. doi: 10.1093/eurpub/ckv012.

27.

Lassi

Moin

Das

Salam

Bhutta

. Systematic review on evidence-based adolescent nutrition interventions. Ann N Y Acad Sci. 2017;1393(1):34–50. doi: 10.1111/nyas.13335.

28.

Salam

Hooda

Das

, et al. Interventions to improve adolescent nutrition: a systematic review and meta-analysis. J Adolescent Health. 2016;59(suppl 4):S29–S39. doi: 10.1016/j.jadohealth.2016.06.022

29.

Christian

Smith

. Adolescent undernutrition: global burden, physiology, and nutritional risks. Ann Nutr Metab. 2018;72(4):316–328.

30.

WHO. Guideline: Implementing Effective Actions for Improving Adolescent Nutrition. Geneva,Switzerland: WHO; 2018.

31.

Krebs

Bagby

Bhutta

Z A

, et al. International summit on the nutrition of adolescent girls and young women: consensus statement. Ann N Y Acad Sci. 2017;1400(1):3–7. doi: 10.1111/nyas.13417.

32.

Global Alliance for Improved Nutrition and World Health Organization. Adolescents—Agents of Change for a Well-Nourished World. Geneva, Switzerland; 2018.

33.

SPRING. Report of the Stakeholders Consultation on Adolescent Girls’ Nutrition: Evidence, Guidance and Gaps. Washington, DC: Strengthening Partnerships, Results and Innovations in Nutrition Globally (SPRING) project; 2018.

34.

Save the Children. Adolescent Nutrition: Policy and Programming in SUN+ Countries. London, England: Save the Children Fund; 2015. https://reliefweb.int/sites/reliefweb.int/files/resources/Adolescent_Nutrition.pdf.

35.

HLPE. Nutrition and Food Systems. A Report by the High Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security. Rome, Italy: Author; 2017:152.

36.

Borenstein

Hedges

Higgins

JPT

Rothstein

. Introduction to Meta-Analysis. West Sussex, UK: John Wiley & Sons; 2007.

37.

United Nations, Department of Economic and Social Affairs. World Population Prospects 2017. https://esa.un.org/unpd/wpp/Download/Standard/Population/. 2017. Accessed July 14, 2018.

38.

Barendregt

Doi

Lee

Norman

Vos

. Meta-analysis of prevalence. J Epidemiol Community Health. 2013;67(11):974–978. doi: 10.1136/jech-2013-203104.

39.

Miller

. The inverse of the freeman – Tukey double arcsine transformation. Am Stat. 1978;32(4):138–138. doi: 10.1080/00031305.1978.10479283.

40.

Knapp

Hartung

. Improved tests for a random effects meta-regression with a single covariate. Stat Med. 2003;22(17):2693–2710. doi: 10.1002/sim.1482.

41.

Viechtbauer

. Conducting meta-analyses in R with the metafor package. J Stat Softw. 2010;36(3):1–48.

42.

Holm

. A simple sequentially rejective multiple test procedure. Scand J Stat. 1979;6(2):65–70.

43.

Higgins

JPT

Green

. Cochrane Handbook for Systematic Reviews of Interventions. Hoboken, NJ: John Wiley & Sons; 2011.

44.

Newens

Walton

. A review of sugar consumption from nationally representative dietary surveys across the world. J Hum Nutr Diet. 2016;29(2):225–240. doi: 10.1111/jhn.12338.

45.

World Health Organization. Food Safety Team. Essential Safety Requirements for Street-Vended Foods. Geneva,Switzerland: World Health Organization; 1996. http://apps.who.int/iris/handle/10665/63265. Accessed September 11, 2018.

46.

Alimi

Workneh

. Consumer awareness and willingness to pay for safety of street foods in developing countries: a review. Int J Consum Stud. 2016;40(2):242–248. doi: 10.1111/ijcs.12248.

47.

Caleyachetty

Thomas

Kengne

, et al. The double burden of malnutrition among adolescents: analysis of data from the global school-based student health and health behavior in school-aged children surveys in 57 low- and middle-income countries. Am J Clin Nutr. 2018. doi: 10.1093/ajcn/nqy105.

48.

Golley

Bell

Hendrie

, et al. Validity of short food questionnaire items to measure intake in children and adolescents: a systematic review. J Hum Nutr Diet. 2017;30(1):36–50. doi: 10.1111/jhn.12399.

49.

Gibson

. Principles of Nutritional Assessment. Oxford, England: Oxford University Press; 2005.

50.

Rivera

JÁ

de Cossío

Pedraza

Aburto

Sánchez

Martorell

. Childhood and adolescent overweight and obesity in Latin America: a systematic review. Lancet Diabetes Endo. 2014;2(4):321–332. doi: 10.1016/S2213-8587(13)70173-6.

51.

Bryan

Yeager

Hinojosa

, et al. Harnessing adolescent values to motivate healthier eating. Proc Natl Acad Sci U S A. 2016;113(39):10830–10835. doi: 10.1073/pnas.1604586113.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.65 MB

Global Patterns of Adolescent Fruit,Vegetable,Carbonated Soft Drink,and Fast-Food Consumption: A Meta-Analysis of Global School-Based Student Health Surveys

Abstract

Background:

Objective:

Methods:

Results:

Conclusions:

Keywords

Introduction

Methods

Overview

Data Analysis

Results

Regional Heterogeneity

Heterogeneity Between Income Groups

Heterogeneity Between Food System Typologies

Discussion

Supplemental Material

Supplemental Material, Appendix_(6) - Global Patterns of Adolescent Fruit, Vegetable, Carbonated Soft Drink, and Fast-Food Consumption: A Meta-Analysis of Global School-Based Student Health Surveys

Footnotes

Authors’ Note

Declaration of Conflicting Interests

Funding

ORCID iD

Supplemental Material

References

Supplementary Material