- Email: [email protected]
Hypercholesterolemia and its associated risk factors—Kingdom of Saudi Arabia, 2013
Annals of Epidemiology 24 (2014) 801e808
Contents lists available at ScienceDirect
Annals of Epidemiology journal homepage: www.annalsofepidemiology.org
Original article
Hypercholesterolemia and its associated risk factorsdKingdom of Saudi Arabia, 2013 Mohammed Basulaiman MD a, Charbel El Bcheraoui PhD, MSc b, Marwa Tuffaha MD b, Margaret Robinson BSc b, Farah Daoud BSc b, Sara Jaber MD b, Sarah Mikhitarian BSc b, Shelley Wilson BA b, Ziad A. Memish MD a, Mohammed Al Saeedi MD b, Mohammad A. AlMazroa MD a, Ali H. Mokdad PhD b, * a b
Ministry of Health of the Kingdom of Saudi Arabia, Riyadh, Saudi Arabia Institute for Health Metrics and Evaluation, University of Washington, Seattle
a r t i c l e i n f o
a b s t r a c t
Article history: Received 1 July 2014 Accepted 14 August 2014 Available online 20 August 2014
Purpose: To assess the prevalence of hypercholesterolemia and its associated factors in the Kingdom of Saudi Arabia. Methods: A national multistage representative sample of Saudis aged 15 years or older was surveyed through face-to-face interviews. Data on sociodemographics, risk factors, and health information were collected, and blood sample analysis was performed. Data were analyzed using SAS 9.3 to account for the sample weights and complex survey design. Results: Between April and June 2013, a total of 10,735 participants completed the survey. Overall, 8.5% of Saudis had hypercholesterolemia. Another 19.6% had borderline hypercholesterolemia. Among hypercholesterolemic Saudis, 65.1% were undiagnosed, 2.3% were treated uncontrolled, 28.3% were treated controlled, and 4.3% were untreated. The risk of being hypercholesterolemic increased with age and among individuals who reported consuming margarine, obese individuals, and those who have been previously diagnosed with hypertension or diabetes. Conclusions: More than a million Saudis have hypercholesterolemia, and 700,000 of them are unaware of their condition which can be controlled through early detection campaigns and lifestyle change and medication. An urgent awareness and screening campaign is needed in Kingdom of Saudi Arabia to prevent and avoid disease progression toward more serious stages. Ó 2014 Elsevier Inc. All rights reserved.
Keywords: Hypercholesterolemia Saudi Arabia Noncommunicable diseases
Introduction Cholesterol is essential for the body, specifically for hormone synthesis, fat digestion, and formation of cell membranes [1]. However, large amounts of cholesterol in the bloodstream lead to myocardial infarction, stroke, ischemic injury, and organ failure [2]. Hypercholesterolemia can be prevented, and measures such as dietary changes and clinical follow-up must be taken for high-risk individuals to reduce adverse events [3e7]. The Global Burden of Disease 2010 study estimated that high cholesterol is a leading risk factor for death from cardiovascular diseases in the Kingdom of Saudi Arabia (KSA). It accounted for Conflicts of interest: The salaries of the authors from the Institute for Health Metrics and Evaluation were fully or partially paid through a grant from the Ministry of Health (MOH) of the Kingdom of Saudi Arabia. The MOH managed the data collection for this study. * Corresponding author. Department of Global Health, Institute for Health Metrics and Evaluation, University of Washington, 2301 5th Avenue, Suite 600, Seattle, WA 98121. Tel.: þ1 206 897 2849; fax: þ1 206 897 2899. E-mail address: [email protected] (A.H. Mokdad). 1047-2797/$ e see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.annepidem.2014.08.001
about 5.14% of total deaths, 3.96% of years of life lost, and 1.99% of disability-adjusted life years. From 1990 to 2010, the burden of hypercholesterolemia remained very high in KSA as it ranked eighth among the risk factors for total deaths [8]. Previous studies reported on prevalence of hypercholesterolemia in KSA. In the early 80s, prevalence of hypercholesterolemia was 7% for males and 8% for females and increased with age and body mass index (BMI) [9,10]. Since the late 90s, data on hypercholesterolemia are limited [11]. However, prevalence of hypercholesterolemia was reported as 18.6% for males and 19.7% for females in 2005 [12]. To assess the current status of hypercholesterolemia in KSA, we analyzed the 2013 Saudi Health Interview Survey (SHIS) to describe the magnitude of hypercholesterolemia in KSA by different sociodemographic groups and risk factors. Materials and methods SHIS is a national multistage survey of individuals aged 15 years or older. Households were randomly selected from a national sampling
802
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
frame maintained and updated by the Census Bureau. KSA was divided into 13 regions. Each region was divided into subregions and blocks. All regions were included, and a probability proportional to size was used to randomly select subregions and blocks. Households were randomly selected from each block. A roster of household members was collected, and an adult aged 15 years or older was randomly selected to be surveyed. Weight, height, and blood pressure were measured at the household by a trained professional. An Omron HN286 scale (SN: 201207-03163F) and Omron M6 Comfort monitor (HEM-7223-E) were used to measure weight and blood pressure. The survey included questions on sociodemographic characteristics, including age, sex, education, and marital status, tobacco consumption, diet, physical activity, health care utilization, different health-related behaviors, and self-reported chronic conditions. We used measured weight and height to calculate BMI as weight (kilogram)/height2 (square meter). Participants were classified into four groups: (1) underweight, BMI less than18.5; (2) normal weight, BMI 18.5 to 24.9; (3) overweight, BMI 25.0 to 29.9; or (4) obese, BMI greater than or equal to 30.0. Respondents were considered to be current smokers if they reported ever smoking any tobacco products and still currently smoking tobacco, and were considered past smokers if they reported smoking in the past but not anymore. We assessed the type of fat most consumed by asking respondents: “What type of oil or fat is most often used for meal preparation in your household?” Respondents could choose between vegetable oils, olive oil, butter or ghee, margarine, or none in particular. Vegetable and olive oil were combined into one category of vegetable oil, and butter and ghee were considered animal fat. We computed the servings of fruits and vegetables and red meats and chicken consumed per day from the detailed dietary questionnaire as the sum of the average daily consumption of fruits, fruit juices, and vegetables, and red meats and chicken. We used the International Physical Activity questionnaire [13] to classify respondents into four groups of physical activity: (1) met vigorous physical activity level, (2) met moderate physical activity level, (3) insufficient physical activity to meet vigorous or moderate levels, and (4) no physical activity. To calculate the time spent on watching television, respondents were asked: “In a typical week, how much time do you usually spend in front of the television or on the computer?” Time was recorded in minutes or hours and then computed to create four categories in hours. To assess diagnosed hypertension, prediabetes, diabetes, and hypercholesterolemia status, respondents were asked four separate questions: “Have you ever been told by a doctor, nurse, or other
health professional that you had: (1) high blood pressure, otherwise known as hypertension; (2) prediabetes mellitus, otherwise known as prediabetes, borderline diabetes, impaired fasting glucose, impaired glucose tolerance, or impaired sugar tolerance; (3) diabetes mellitus, otherwise known as diabetes, sugar diabetes, high blood glucose, or high blood sugar; (4) hypercholesterolemia, otherwise known as high or abnormal blood cholesterol?” Women diagnosed with diabetes or hypertension only during pregnancy were counted as not having these conditions. Those who were diagnosed with either of these conditions were further asked if they are currently receiving any treatment for their condition. Similarly, the same types of questions were used to determine previous diagnosis of stroke, myocardial infarction, atrial fibrillation, cardiac arrest, congestive heart failure, chronic obstructive pulmonary disease, asthma, renal failure, and cancer. We considered a person to be diagnosed with a chronic condition if they reported being diagnosed with any of the conditions cited earlier. Respondents who reported being diagnosed with hypercholesterolemia were asked: “During the past 30 days, or since your diagnosis, have you ever taken medication for this condition?” Those who reported being on treatment were asked to provide the name of the drugs they are using. Respondents who completed the questionnaire were invited to local primary health care clinics to provide a blood sample for laboratory analysis. A Roche Hitachi COBAS 8000 system was used to measure cholesterol. Respondents were considered to be hypercholesterolemic if they met any of the following criteria: (1) measured cholesterol equal to or exceeding 6.2 mmol/L or (2) measured cholesterol not exceeding 6.2 mmol/L, but the respondent reported taking medications for hypercholesterolemia. Respondents were considered to have borderline hypercholesterolemia if (1) they did not report taking drugs for hypercholesterolemia and (2) their measured blood cholesterol level was between 5.18 and 6.2 mmol/L. Respondents under treatment for hypercholesterolemia were considered controlled if their measured cholesterol levels were below 6.2 mmol/L. Levels of high-density lipoproteins (HDL), low-density lipoproteins (LDL), and triglycerides were also measured. Respondents were considered to have low HDL if their measured HDL was below 1.0 mmol/L or 1.3 mol/L for males and females, respectively. Respondents were considered to have high LDL if their measured LDL was greater or equal to 4.13 mmol/L. They were considered to have hypertriglyceridemia if their measured triglycerides level was greater or equal to 2.3 mmol/L.
Table 1 Sociodemographic characteristics of Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y)
Marital status
Education
Categories
Males Females 15e24 25e34 35e44 45e54 55e64 65þ Currently married Never married Separated, divorced, or widowed Primary school* or less Elementaryy or high school completed College degree or higher education
SE ¼ Standard error. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
Complete sample
Clinical examination sample
N
Weighted %
SE
N
Weighted %
SE
5253 5482 2382 2757 2339 1520 862 875 3286 4872 2557 6976 2829 897
50.64 49.36 40.31 21.50 15.18 12.38 6.47 4.17 26.32 52.83 20.85 49.35 45.87 4.78
0.69 0.69 0.73 0.52 0.41 0.40 0.29 0.20 0.58 0.68 0.53 0.69 0.71 0.23
2576 3014 1163 1299 1241 860 485 542 2024 2325 1236 3657 1423 492
52.86 47.14 42.32 22.05 14.35 11.51 6.03 3.74 28.39 51.92 19.69 46.52 49.74 3.75
1.08 1.08 1.14 0.84 0.58 0.55 0.40 0.24 0.91 1.07 0.82 1.05 1.08 0.27
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
803
Table 2 Sociodemographic characteristics of diagnosed hypercholesterolemic, undiagnosed hypercholesterolemic, hypercholesterolemic, and borderline hypercholesterolemic, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Diagnosed hypercholesterolemic N
Sex Males Females Age (y) 15e24 25e34 35e44 45e54 55e64 65þ Marital status Currently married Never married Separated, divorced, or widowed Education Primary school* or less Elementaryy or high school completed College degree or higher education Smoking status Never smoked Ex-smoker Current smoker Type of fat consumed Vegetable oils Margarine Animal fat None in particular Daily servings of fruits and vegetables 0 0e3 3e5 5þ Daily servings of meat and chicken 0e1 1e2 2e3 3þ Daily hours spent on watching television 0e1 1e3 3e5 5þ Level of physical activity None Low Moderate High Obesity Not obese Obese History of diagnosis with hypertension No Yes History of diagnosis with prediabetes No Yes History of diagnosis with diabetes No Yes Diagnosis of chronic condition No Yes
Weighted %
Hypercholesterolemic undiagnosed SE
N
Weighted %
Hypercholesterolemic SE
N
Weighted %
Borderline hypercholesterolemic SE
N
Weighted %
SE
434 325
6.50 4.33
0.41 0.32
174 186
6.37 5.36
0.76 0.63
330 314
9.54 7.29
0.80 0.64
542 614
19.00 20.28
1.16 1.22
4 51 139 195 181 189
0.18 1.66 5.98 14.81 20.09 25.12
0.10 0.29 0.64 1.24 1.84 2.22
46 85 93 67 38 31
3.78 5.75 7.82 10.16 7.97 7.59
0.86 0.82 1.13 1.66 1.82 1.82
46 92 133 130 114 129
3.54 5.66 10.93 16.27 20.18 28.70
0.80 0.78 1.27 1.85 2.44 2.91
129 269 300 226 109 123
12.03 21.67 27.56 29.32 27.61 20.51
1.40 1.73 1.88 2.26 3.32 2.56
596 23 140
9.14 0.49 15.54
0.47 0.14 1.70
259 62 38
7.71 4.08 8.19
0.66 0.78 1.77
484 68 90
12.45 4.02 18.40
0.76 0.73 2.48
870 167 114
27.70 11.84 20.84
1.16 1.22 2.67
388 226 143
9.59 3.47 5.17
0.66 0.30 0.54
131 136 93
7.86 4.73 6.17
1.08 0.68 0.87
301 214 129
12.81 6.34 7.95
1.14 0.69 0.94
417 439 297
20.56 17.08 24.55
1.49 1.18 1.97
577 67 113
4.86 12.18 7.01
0.27 1.91 0.88
299 12 47
5.21 5.95 10.84
0.48 2.30 2.28
520 38 83
7.59 10.23 14.34
0.52 2.40 2.28
968 61 124
18.76 26.02 23.23
0.91 4.01 2.68
680 19 17 31
5.41 8.52 4.73 5.10
0.28 0.28 1.36 1.11
300 13 13 29
5.51 10.22 11.58 8.38
0.53 3.26 5.28 1.93
543 23 20 45
7.93 15.15 14.95 11.39
0.55 3.53 5.07 2.06
989 32 33 76
19.40 18.23 29.78 18.47
0.91 5.10 5.95 2.49
24 536 108 72
4.68 4.75 8.17 8.42
1.29 0.28 0.99 1.25
10 276 47 19
5.25 6.22 5.53 2.90
1.90 0.61 1.17 0.98
18 472 88 44
5.57 8.42 9.62 7.14
1.75 0.62 1.40 1.50
38 850 153 78
19.24 19.46 20.32 18.24
3.90 0.99 2.27 2.93
234 264 126 123
5.69 4.86 6.22 5.32
0.50 0.39 0.75 0.62
119 124 55 50
7.04 5.54 5.18 5.02
0.99 0.81 1.06 1.17
201 216 104 100
9.55 7.68 8.38 7.78
1.02 0.83 1.20 1.24
354 421 179 173
21.41 21.34 18.30 14.24
1.60 1.50 1.98 1.70
67 269 172 126
11.07 6.70 4.87 4.64
1.64 0.52 0.51 0.58
41 113 72 56
7.49 4.88 6.16 9.62
1.76 0.65 0.98 2.11
80 213 142 86
13.57 8.22 9.23 10.23
2.15 0.78 1.10 1.97
82 401 229 181
16.07 19.16 19.01 23.23
2.42 1.38 1.92 2.40
317 189 112 141
5.92 6.04 6.47 3.70
0.45 0.58 0.81 0.40
152 89 36 83
7.60 5.58 4.56 5.24
0.90 1.00 1.11 0.95
277 157 83 127
11.57 7.71 8.21 6.44
1.00 1.01 1.30 0.93
465 308 141 242
21.79 21.69 16.58 17.05
1.42 1.81 2.04 1.51
324 405
3.40 10.35
0.25 0.67
218 128
5.35 6.87
0.62 0.81
328 286
6.79 12.05
0.62 0.96
660 460
18.26 23.00
1.04 1.49
420 337
3.28 34.74
0.21 2.09
331 28
5.94 5.52
0.53 1.37
462 177
7.11 28.78
0.53 2.56
1006 135
19.21 23.79
0.88 2.69
651 64
4.79 32.70
0.25 4.41
349 6
6.00 2.28
0.51 1.09
578 36
8.13 24.17
0.53 4.88
1101 28
19.66 18.96
0.86 4.69
379 377
3.13 30.95
0.21 1.83
311 46
5.91 5.95
0.54 1.08
421 216
7.01 25.85
0.53 2.19
975 159
18.96 24.59
0.89 2.33
602 157
4.53 19.95
0.24 2.07
343 17
5.91 5.70
0.51 2.40
572 67
8.12 13.05
0.53 2.60
1084 68
19.17 24.81
0.85 4.06
SE ¼ Standard error. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
Weighting methodology Two sets of sample weights were generated and incorporated into the data set for analysis and extrapolation to the Saudi
population. First, we created an individual weight for all respondents to account for (1) the probability of selection of an eligible respondent within a household, (2) the probability of selection of the household within a stratum, and (3) the poststratification
804
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
differences in age and sex distribution between the sample and the Saudi population. For individuals who completed the laboratory-based blood analysis, we computed an additional weight used in analyzing data from clinic visits to account for (1) the individual weight described previously, (2) the probability of visiting a clinic, (3) sociodemographic, behavioral, and health differences between respondents who visited the clinic and those who did not, and (4) the poststratification differences in age and sex distribution between the respondents who visited the clinic and the general Saudi population. Statistical analysis We used a multivariate logistic regression model to measure association between outcome variables and sociodemographic factors first. Then, we used a backward elimination multivariate logistic regression model to measure the association between outcome variables and all associated factors. All factors were first included in the models. Then variables were eliminated based on a Wald c2 test for the analysis of effect. Variables were removed one by one based on the significance level of their effect on the model, starting with the variable with the highest P > .5, till all variables kept had a P .5 in the analysis of effect. We used SAS 9.2 (SAS Institute Inc., Cary, NC) for the analyses and to account for weights and the complex sampling design. Results Between April and June 2013, a total of 10,735 participants completed the SHISdresponse rate of 89.4%dand were invited to the local health clinics. The remaining 1265 participants completed part of the household enumeration, or all of it, but the selected adult did not complete the survey. A total of 5590 individuals went to the local clinics and provided blood samples for analysesda response rate of 52.1%. The characteristics of respondents who completed the questionnaire and the laboratory examination are presented in Table 1. Overall, an estimated 693,106 Saudis (5.4%) reported being diagnosed with hypercholesterolemia. However, an estimated 1,107,296 Saudis (8.5%) aged 15 years or older had hypercholesterolemia as measured by our laboratory tests with blood cholesterol level greater or equal to 6.2 mmol/L or reported currently taking cholesterol medication. Of these, 65.1% were undiagnosed or unaware of their condition. Moreover, 19.6% of Saudis, an estimated 2,397,211 participants, had borderline hypercholesterolemia with measured blood cholesterol levels between 5.18 and 6.2 mmol/L and not currently taking blood cholesterol lowering medication. Characteristics of respondents with undiagnosed hypercholesterolemia, hypercholesterolemia, and borderline hypercholesterolemia are presented in Table 2. Among participants diagnosed with hypercholesterolemia, 92.9% reported taking medication for their condition. More than 95% of those being treated were taking statins for their condition. The vast majority of hypercholesterolemic patients (92.6%) who were on medication had their blood cholesterol level controlled. Among those who were hypercholesterolemic, 65.1% were undiagnosed, 2.3% were treated uncontrolled, 28.3% were treated controlled, and 4.3% were untreated (Fig. 1). A high percent of Saudis (48.7%) have low HDL, 7.4% have high LDL, and 8.5% have hypertriglyceridemia. Distributions of these three conditions by sociodemographic and risk factors are presented in Table 3. Age, type of fat mostly consumed, obesity, and diagnosis history of hypertension and diabetes were associated with hypercholesterolemia (Table 4). The risk of being hypercholesterolemic
increased with age (adjusted odds ratios [AOR] ¼ 1.03; 95% confidence intervals [CI], 1.01e1.04), among individuals who reported consuming margarine (AOR ¼ 2.17; 95% CI, 1.17e4.00), obese individuals (AOR ¼ 1.40; 95% CI, 1.03e1.89), and those who have been previously diagnosed with hypertension (AOR ¼ 1.82; 95% CI, 1.31e2.53) and diabetes (AOR ¼ 1.85; 95% CI, 1.28e2.69). On the other hand, sex, marital status, education, smoking status, time spent on watching television, and level of physical activity were not associated with the risk of hypercholesterolemia (Table 4). The risk of borderline hypercholesterolemia increased with age and time spent on watching television (Table 5). The risk of being undiagnosed when one is hypercholesterolemic was associated only with age (Table 6). Older individuals were more likely to be undiagnosed while hypercholesterolemic (AOR ¼ 1.02; 95% CI, 1.01e1.03).
Discussion To our knowledge, this is the first national study to report on cholesterol levels in KSA. We found high rates of undiagnosed cholesterol and borderline levels in a country where chronic diseases are the major health burden. Our findings call for awareness campaigns to educate the public about the danger of high cholesterol levels and the importance of early diagnosis. Indeed, because Saudis have free medical care, receive free medication when needed, and our data show high control levels of cholesterol once on medication, our recommendation will reduce the burden of cholesterol-related outcomes in KSA. Moreover, if these campaigns include messages on changing health behaviors such as diet and physical activity, they will have a larger impact on health. Our findings about lack of diagnosis and treatment of hypercholesterolemia in KSA have been previously reported; the majority of hypercholesterolemic Saudi individuals are unaware of their condition, and hypercholesterolemia remains undertreated [14]. In KSA, ischemic heart disease is the leading cause of death [8] as is the case for the whole Arab world, geographically comprised 22 countries known as the Arab league [15], and it has been reported that hypercholesterolemia is a major risk factor for this disease, both in KSA and worldwide [16,17]. Hypercholesterolemia has been reported as the eighth risk factor for death in the Arab World in Global Burden of Disease 2010 study. In our study, type of fat consumed, obesity, and sedentary activities were associated with hypercholesterolemia and borderline hypercholesterolemia. Hence, health authorities in KSA should develop public health campaigns that promote healthier lifestyles, encourage people to exercise, and make informed decisions around their eating habits.
Fig. 1. Percent distribution of diagnosis and treatment status among hypercholesterolemic, Saudis aged 15 years older, 2013.
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
805
Table 3 Distribution of HDL, LDL, and triglycerides by sociodemographic and risk factors Sociodemographic and risk factors
Sex Males Females Age (y) 15e24 25e34 35e44 45e54 55e64 65þ Marital status Currently married Never married Separated, divorced, or widowed Education Primary school* or less Elementaryy or high school completed College degree or higher education Smoking status Never smoked Ex-smoker Current smoker Type of fat consumed Vegetable oils Margarine Animal fat None in particular Daily servings of fruits and vegetables 0 0e3 3e5 5þ Daily servings of meat and chicken 0e1 1e2 2e3 3þ Daily hours spent on watching television 0e1 1e3 3e5 5þ Level of physical activity None Low Moderate High Obesity Not obese obese History of diagnosis with hypertension No Yes History of diagnosis with prediabetes No Yes History of diagnosis with diabetes No Yes
HDL
LDL
Hypertriglyceridemia
Not diagnosed or treated
Diagnosed and treated Not diagnosed or treated
Diagnosed and treated
Not diagnosed or treated
Diagnosed and treated
N
N
N Weighted % SE
N
N Weighted % SE
Weighted % SE
947 38.89 1717 59.77
Weighted % SE
1.57 1.56
59 39.84 72 63.77
45.33 47.91 53.42 54.74 56.31 50.04
2.04 2.11 2.09 2.57 3.79 3.66
0 2 11 31 38 49
1741 53.16 665 44.39 250 58.43
1.26 1.86 4.15
997 55.29 1068 46.10
597 47.34
558 617 641 428 214 206
N
Weighted % SE
8.29 6.67
0.87 0.84
6 5
4.74 5.44
2.42 249 3.20 185
3.69 7.52 11.62 13.85 13.32 13.53
0.88 1.17 1.62 2.10 2.73 3.07
0 0 4 3 0 4
0 0 9.06 6.84 0 5.07
0 43 4.14 0 81 7.13 5.99 121 13.48 4.58 93 15.07 0 53 18.78 3.35 43 12.54
0.86 0 1.08 1 1.56 3 2.01 7 3.21 8 2.49 18
0 12.84 8.40 11.44 14.81 15.05
0 13.78 5.13 5.98 7.14 4.06
95 43.54 1 44.10 35 75.36
5.22 263 11.01 31.99 60 4.31 7.93 41 11.89
0.93 0.82 2.54
7 0 4
4.84 0 6.84
2.25 348 12.99 0 57 4.27 4.31 28 9.34
0.92 30 12.40 0.78 0 0 2.24 7 19.66
3.36 0 7.48
1.92 1.66
85 44.51 31 58.42
5.89 136 10.37 8.90 144 5.88
1.36 0.78
5 5
3.26 7.46
1.97 161 4.39 168
1.11 24 12.58 0.79 9 16.65
2.90 7.88
2.37
15 39.78
8.15
1.23
1
6.02
5.87 104 10.65
1.44
7.35
4.47
0 63.98 45.03 48.37 53.99 43.94
5.94 194 6.31 171
Weighted % SE
0 29.86 14.22 10.30 8.37 7.05
10.18
41 79 95 68 47 35
85
9.93 6.52
9.87 6.61
0.91 23 12.85 0.72 14 13.60
4
4.08 3.87
2359 49.47 82 39.92 221 47.29
1.22 102 46.52 5.29 12 40.31 3.48 17 60.60
5.29 300 6.75 14.77 16 8.94 12.61 47 12.37
0.61 2.87 2.50
8 0 3
4.96 0 7.84
2.24 341 6.94 0 26 13.73 6.46 65 16.91
0.56 26 11.72 3.97 6 12.22 2.69 5 21.32
2.80 6.53 12.77
2297 81 75 172
48.83 53.28 54.50 46.79
1.23 117 49.06 5.95 4 35.31 6.52 4 73.39 3.80 3 31.54
5.17 303 7.11 18.16 16 11.88 18.94 10 12.67 19.29 29 10.78
0.64 11 3.64 0 6.03 0 2.51 0
6.21 0 0 0
2.47 373 0 14 0 6 0 32
8.39 6.24 5.58 9.49
0.65 35 15.07 2.28 0 0 3.00 0 0 1.97 1 7.01
3.51 0 0 7.05
71 2057 322 178
49.90 49.05 46.31 48.56
5.96 1.30 3.16 4.29
4 98 17 7
35.54 49.72 39.73 47.67
26.00 8 5.67 283 10.72 49 15.72 21
4.78 7.74 8.07 5.77
1.92 0 0.72 10 1.74 0 1.59 1
0 6.95 0 1.52
0 10 2.83 332 0 54 1.57 30
6.71 8.49 8.01 7.51
2.42 0 0 0.71 28 12.41 1.42 5 8.68 1.82 4 30.16
0 3.01 4.25 16.53
789 974 415 458
49.40 49.48 46.92 47.83
2.05 1.88 2.88 2.68
40 41 29 19
42.97 49.47 56.79 46.86
8.63 119 8.00 134 11.18 65 10.90 43
8.61 7.79 7.18 5.48
1.03 1.13 1.34 1.34
2 4 2 3
6.21 3.74 5.48 4.24
4.39 123 2.94 159 4.32 76 2.83 70
7.57 8.95 9.51 7.34
0.96 8 12.16 1.05 15 13.38 1.61 5 13.43 1.36 9 14.92
6.07 4.08 8.92 5.59
186 915 543 399
53.74 48.64 45.28 48.43
4.77 1.93 2.46 2.88
15 48 40 8
35.00 58.57 40.63 24.60
10.12 36 8.72 7.67 122 7.07 8.08 82 8.62 10.04 59 10.77
2.09 0.99 1.34 2.20
3 12.27 1 0.45 2 4.42 2 17.04
7.89 29 0.46 150 3.69 96 13.52 57
6.48 8.41 7.99 8.43
1.63 4 12.89 0.99 16 16.65 1.23 9 12.52 1.70 4 19.89
6.48 6.09 5.56 10.00
1091 674 316 583
58.45 50.70 43.46 40.98
1.78 2.20 3.05 2.12
53 37 20 21
48.67 51.93 32.64 59.90
7.71 141 8.21 102 9.08 37 11.77 85
9.58 7.81 5.79 6.30
1.26 1.23 1.30 0.99
5 3 1 2
3.98 8.66 3.78 2.65
2.60 158 5.39 114 3.77 54 1.93 108
8.47 8.48 6.87 8.67
0.91 12 12.76 1.25 9 10.49 1.31 10 16.54 1.16 6 14.27
5.04 3.69 7.96 6.65
1575 45.81 1005 56.58
1.41 1.87
48 41.04 79 55.01
7.03 207 6.18 147
6.41 9.81
0.70 1.14
7 4
7.05 3.64
3.55 244 6.73 2.32 183 12.46
0.66 11 6.20 1.27 24 19.18
2.27 5.09
2432 48.54 220 56.74
1.17 3.72
59 43.08 72 51.96
7.50 328 7.42 5.80 35 10.12
0.63 2.30
6 5
5.75 4.44
3.10 389 8.09 2.53 45 15.53
0.61 10 5.37 2.49 27 18.71
2.02 4.64
2563 48.63 54 59.44
1.15 111 47.97 7.27 16 61.82
0.62 10 1.55 0
5.01 0
2.13 410 8.12 0 13 18.82
0.60 30 13.96 5.98 5 7.10
3.48 3.87
2356 48.62 289 55.54
1.19 2.98
0.64 1.90
0.89 8.34
0.66 341 7.51 3.50 92 21.38
0.61 10 10.13 2.48 27 15.34
4.10 4.18
47 51.98 84 45.39
5.26 351 11.03 8
7.64 3.68
7.51 317 7.35 5.82 43 10.24
2 9
(continued on next page)
806
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
Table 3 (continued ) Sociodemographic and risk factors
HDL
LDL
Hypertriglyceridemia
Not diagnosed or treated
Diagnosed and treated Not diagnosed or treated
Diagnosed and treated
Not diagnosed or treated
Diagnosed and treated
N
N
N Weighted % SE
N
N
Weighted % SE
Diagnosis of chronic condition No 2526 48.93 Yes 138 47.08
Weighted % SE
1.16 102 46.79 4.65 29 53.62
N
5.19 347 11.01 18
Weighted % SE 7.53 7.56
0.62 10 2.90 1
5.60 2.48
Weighted % SE
2.38 406 8.18 2.48 28 10.42
Weighted % SE
0.60 24 10.03 2.85 13 25.87
2.64 10.08
SE ¼ Standard error. * Six years of schooling following kindergarten. y Three years of schooling following primary school.
Our findings showed that older individuals should be targeted for early detection and encouraged to change their lifestyle or seek treatment, especially because health care and medications in KSA is free and readily available to citizens. As seen in our data, most Saudis treated for hypercholesterolemia are prescribed statins, and the fact that 92% of treated subjects are able to control their cholesterol blood level is an encouraging finding in our study. Moreover, the risk of hypercholesterolemia increased among diabetic and hypertensive patients. Screening for hypercholesterolemia should be prioritized among these patients as comorbid individuals who are at higher risk of serious health complications. The last reported study of hypercholesterolemia rates in KSA was in 2005 and revealed a 19.1% prevalence of hypercholesterolemiadcutoff point of 5.2 mmol/Ldamong Saudis aged 15 to 64 years [12]. When restricted to the same age groups and using the same cutoff point, our data revealed a prevalence of 25.3%. However, in our survey, we applied poststratification weights to
reflect the Saudi population and adjusted for the increased probability of sick individuals to agree to visit the health facility for physical and blood samples. This was not done for the previous survey. Indeed, when we did not apply adjustments based on poststratification and the predicted probability of completing the clinic visit; our hypercholesterolemia estimates were 30.1% (95% CI, 29.0e32.4). Clearly, this indicates a rise in the prevalence of cholesterol in KSA and calls for programs to prevent and control cholesterol levels given its role as a major risk factor for ischemic heart disease, the leading cause of death in KSA [8]. Our study has some limitations. First, our data are from a crosssectional study; hence, we cannot assess causality. Second, many of our behavioral data, such as diet and physical activity, are selfreported and subject to recall and social desirability biases. On the other hand, our study is based on a large sample size and used a standardized methodology for all its measures. Third, only 52% of respondents completed the visit to the health clinics and had their blood drawn for analysis. However, our weighting methodology
Table 4 Multivariate logistic regression for sociodemographic characteristics, risk factors, and hypercholesterolemia, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y) Marital status
Education
Smoking status
Type of fat consumed
Daily hours spent on watching television
Level of physical activity
Obesity History of diagnosis with hypertension History of diagnosis with diabetes AOR ¼ Adjusted odds ratio; CI ¼ Confidence interval. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
Categories
Males Females Currently married Never married Separated, divorced, or widowed Primary school* or less Elementaryy or high school completed College degree or higher education Never smoked Ex-smoker Current smoker Vegetable oils Margarine Animal fat None in particular 0e1 1e3 3e5 5þ None Low Moderate High Not obese Obese No Yes No Yes
Sociodemographic model
Full model
AOR
95% CI
AOR
95% CI
d 0.73 1.04 d 0.71 1.23 d 0.97 1.00 d d d d d d d d d d d d d d d d d d d d d
d 0.56e0.95 1.03e1.05 d 0.47e1.06 0.83e1.82 d 0.65e1.44 0.66e1.50 d d d d d d d d d d d d d d d d d d d d d
d 0.72 1.03 d 0.79 1.49 d d d d 0.62 1.70 d 2.17 1.00 1.25 d 0.82 1.14 1.19 d 0.73 0.83 0.68 d 1.40 d 1.85 d 1.88
d 0.50e1.03 1.01e1.04 d 0.49e1.27 0.96e2.33 d d d d 0.28e1.34 0.98e2.94 d 1.17e4.04 0.34e2.95 0.72e2.19 d 0.52e1.27 0.70e1.87 0.68e2.06 d 0.48e1.12 0.52e1.33 0.43e1.09 d 1.03e1.89 d 1.28e2.69 d 1.31e2.70
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
807
Table 5 Multivariate logistic regression for sociodemographic characteristics, risk factors, and borderline hypercholesterolemia, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y) Marital status
Education
Type of fat consumed
Daily servings of meat and chicken
Daily hours spent on watching television
Level of physical activity
History of diagnosis with prediabetes History of diagnosis with diabetes Diagnosis of chronic condition
Categories
Males Females Currently married Never married Separated, divorced, or widowed Primary* school or less Elementaryy or high school completed College degree or higher education Vegetable oils Margarine Animal fat None in particular 0e1 1e2 2e3 3þ 0e1 1e3 3e5 5þ None Low Moderate High No Yes No Yes No Yes
Sociodemographic model
Full model
AOR
95% CI
AOR
95% CI
d 1.09 1.02 d 0.43 0.67 d 1.32 1.63 d d d d d d d d d d d d d d d d d d d d d d
d 0.86e1.37 1.01e1.03 d 0.31e0.60 0.46e0.98 d 0.98e1.78 1.17e2.26 d d d d d d d d d d d d d d d d d d d d d d
d 1.02 1.02 d 0.43 0.69 d 1.19 1.29 d 1.25 1.96 1.09 d 1.00 0.76 0.74 d 1.44 1.66 2.00 d 1.00 0.73 0.81 d 0.54 d 1.19 d 1.31
d 0.76e1.35 1.00e1.03 d 0.29e0.64 0.44e1.10 d 0.84e1.69 0.88e1.91 d 0.54e2.87 0.99e3.87 0.69e1.73 d 0.73e1.38 0.51e1.12 0.49e1.11 d 0.96e2.17 1.05e2.63 1.25e3.20 d 0.72e1.39 0.47e1.12 0.57e1.14 d 0.27e1.08 d 0.80e1.75 d 0.65e2.62
AOR ¼ Adjusted odds ratio; CI ¼ Confidence interval. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
accounted for this bias by applying a poststratification adjustment using sociodemographic characteristics, health behavior, previously diagnosed noncommunicable diseases, and anthropometric measurements of respondents from the household survey. The prevalence of hypercholesterolemia in KSA is not as high as in many developing countries but is a shared risk factor for death with the remaining Arab World [15]. For instance, in Switzerland, 17.4% of adults aged 15 years or older were selfreportedly hypercholesterolemic in 2007 [18], double the prevalence of KSA. In the United States in 2005 to 2008, 23% of adults aged 20 to 64 years had hypercholesterolemia, defined similarly
to our study [19]. In KSA, only 6.7% of adults of the same age group were affected in 2013. On the other hand, the prevalence in KSA is similar to its neighboring countries. For example, in Kuwait, about 7% of individuals aged 20 to 69 years had blood cholesterol levels at or above 6.2 mmol/L in 2009 [20]. We found a large prevalence of undiagnosed hypercholesterolemia, a major risk factor for ischemic heart disease, the leading cause of death in KSA [8]. Despite its relatively low prevalence, the issue of hypercholesterolemia should be addressed to reduce the risk of ischemic heart disease and death, especially because Saudi patients have been successful in lowering their blood cholesterol levels.
Table 6 Multivariate logistic regression for sociodemographic characteristics, risk factors, and undiagnosis of hypercholesterolemia, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y) Marital status
Education
History of diagnosis with prediabetes History of diagnosis with diabetes History of diagnosis with hypertension AOR ¼ Adjusted odds ratio; CI ¼ Confidence interval.
Categories
Males Females Currently married Never married Separated, divorced, or widowed Primary school or less Elementary or high school completed College degree or higher education No Yes No Yes No Yes
Sociodemographic model
Full model
AOR
95% CI
AOR
95% CI
d 0.75 1.01 d 0.64 1.04 d 0.74 0.85 d d d d d d
d 0.54e1.04 1.00e1.03 d 0.41e1.01 0.61e1.78 d 0.45e1.21 0.52e1.38 d d d d d d
d 0.78 1.02 d 0.68 1.07 d d d d 0.27 d 0.80 d 0.68
d 0.54e1.13 1.01e1.03 d 0.43e1.07 0.62e1.84 d d d d 0.09e0.79 d 0.49e1.31 d 0.37e1.27
808
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
Indeed, concerned Saudi health professionals are already calling for steps, such as food labeling, to prevent cardiovascular diseases [21]. Conclusions Our findings call for urgent measures by the Saudi Ministry of Health to reduce the burden of hypercholesterolemia in KSA. Targeted blood cholesterol screening campaigns are required to detect individuals with hypercholesterolemia and refer them for medical follow-up. Moreover, the early detection campaigns should offer those with borderline hypercholesterolemia means to control their cholesterol levels through behavioral changes or medication if they have other risk factors. In addition, these early detection campaigns should include screening for blood pressure or diabetes to maximize their benefits. Finally, educational campaigns are needed to promote healthier lifestyles through physical activity and healthy eating habits. Indeed, changing health behaviors among Saudis will have a tremendous impact on the burden of disease in KSA. Acknowledgment This study was financially supported by a grant from the Ministry of Health (MOH) of the Kingdom of Saudi Arabia. The authors would like to acknowledge Adrienne Chew at the Institute for Health Metrics and Evaluation, Seattle, WA for editing this paper. References [1] Widmaier E, Raff H, Strang K. Vander’s Human physiology: the mechanisms of body function. 13th ed. New York: McGraw-Hill Science/Engineering/Math; 2013. [2] Kashyap ML. Cholesterol and atherosclerosis: a contemporary perspective. Ann Acad Med Singapore 1997;26:517e23.
[3] Ziegler O, Got I, Jan P, Drouin P. [Diet therapy of hypercholesterolemia. From theory to practice]. Ann Cardiol Angéiol 1989;38:249e53. [4] Darnton-Hill I, Nishida C, James WPT. A life course approach to diet, nutrition and the prevention of chronic diseases. Public Health Nutr 2004;7:101e21. [5] Diet, nutrition and the prevention of chronic diseases. World Health Organ Tech Rep Ser 2003;916:ieviii. 1e149, backcover. [6] Griffin BA. Nonpharmacological approaches for reducing serum low-density lipoprotein cholesterol. Curr Opin Cardiol 2014. [7] McNamara DJ, Howell WH. Epidemiologic data linking diet to hyperlipidemia and arteriosclerosis. Semin Liver Dis 1992;12:347e55. [8] GBD Compare. IHME n.d. [9] Al-Nuaim AR, Al-Rubeaan K, Al-Mazrou Y, Al-Attas O, Al-Daghari N. Prevalence of hypercholesterolemia in Saudi Arabia, epidemiological study. Int J Cardiol 1996;54:41e9. [10] Al-Nuaim AR. Serum total and fractionated cholesterol distribution and prevalence of hypercholesterolemia in urban and rural communities in Saudi Arabia. Int J Cardiol 1997;58:141e9. [11] Abalkhail BA, Shawky S, Ghabrah TM, Milaat WA. Hypercholesterolemia and 5-year risk of development of coronary heart disease among university and school workers in Jeddah, Saudi Arabia. Prev Med 2000;31:390e5. [12] WHO j STEPwise approach to chronic disease risk factor surveillance. WHO n.d. [13] Downloadable questionnaires - International Physical Activity Questionnaire n.d. [14] Arafah MR, Mahmeed AAW. The undertreatment of hypercholesterolemia in Saudi Arabia and the Gulf states: results of the CEPHEUS study. J Saudi Heart Assoc 2012;24:276e7. [15] Mokdad AH, Jaber S, Aziz MIA, AlBuhairan F, AlGhaithi A, AlHamad NM, et al. The state of health in the Arab world, 1990-2010: an analysis of the burden of diseases, injuries, and risk factors. Lancet 2014;383:309e20. [16] Al-Nozha MM, Arafah MR, Al-Mazrou YY, Al-Maatouq MA, Khan NB, Khalil MZ, et al. Coronary artery disease in Saudi Arabia. Saudi Med J 2004;25:1165e71. [17] Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97:1837e47. [18] Estoppey D, Paccaud F, Vollenweider P, Marques-Vidal P. Trends in selfreported prevalence and management of hypertension, hypercholesterolemia and diabetes in Swiss adults, 1997-2007. BMC Public Health 2011;11:114. [19] Products - Data briefs - number 57-January 2011 n.d. [20] Ahmed F, Waslien C, Al-Sumaie M, Prakash P. Trends and risk factors of hypercholesterolemia among Kuwaiti adults: National Nutrition Surveillance Data from 1998 to 2009. Nutrition 2012;28:917e23. [21] Mandatory food labeling urged to warn consumers of risks to health. Yahoo Maktoob News n.d.
Contents lists available at ScienceDirect
Annals of Epidemiology journal homepage: www.annalsofepidemiology.org
Original article
Hypercholesterolemia and its associated risk factorsdKingdom of Saudi Arabia, 2013 Mohammed Basulaiman MD a, Charbel El Bcheraoui PhD, MSc b, Marwa Tuffaha MD b, Margaret Robinson BSc b, Farah Daoud BSc b, Sara Jaber MD b, Sarah Mikhitarian BSc b, Shelley Wilson BA b, Ziad A. Memish MD a, Mohammed Al Saeedi MD b, Mohammad A. AlMazroa MD a, Ali H. Mokdad PhD b, * a b
Ministry of Health of the Kingdom of Saudi Arabia, Riyadh, Saudi Arabia Institute for Health Metrics and Evaluation, University of Washington, Seattle
a r t i c l e i n f o
a b s t r a c t
Article history: Received 1 July 2014 Accepted 14 August 2014 Available online 20 August 2014
Purpose: To assess the prevalence of hypercholesterolemia and its associated factors in the Kingdom of Saudi Arabia. Methods: A national multistage representative sample of Saudis aged 15 years or older was surveyed through face-to-face interviews. Data on sociodemographics, risk factors, and health information were collected, and blood sample analysis was performed. Data were analyzed using SAS 9.3 to account for the sample weights and complex survey design. Results: Between April and June 2013, a total of 10,735 participants completed the survey. Overall, 8.5% of Saudis had hypercholesterolemia. Another 19.6% had borderline hypercholesterolemia. Among hypercholesterolemic Saudis, 65.1% were undiagnosed, 2.3% were treated uncontrolled, 28.3% were treated controlled, and 4.3% were untreated. The risk of being hypercholesterolemic increased with age and among individuals who reported consuming margarine, obese individuals, and those who have been previously diagnosed with hypertension or diabetes. Conclusions: More than a million Saudis have hypercholesterolemia, and 700,000 of them are unaware of their condition which can be controlled through early detection campaigns and lifestyle change and medication. An urgent awareness and screening campaign is needed in Kingdom of Saudi Arabia to prevent and avoid disease progression toward more serious stages. Ó 2014 Elsevier Inc. All rights reserved.
Keywords: Hypercholesterolemia Saudi Arabia Noncommunicable diseases
Introduction Cholesterol is essential for the body, specifically for hormone synthesis, fat digestion, and formation of cell membranes [1]. However, large amounts of cholesterol in the bloodstream lead to myocardial infarction, stroke, ischemic injury, and organ failure [2]. Hypercholesterolemia can be prevented, and measures such as dietary changes and clinical follow-up must be taken for high-risk individuals to reduce adverse events [3e7]. The Global Burden of Disease 2010 study estimated that high cholesterol is a leading risk factor for death from cardiovascular diseases in the Kingdom of Saudi Arabia (KSA). It accounted for Conflicts of interest: The salaries of the authors from the Institute for Health Metrics and Evaluation were fully or partially paid through a grant from the Ministry of Health (MOH) of the Kingdom of Saudi Arabia. The MOH managed the data collection for this study. * Corresponding author. Department of Global Health, Institute for Health Metrics and Evaluation, University of Washington, 2301 5th Avenue, Suite 600, Seattle, WA 98121. Tel.: þ1 206 897 2849; fax: þ1 206 897 2899. E-mail address: [email protected] (A.H. Mokdad). 1047-2797/$ e see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.annepidem.2014.08.001
about 5.14% of total deaths, 3.96% of years of life lost, and 1.99% of disability-adjusted life years. From 1990 to 2010, the burden of hypercholesterolemia remained very high in KSA as it ranked eighth among the risk factors for total deaths [8]. Previous studies reported on prevalence of hypercholesterolemia in KSA. In the early 80s, prevalence of hypercholesterolemia was 7% for males and 8% for females and increased with age and body mass index (BMI) [9,10]. Since the late 90s, data on hypercholesterolemia are limited [11]. However, prevalence of hypercholesterolemia was reported as 18.6% for males and 19.7% for females in 2005 [12]. To assess the current status of hypercholesterolemia in KSA, we analyzed the 2013 Saudi Health Interview Survey (SHIS) to describe the magnitude of hypercholesterolemia in KSA by different sociodemographic groups and risk factors. Materials and methods SHIS is a national multistage survey of individuals aged 15 years or older. Households were randomly selected from a national sampling
802
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
frame maintained and updated by the Census Bureau. KSA was divided into 13 regions. Each region was divided into subregions and blocks. All regions were included, and a probability proportional to size was used to randomly select subregions and blocks. Households were randomly selected from each block. A roster of household members was collected, and an adult aged 15 years or older was randomly selected to be surveyed. Weight, height, and blood pressure were measured at the household by a trained professional. An Omron HN286 scale (SN: 201207-03163F) and Omron M6 Comfort monitor (HEM-7223-E) were used to measure weight and blood pressure. The survey included questions on sociodemographic characteristics, including age, sex, education, and marital status, tobacco consumption, diet, physical activity, health care utilization, different health-related behaviors, and self-reported chronic conditions. We used measured weight and height to calculate BMI as weight (kilogram)/height2 (square meter). Participants were classified into four groups: (1) underweight, BMI less than18.5; (2) normal weight, BMI 18.5 to 24.9; (3) overweight, BMI 25.0 to 29.9; or (4) obese, BMI greater than or equal to 30.0. Respondents were considered to be current smokers if they reported ever smoking any tobacco products and still currently smoking tobacco, and were considered past smokers if they reported smoking in the past but not anymore. We assessed the type of fat most consumed by asking respondents: “What type of oil or fat is most often used for meal preparation in your household?” Respondents could choose between vegetable oils, olive oil, butter or ghee, margarine, or none in particular. Vegetable and olive oil were combined into one category of vegetable oil, and butter and ghee were considered animal fat. We computed the servings of fruits and vegetables and red meats and chicken consumed per day from the detailed dietary questionnaire as the sum of the average daily consumption of fruits, fruit juices, and vegetables, and red meats and chicken. We used the International Physical Activity questionnaire [13] to classify respondents into four groups of physical activity: (1) met vigorous physical activity level, (2) met moderate physical activity level, (3) insufficient physical activity to meet vigorous or moderate levels, and (4) no physical activity. To calculate the time spent on watching television, respondents were asked: “In a typical week, how much time do you usually spend in front of the television or on the computer?” Time was recorded in minutes or hours and then computed to create four categories in hours. To assess diagnosed hypertension, prediabetes, diabetes, and hypercholesterolemia status, respondents were asked four separate questions: “Have you ever been told by a doctor, nurse, or other
health professional that you had: (1) high blood pressure, otherwise known as hypertension; (2) prediabetes mellitus, otherwise known as prediabetes, borderline diabetes, impaired fasting glucose, impaired glucose tolerance, or impaired sugar tolerance; (3) diabetes mellitus, otherwise known as diabetes, sugar diabetes, high blood glucose, or high blood sugar; (4) hypercholesterolemia, otherwise known as high or abnormal blood cholesterol?” Women diagnosed with diabetes or hypertension only during pregnancy were counted as not having these conditions. Those who were diagnosed with either of these conditions were further asked if they are currently receiving any treatment for their condition. Similarly, the same types of questions were used to determine previous diagnosis of stroke, myocardial infarction, atrial fibrillation, cardiac arrest, congestive heart failure, chronic obstructive pulmonary disease, asthma, renal failure, and cancer. We considered a person to be diagnosed with a chronic condition if they reported being diagnosed with any of the conditions cited earlier. Respondents who reported being diagnosed with hypercholesterolemia were asked: “During the past 30 days, or since your diagnosis, have you ever taken medication for this condition?” Those who reported being on treatment were asked to provide the name of the drugs they are using. Respondents who completed the questionnaire were invited to local primary health care clinics to provide a blood sample for laboratory analysis. A Roche Hitachi COBAS 8000 system was used to measure cholesterol. Respondents were considered to be hypercholesterolemic if they met any of the following criteria: (1) measured cholesterol equal to or exceeding 6.2 mmol/L or (2) measured cholesterol not exceeding 6.2 mmol/L, but the respondent reported taking medications for hypercholesterolemia. Respondents were considered to have borderline hypercholesterolemia if (1) they did not report taking drugs for hypercholesterolemia and (2) their measured blood cholesterol level was between 5.18 and 6.2 mmol/L. Respondents under treatment for hypercholesterolemia were considered controlled if their measured cholesterol levels were below 6.2 mmol/L. Levels of high-density lipoproteins (HDL), low-density lipoproteins (LDL), and triglycerides were also measured. Respondents were considered to have low HDL if their measured HDL was below 1.0 mmol/L or 1.3 mol/L for males and females, respectively. Respondents were considered to have high LDL if their measured LDL was greater or equal to 4.13 mmol/L. They were considered to have hypertriglyceridemia if their measured triglycerides level was greater or equal to 2.3 mmol/L.
Table 1 Sociodemographic characteristics of Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y)
Marital status
Education
Categories
Males Females 15e24 25e34 35e44 45e54 55e64 65þ Currently married Never married Separated, divorced, or widowed Primary school* or less Elementaryy or high school completed College degree or higher education
SE ¼ Standard error. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
Complete sample
Clinical examination sample
N
Weighted %
SE
N
Weighted %
SE
5253 5482 2382 2757 2339 1520 862 875 3286 4872 2557 6976 2829 897
50.64 49.36 40.31 21.50 15.18 12.38 6.47 4.17 26.32 52.83 20.85 49.35 45.87 4.78
0.69 0.69 0.73 0.52 0.41 0.40 0.29 0.20 0.58 0.68 0.53 0.69 0.71 0.23
2576 3014 1163 1299 1241 860 485 542 2024 2325 1236 3657 1423 492
52.86 47.14 42.32 22.05 14.35 11.51 6.03 3.74 28.39 51.92 19.69 46.52 49.74 3.75
1.08 1.08 1.14 0.84 0.58 0.55 0.40 0.24 0.91 1.07 0.82 1.05 1.08 0.27
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
803
Table 2 Sociodemographic characteristics of diagnosed hypercholesterolemic, undiagnosed hypercholesterolemic, hypercholesterolemic, and borderline hypercholesterolemic, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Diagnosed hypercholesterolemic N
Sex Males Females Age (y) 15e24 25e34 35e44 45e54 55e64 65þ Marital status Currently married Never married Separated, divorced, or widowed Education Primary school* or less Elementaryy or high school completed College degree or higher education Smoking status Never smoked Ex-smoker Current smoker Type of fat consumed Vegetable oils Margarine Animal fat None in particular Daily servings of fruits and vegetables 0 0e3 3e5 5þ Daily servings of meat and chicken 0e1 1e2 2e3 3þ Daily hours spent on watching television 0e1 1e3 3e5 5þ Level of physical activity None Low Moderate High Obesity Not obese Obese History of diagnosis with hypertension No Yes History of diagnosis with prediabetes No Yes History of diagnosis with diabetes No Yes Diagnosis of chronic condition No Yes
Weighted %
Hypercholesterolemic undiagnosed SE
N
Weighted %
Hypercholesterolemic SE
N
Weighted %
Borderline hypercholesterolemic SE
N
Weighted %
SE
434 325
6.50 4.33
0.41 0.32
174 186
6.37 5.36
0.76 0.63
330 314
9.54 7.29
0.80 0.64
542 614
19.00 20.28
1.16 1.22
4 51 139 195 181 189
0.18 1.66 5.98 14.81 20.09 25.12
0.10 0.29 0.64 1.24 1.84 2.22
46 85 93 67 38 31
3.78 5.75 7.82 10.16 7.97 7.59
0.86 0.82 1.13 1.66 1.82 1.82
46 92 133 130 114 129
3.54 5.66 10.93 16.27 20.18 28.70
0.80 0.78 1.27 1.85 2.44 2.91
129 269 300 226 109 123
12.03 21.67 27.56 29.32 27.61 20.51
1.40 1.73 1.88 2.26 3.32 2.56
596 23 140
9.14 0.49 15.54
0.47 0.14 1.70
259 62 38
7.71 4.08 8.19
0.66 0.78 1.77
484 68 90
12.45 4.02 18.40
0.76 0.73 2.48
870 167 114
27.70 11.84 20.84
1.16 1.22 2.67
388 226 143
9.59 3.47 5.17
0.66 0.30 0.54
131 136 93
7.86 4.73 6.17
1.08 0.68 0.87
301 214 129
12.81 6.34 7.95
1.14 0.69 0.94
417 439 297
20.56 17.08 24.55
1.49 1.18 1.97
577 67 113
4.86 12.18 7.01
0.27 1.91 0.88
299 12 47
5.21 5.95 10.84
0.48 2.30 2.28
520 38 83
7.59 10.23 14.34
0.52 2.40 2.28
968 61 124
18.76 26.02 23.23
0.91 4.01 2.68
680 19 17 31
5.41 8.52 4.73 5.10
0.28 0.28 1.36 1.11
300 13 13 29
5.51 10.22 11.58 8.38
0.53 3.26 5.28 1.93
543 23 20 45
7.93 15.15 14.95 11.39
0.55 3.53 5.07 2.06
989 32 33 76
19.40 18.23 29.78 18.47
0.91 5.10 5.95 2.49
24 536 108 72
4.68 4.75 8.17 8.42
1.29 0.28 0.99 1.25
10 276 47 19
5.25 6.22 5.53 2.90
1.90 0.61 1.17 0.98
18 472 88 44
5.57 8.42 9.62 7.14
1.75 0.62 1.40 1.50
38 850 153 78
19.24 19.46 20.32 18.24
3.90 0.99 2.27 2.93
234 264 126 123
5.69 4.86 6.22 5.32
0.50 0.39 0.75 0.62
119 124 55 50
7.04 5.54 5.18 5.02
0.99 0.81 1.06 1.17
201 216 104 100
9.55 7.68 8.38 7.78
1.02 0.83 1.20 1.24
354 421 179 173
21.41 21.34 18.30 14.24
1.60 1.50 1.98 1.70
67 269 172 126
11.07 6.70 4.87 4.64
1.64 0.52 0.51 0.58
41 113 72 56
7.49 4.88 6.16 9.62
1.76 0.65 0.98 2.11
80 213 142 86
13.57 8.22 9.23 10.23
2.15 0.78 1.10 1.97
82 401 229 181
16.07 19.16 19.01 23.23
2.42 1.38 1.92 2.40
317 189 112 141
5.92 6.04 6.47 3.70
0.45 0.58 0.81 0.40
152 89 36 83
7.60 5.58 4.56 5.24
0.90 1.00 1.11 0.95
277 157 83 127
11.57 7.71 8.21 6.44
1.00 1.01 1.30 0.93
465 308 141 242
21.79 21.69 16.58 17.05
1.42 1.81 2.04 1.51
324 405
3.40 10.35
0.25 0.67
218 128
5.35 6.87
0.62 0.81
328 286
6.79 12.05
0.62 0.96
660 460
18.26 23.00
1.04 1.49
420 337
3.28 34.74
0.21 2.09
331 28
5.94 5.52
0.53 1.37
462 177
7.11 28.78
0.53 2.56
1006 135
19.21 23.79
0.88 2.69
651 64
4.79 32.70
0.25 4.41
349 6
6.00 2.28
0.51 1.09
578 36
8.13 24.17
0.53 4.88
1101 28
19.66 18.96
0.86 4.69
379 377
3.13 30.95
0.21 1.83
311 46
5.91 5.95
0.54 1.08
421 216
7.01 25.85
0.53 2.19
975 159
18.96 24.59
0.89 2.33
602 157
4.53 19.95
0.24 2.07
343 17
5.91 5.70
0.51 2.40
572 67
8.12 13.05
0.53 2.60
1084 68
19.17 24.81
0.85 4.06
SE ¼ Standard error. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
Weighting methodology Two sets of sample weights were generated and incorporated into the data set for analysis and extrapolation to the Saudi
population. First, we created an individual weight for all respondents to account for (1) the probability of selection of an eligible respondent within a household, (2) the probability of selection of the household within a stratum, and (3) the poststratification
804
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
differences in age and sex distribution between the sample and the Saudi population. For individuals who completed the laboratory-based blood analysis, we computed an additional weight used in analyzing data from clinic visits to account for (1) the individual weight described previously, (2) the probability of visiting a clinic, (3) sociodemographic, behavioral, and health differences between respondents who visited the clinic and those who did not, and (4) the poststratification differences in age and sex distribution between the respondents who visited the clinic and the general Saudi population. Statistical analysis We used a multivariate logistic regression model to measure association between outcome variables and sociodemographic factors first. Then, we used a backward elimination multivariate logistic regression model to measure the association between outcome variables and all associated factors. All factors were first included in the models. Then variables were eliminated based on a Wald c2 test for the analysis of effect. Variables were removed one by one based on the significance level of their effect on the model, starting with the variable with the highest P > .5, till all variables kept had a P .5 in the analysis of effect. We used SAS 9.2 (SAS Institute Inc., Cary, NC) for the analyses and to account for weights and the complex sampling design. Results Between April and June 2013, a total of 10,735 participants completed the SHISdresponse rate of 89.4%dand were invited to the local health clinics. The remaining 1265 participants completed part of the household enumeration, or all of it, but the selected adult did not complete the survey. A total of 5590 individuals went to the local clinics and provided blood samples for analysesda response rate of 52.1%. The characteristics of respondents who completed the questionnaire and the laboratory examination are presented in Table 1. Overall, an estimated 693,106 Saudis (5.4%) reported being diagnosed with hypercholesterolemia. However, an estimated 1,107,296 Saudis (8.5%) aged 15 years or older had hypercholesterolemia as measured by our laboratory tests with blood cholesterol level greater or equal to 6.2 mmol/L or reported currently taking cholesterol medication. Of these, 65.1% were undiagnosed or unaware of their condition. Moreover, 19.6% of Saudis, an estimated 2,397,211 participants, had borderline hypercholesterolemia with measured blood cholesterol levels between 5.18 and 6.2 mmol/L and not currently taking blood cholesterol lowering medication. Characteristics of respondents with undiagnosed hypercholesterolemia, hypercholesterolemia, and borderline hypercholesterolemia are presented in Table 2. Among participants diagnosed with hypercholesterolemia, 92.9% reported taking medication for their condition. More than 95% of those being treated were taking statins for their condition. The vast majority of hypercholesterolemic patients (92.6%) who were on medication had their blood cholesterol level controlled. Among those who were hypercholesterolemic, 65.1% were undiagnosed, 2.3% were treated uncontrolled, 28.3% were treated controlled, and 4.3% were untreated (Fig. 1). A high percent of Saudis (48.7%) have low HDL, 7.4% have high LDL, and 8.5% have hypertriglyceridemia. Distributions of these three conditions by sociodemographic and risk factors are presented in Table 3. Age, type of fat mostly consumed, obesity, and diagnosis history of hypertension and diabetes were associated with hypercholesterolemia (Table 4). The risk of being hypercholesterolemic
increased with age (adjusted odds ratios [AOR] ¼ 1.03; 95% confidence intervals [CI], 1.01e1.04), among individuals who reported consuming margarine (AOR ¼ 2.17; 95% CI, 1.17e4.00), obese individuals (AOR ¼ 1.40; 95% CI, 1.03e1.89), and those who have been previously diagnosed with hypertension (AOR ¼ 1.82; 95% CI, 1.31e2.53) and diabetes (AOR ¼ 1.85; 95% CI, 1.28e2.69). On the other hand, sex, marital status, education, smoking status, time spent on watching television, and level of physical activity were not associated with the risk of hypercholesterolemia (Table 4). The risk of borderline hypercholesterolemia increased with age and time spent on watching television (Table 5). The risk of being undiagnosed when one is hypercholesterolemic was associated only with age (Table 6). Older individuals were more likely to be undiagnosed while hypercholesterolemic (AOR ¼ 1.02; 95% CI, 1.01e1.03).
Discussion To our knowledge, this is the first national study to report on cholesterol levels in KSA. We found high rates of undiagnosed cholesterol and borderline levels in a country where chronic diseases are the major health burden. Our findings call for awareness campaigns to educate the public about the danger of high cholesterol levels and the importance of early diagnosis. Indeed, because Saudis have free medical care, receive free medication when needed, and our data show high control levels of cholesterol once on medication, our recommendation will reduce the burden of cholesterol-related outcomes in KSA. Moreover, if these campaigns include messages on changing health behaviors such as diet and physical activity, they will have a larger impact on health. Our findings about lack of diagnosis and treatment of hypercholesterolemia in KSA have been previously reported; the majority of hypercholesterolemic Saudi individuals are unaware of their condition, and hypercholesterolemia remains undertreated [14]. In KSA, ischemic heart disease is the leading cause of death [8] as is the case for the whole Arab world, geographically comprised 22 countries known as the Arab league [15], and it has been reported that hypercholesterolemia is a major risk factor for this disease, both in KSA and worldwide [16,17]. Hypercholesterolemia has been reported as the eighth risk factor for death in the Arab World in Global Burden of Disease 2010 study. In our study, type of fat consumed, obesity, and sedentary activities were associated with hypercholesterolemia and borderline hypercholesterolemia. Hence, health authorities in KSA should develop public health campaigns that promote healthier lifestyles, encourage people to exercise, and make informed decisions around their eating habits.
Fig. 1. Percent distribution of diagnosis and treatment status among hypercholesterolemic, Saudis aged 15 years older, 2013.
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
805
Table 3 Distribution of HDL, LDL, and triglycerides by sociodemographic and risk factors Sociodemographic and risk factors
Sex Males Females Age (y) 15e24 25e34 35e44 45e54 55e64 65þ Marital status Currently married Never married Separated, divorced, or widowed Education Primary school* or less Elementaryy or high school completed College degree or higher education Smoking status Never smoked Ex-smoker Current smoker Type of fat consumed Vegetable oils Margarine Animal fat None in particular Daily servings of fruits and vegetables 0 0e3 3e5 5þ Daily servings of meat and chicken 0e1 1e2 2e3 3þ Daily hours spent on watching television 0e1 1e3 3e5 5þ Level of physical activity None Low Moderate High Obesity Not obese obese History of diagnosis with hypertension No Yes History of diagnosis with prediabetes No Yes History of diagnosis with diabetes No Yes
HDL
LDL
Hypertriglyceridemia
Not diagnosed or treated
Diagnosed and treated Not diagnosed or treated
Diagnosed and treated
Not diagnosed or treated
Diagnosed and treated
N
N
N Weighted % SE
N
N Weighted % SE
Weighted % SE
947 38.89 1717 59.77
Weighted % SE
1.57 1.56
59 39.84 72 63.77
45.33 47.91 53.42 54.74 56.31 50.04
2.04 2.11 2.09 2.57 3.79 3.66
0 2 11 31 38 49
1741 53.16 665 44.39 250 58.43
1.26 1.86 4.15
997 55.29 1068 46.10
597 47.34
558 617 641 428 214 206
N
Weighted % SE
8.29 6.67
0.87 0.84
6 5
4.74 5.44
2.42 249 3.20 185
3.69 7.52 11.62 13.85 13.32 13.53
0.88 1.17 1.62 2.10 2.73 3.07
0 0 4 3 0 4
0 0 9.06 6.84 0 5.07
0 43 4.14 0 81 7.13 5.99 121 13.48 4.58 93 15.07 0 53 18.78 3.35 43 12.54
0.86 0 1.08 1 1.56 3 2.01 7 3.21 8 2.49 18
0 12.84 8.40 11.44 14.81 15.05
0 13.78 5.13 5.98 7.14 4.06
95 43.54 1 44.10 35 75.36
5.22 263 11.01 31.99 60 4.31 7.93 41 11.89
0.93 0.82 2.54
7 0 4
4.84 0 6.84
2.25 348 12.99 0 57 4.27 4.31 28 9.34
0.92 30 12.40 0.78 0 0 2.24 7 19.66
3.36 0 7.48
1.92 1.66
85 44.51 31 58.42
5.89 136 10.37 8.90 144 5.88
1.36 0.78
5 5
3.26 7.46
1.97 161 4.39 168
1.11 24 12.58 0.79 9 16.65
2.90 7.88
2.37
15 39.78
8.15
1.23
1
6.02
5.87 104 10.65
1.44
7.35
4.47
0 63.98 45.03 48.37 53.99 43.94
5.94 194 6.31 171
Weighted % SE
0 29.86 14.22 10.30 8.37 7.05
10.18
41 79 95 68 47 35
85
9.93 6.52
9.87 6.61
0.91 23 12.85 0.72 14 13.60
4
4.08 3.87
2359 49.47 82 39.92 221 47.29
1.22 102 46.52 5.29 12 40.31 3.48 17 60.60
5.29 300 6.75 14.77 16 8.94 12.61 47 12.37
0.61 2.87 2.50
8 0 3
4.96 0 7.84
2.24 341 6.94 0 26 13.73 6.46 65 16.91
0.56 26 11.72 3.97 6 12.22 2.69 5 21.32
2.80 6.53 12.77
2297 81 75 172
48.83 53.28 54.50 46.79
1.23 117 49.06 5.95 4 35.31 6.52 4 73.39 3.80 3 31.54
5.17 303 7.11 18.16 16 11.88 18.94 10 12.67 19.29 29 10.78
0.64 11 3.64 0 6.03 0 2.51 0
6.21 0 0 0
2.47 373 0 14 0 6 0 32
8.39 6.24 5.58 9.49
0.65 35 15.07 2.28 0 0 3.00 0 0 1.97 1 7.01
3.51 0 0 7.05
71 2057 322 178
49.90 49.05 46.31 48.56
5.96 1.30 3.16 4.29
4 98 17 7
35.54 49.72 39.73 47.67
26.00 8 5.67 283 10.72 49 15.72 21
4.78 7.74 8.07 5.77
1.92 0 0.72 10 1.74 0 1.59 1
0 6.95 0 1.52
0 10 2.83 332 0 54 1.57 30
6.71 8.49 8.01 7.51
2.42 0 0 0.71 28 12.41 1.42 5 8.68 1.82 4 30.16
0 3.01 4.25 16.53
789 974 415 458
49.40 49.48 46.92 47.83
2.05 1.88 2.88 2.68
40 41 29 19
42.97 49.47 56.79 46.86
8.63 119 8.00 134 11.18 65 10.90 43
8.61 7.79 7.18 5.48
1.03 1.13 1.34 1.34
2 4 2 3
6.21 3.74 5.48 4.24
4.39 123 2.94 159 4.32 76 2.83 70
7.57 8.95 9.51 7.34
0.96 8 12.16 1.05 15 13.38 1.61 5 13.43 1.36 9 14.92
6.07 4.08 8.92 5.59
186 915 543 399
53.74 48.64 45.28 48.43
4.77 1.93 2.46 2.88
15 48 40 8
35.00 58.57 40.63 24.60
10.12 36 8.72 7.67 122 7.07 8.08 82 8.62 10.04 59 10.77
2.09 0.99 1.34 2.20
3 12.27 1 0.45 2 4.42 2 17.04
7.89 29 0.46 150 3.69 96 13.52 57
6.48 8.41 7.99 8.43
1.63 4 12.89 0.99 16 16.65 1.23 9 12.52 1.70 4 19.89
6.48 6.09 5.56 10.00
1091 674 316 583
58.45 50.70 43.46 40.98
1.78 2.20 3.05 2.12
53 37 20 21
48.67 51.93 32.64 59.90
7.71 141 8.21 102 9.08 37 11.77 85
9.58 7.81 5.79 6.30
1.26 1.23 1.30 0.99
5 3 1 2
3.98 8.66 3.78 2.65
2.60 158 5.39 114 3.77 54 1.93 108
8.47 8.48 6.87 8.67
0.91 12 12.76 1.25 9 10.49 1.31 10 16.54 1.16 6 14.27
5.04 3.69 7.96 6.65
1575 45.81 1005 56.58
1.41 1.87
48 41.04 79 55.01
7.03 207 6.18 147
6.41 9.81
0.70 1.14
7 4
7.05 3.64
3.55 244 6.73 2.32 183 12.46
0.66 11 6.20 1.27 24 19.18
2.27 5.09
2432 48.54 220 56.74
1.17 3.72
59 43.08 72 51.96
7.50 328 7.42 5.80 35 10.12
0.63 2.30
6 5
5.75 4.44
3.10 389 8.09 2.53 45 15.53
0.61 10 5.37 2.49 27 18.71
2.02 4.64
2563 48.63 54 59.44
1.15 111 47.97 7.27 16 61.82
0.62 10 1.55 0
5.01 0
2.13 410 8.12 0 13 18.82
0.60 30 13.96 5.98 5 7.10
3.48 3.87
2356 48.62 289 55.54
1.19 2.98
0.64 1.90
0.89 8.34
0.66 341 7.51 3.50 92 21.38
0.61 10 10.13 2.48 27 15.34
4.10 4.18
47 51.98 84 45.39
5.26 351 11.03 8
7.64 3.68
7.51 317 7.35 5.82 43 10.24
2 9
(continued on next page)
806
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
Table 3 (continued ) Sociodemographic and risk factors
HDL
LDL
Hypertriglyceridemia
Not diagnosed or treated
Diagnosed and treated Not diagnosed or treated
Diagnosed and treated
Not diagnosed or treated
Diagnosed and treated
N
N
N Weighted % SE
N
N
Weighted % SE
Diagnosis of chronic condition No 2526 48.93 Yes 138 47.08
Weighted % SE
1.16 102 46.79 4.65 29 53.62
N
5.19 347 11.01 18
Weighted % SE 7.53 7.56
0.62 10 2.90 1
5.60 2.48
Weighted % SE
2.38 406 8.18 2.48 28 10.42
Weighted % SE
0.60 24 10.03 2.85 13 25.87
2.64 10.08
SE ¼ Standard error. * Six years of schooling following kindergarten. y Three years of schooling following primary school.
Our findings showed that older individuals should be targeted for early detection and encouraged to change their lifestyle or seek treatment, especially because health care and medications in KSA is free and readily available to citizens. As seen in our data, most Saudis treated for hypercholesterolemia are prescribed statins, and the fact that 92% of treated subjects are able to control their cholesterol blood level is an encouraging finding in our study. Moreover, the risk of hypercholesterolemia increased among diabetic and hypertensive patients. Screening for hypercholesterolemia should be prioritized among these patients as comorbid individuals who are at higher risk of serious health complications. The last reported study of hypercholesterolemia rates in KSA was in 2005 and revealed a 19.1% prevalence of hypercholesterolemiadcutoff point of 5.2 mmol/Ldamong Saudis aged 15 to 64 years [12]. When restricted to the same age groups and using the same cutoff point, our data revealed a prevalence of 25.3%. However, in our survey, we applied poststratification weights to
reflect the Saudi population and adjusted for the increased probability of sick individuals to agree to visit the health facility for physical and blood samples. This was not done for the previous survey. Indeed, when we did not apply adjustments based on poststratification and the predicted probability of completing the clinic visit; our hypercholesterolemia estimates were 30.1% (95% CI, 29.0e32.4). Clearly, this indicates a rise in the prevalence of cholesterol in KSA and calls for programs to prevent and control cholesterol levels given its role as a major risk factor for ischemic heart disease, the leading cause of death in KSA [8]. Our study has some limitations. First, our data are from a crosssectional study; hence, we cannot assess causality. Second, many of our behavioral data, such as diet and physical activity, are selfreported and subject to recall and social desirability biases. On the other hand, our study is based on a large sample size and used a standardized methodology for all its measures. Third, only 52% of respondents completed the visit to the health clinics and had their blood drawn for analysis. However, our weighting methodology
Table 4 Multivariate logistic regression for sociodemographic characteristics, risk factors, and hypercholesterolemia, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y) Marital status
Education
Smoking status
Type of fat consumed
Daily hours spent on watching television
Level of physical activity
Obesity History of diagnosis with hypertension History of diagnosis with diabetes AOR ¼ Adjusted odds ratio; CI ¼ Confidence interval. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
Categories
Males Females Currently married Never married Separated, divorced, or widowed Primary school* or less Elementaryy or high school completed College degree or higher education Never smoked Ex-smoker Current smoker Vegetable oils Margarine Animal fat None in particular 0e1 1e3 3e5 5þ None Low Moderate High Not obese Obese No Yes No Yes
Sociodemographic model
Full model
AOR
95% CI
AOR
95% CI
d 0.73 1.04 d 0.71 1.23 d 0.97 1.00 d d d d d d d d d d d d d d d d d d d d d
d 0.56e0.95 1.03e1.05 d 0.47e1.06 0.83e1.82 d 0.65e1.44 0.66e1.50 d d d d d d d d d d d d d d d d d d d d d
d 0.72 1.03 d 0.79 1.49 d d d d 0.62 1.70 d 2.17 1.00 1.25 d 0.82 1.14 1.19 d 0.73 0.83 0.68 d 1.40 d 1.85 d 1.88
d 0.50e1.03 1.01e1.04 d 0.49e1.27 0.96e2.33 d d d d 0.28e1.34 0.98e2.94 d 1.17e4.04 0.34e2.95 0.72e2.19 d 0.52e1.27 0.70e1.87 0.68e2.06 d 0.48e1.12 0.52e1.33 0.43e1.09 d 1.03e1.89 d 1.28e2.69 d 1.31e2.70
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
807
Table 5 Multivariate logistic regression for sociodemographic characteristics, risk factors, and borderline hypercholesterolemia, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y) Marital status
Education
Type of fat consumed
Daily servings of meat and chicken
Daily hours spent on watching television
Level of physical activity
History of diagnosis with prediabetes History of diagnosis with diabetes Diagnosis of chronic condition
Categories
Males Females Currently married Never married Separated, divorced, or widowed Primary* school or less Elementaryy or high school completed College degree or higher education Vegetable oils Margarine Animal fat None in particular 0e1 1e2 2e3 3þ 0e1 1e3 3e5 5þ None Low Moderate High No Yes No Yes No Yes
Sociodemographic model
Full model
AOR
95% CI
AOR
95% CI
d 1.09 1.02 d 0.43 0.67 d 1.32 1.63 d d d d d d d d d d d d d d d d d d d d d d
d 0.86e1.37 1.01e1.03 d 0.31e0.60 0.46e0.98 d 0.98e1.78 1.17e2.26 d d d d d d d d d d d d d d d d d d d d d d
d 1.02 1.02 d 0.43 0.69 d 1.19 1.29 d 1.25 1.96 1.09 d 1.00 0.76 0.74 d 1.44 1.66 2.00 d 1.00 0.73 0.81 d 0.54 d 1.19 d 1.31
d 0.76e1.35 1.00e1.03 d 0.29e0.64 0.44e1.10 d 0.84e1.69 0.88e1.91 d 0.54e2.87 0.99e3.87 0.69e1.73 d 0.73e1.38 0.51e1.12 0.49e1.11 d 0.96e2.17 1.05e2.63 1.25e3.20 d 0.72e1.39 0.47e1.12 0.57e1.14 d 0.27e1.08 d 0.80e1.75 d 0.65e2.62
AOR ¼ Adjusted odds ratio; CI ¼ Confidence interval. * Six years of schooling after kindergarten. y Three years of schooling after primary school.
accounted for this bias by applying a poststratification adjustment using sociodemographic characteristics, health behavior, previously diagnosed noncommunicable diseases, and anthropometric measurements of respondents from the household survey. The prevalence of hypercholesterolemia in KSA is not as high as in many developing countries but is a shared risk factor for death with the remaining Arab World [15]. For instance, in Switzerland, 17.4% of adults aged 15 years or older were selfreportedly hypercholesterolemic in 2007 [18], double the prevalence of KSA. In the United States in 2005 to 2008, 23% of adults aged 20 to 64 years had hypercholesterolemia, defined similarly
to our study [19]. In KSA, only 6.7% of adults of the same age group were affected in 2013. On the other hand, the prevalence in KSA is similar to its neighboring countries. For example, in Kuwait, about 7% of individuals aged 20 to 69 years had blood cholesterol levels at or above 6.2 mmol/L in 2009 [20]. We found a large prevalence of undiagnosed hypercholesterolemia, a major risk factor for ischemic heart disease, the leading cause of death in KSA [8]. Despite its relatively low prevalence, the issue of hypercholesterolemia should be addressed to reduce the risk of ischemic heart disease and death, especially because Saudi patients have been successful in lowering their blood cholesterol levels.
Table 6 Multivariate logistic regression for sociodemographic characteristics, risk factors, and undiagnosis of hypercholesterolemia, Saudis aged 15 years or older, 2013 Sociodemographic and risk factors
Sex Age (y) Marital status
Education
History of diagnosis with prediabetes History of diagnosis with diabetes History of diagnosis with hypertension AOR ¼ Adjusted odds ratio; CI ¼ Confidence interval.
Categories
Males Females Currently married Never married Separated, divorced, or widowed Primary school or less Elementary or high school completed College degree or higher education No Yes No Yes No Yes
Sociodemographic model
Full model
AOR
95% CI
AOR
95% CI
d 0.75 1.01 d 0.64 1.04 d 0.74 0.85 d d d d d d
d 0.54e1.04 1.00e1.03 d 0.41e1.01 0.61e1.78 d 0.45e1.21 0.52e1.38 d d d d d d
d 0.78 1.02 d 0.68 1.07 d d d d 0.27 d 0.80 d 0.68
d 0.54e1.13 1.01e1.03 d 0.43e1.07 0.62e1.84 d d d d 0.09e0.79 d 0.49e1.31 d 0.37e1.27
808
M. Basulaiman et al. / Annals of Epidemiology 24 (2014) 801e808
Indeed, concerned Saudi health professionals are already calling for steps, such as food labeling, to prevent cardiovascular diseases [21]. Conclusions Our findings call for urgent measures by the Saudi Ministry of Health to reduce the burden of hypercholesterolemia in KSA. Targeted blood cholesterol screening campaigns are required to detect individuals with hypercholesterolemia and refer them for medical follow-up. Moreover, the early detection campaigns should offer those with borderline hypercholesterolemia means to control their cholesterol levels through behavioral changes or medication if they have other risk factors. In addition, these early detection campaigns should include screening for blood pressure or diabetes to maximize their benefits. Finally, educational campaigns are needed to promote healthier lifestyles through physical activity and healthy eating habits. Indeed, changing health behaviors among Saudis will have a tremendous impact on the burden of disease in KSA. Acknowledgment This study was financially supported by a grant from the Ministry of Health (MOH) of the Kingdom of Saudi Arabia. The authors would like to acknowledge Adrienne Chew at the Institute for Health Metrics and Evaluation, Seattle, WA for editing this paper. References [1] Widmaier E, Raff H, Strang K. Vander’s Human physiology: the mechanisms of body function. 13th ed. New York: McGraw-Hill Science/Engineering/Math; 2013. [2] Kashyap ML. Cholesterol and atherosclerosis: a contemporary perspective. Ann Acad Med Singapore 1997;26:517e23.
[3] Ziegler O, Got I, Jan P, Drouin P. [Diet therapy of hypercholesterolemia. From theory to practice]. Ann Cardiol Angéiol 1989;38:249e53. [4] Darnton-Hill I, Nishida C, James WPT. A life course approach to diet, nutrition and the prevention of chronic diseases. Public Health Nutr 2004;7:101e21. [5] Diet, nutrition and the prevention of chronic diseases. World Health Organ Tech Rep Ser 2003;916:ieviii. 1e149, backcover. [6] Griffin BA. Nonpharmacological approaches for reducing serum low-density lipoprotein cholesterol. Curr Opin Cardiol 2014. [7] McNamara DJ, Howell WH. Epidemiologic data linking diet to hyperlipidemia and arteriosclerosis. Semin Liver Dis 1992;12:347e55. [8] GBD Compare. IHME n.d. [9] Al-Nuaim AR, Al-Rubeaan K, Al-Mazrou Y, Al-Attas O, Al-Daghari N. Prevalence of hypercholesterolemia in Saudi Arabia, epidemiological study. Int J Cardiol 1996;54:41e9. [10] Al-Nuaim AR. Serum total and fractionated cholesterol distribution and prevalence of hypercholesterolemia in urban and rural communities in Saudi Arabia. Int J Cardiol 1997;58:141e9. [11] Abalkhail BA, Shawky S, Ghabrah TM, Milaat WA. Hypercholesterolemia and 5-year risk of development of coronary heart disease among university and school workers in Jeddah, Saudi Arabia. Prev Med 2000;31:390e5. [12] WHO j STEPwise approach to chronic disease risk factor surveillance. WHO n.d. [13] Downloadable questionnaires - International Physical Activity Questionnaire n.d. [14] Arafah MR, Mahmeed AAW. The undertreatment of hypercholesterolemia in Saudi Arabia and the Gulf states: results of the CEPHEUS study. J Saudi Heart Assoc 2012;24:276e7. [15] Mokdad AH, Jaber S, Aziz MIA, AlBuhairan F, AlGhaithi A, AlHamad NM, et al. The state of health in the Arab world, 1990-2010: an analysis of the burden of diseases, injuries, and risk factors. Lancet 2014;383:309e20. [16] Al-Nozha MM, Arafah MR, Al-Mazrou YY, Al-Maatouq MA, Khan NB, Khalil MZ, et al. Coronary artery disease in Saudi Arabia. Saudi Med J 2004;25:1165e71. [17] Wilson PWF, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97:1837e47. [18] Estoppey D, Paccaud F, Vollenweider P, Marques-Vidal P. Trends in selfreported prevalence and management of hypertension, hypercholesterolemia and diabetes in Swiss adults, 1997-2007. BMC Public Health 2011;11:114. [19] Products - Data briefs - number 57-January 2011 n.d. [20] Ahmed F, Waslien C, Al-Sumaie M, Prakash P. Trends and risk factors of hypercholesterolemia among Kuwaiti adults: National Nutrition Surveillance Data from 1998 to 2009. Nutrition 2012;28:917e23. [21] Mandatory food labeling urged to warn consumers of risks to health. Yahoo Maktoob News n.d.