Aging in Asia Findings from New and Emerging Data Initiatives (2012) / Chapter Skim
Currently Skimming:
13 Socioeconomic Success and Health in Later Life: Evidence from the Indonesia Family Life Survey--Firman Witoelar, John Strauss, and Bondan Sikoki
13 Socioeconomic Success and Health in Later Life: Evidence from the Indonesia Family Life Survey--Firman Witoelar, John Strauss, and Bondan Sikoki
Pages 309-341
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 309... ...
On the other hand, hemoglobin levels have also been rising, though from low levels, leading to improved health. Yet, other health measures have been fairly steady as shown in the Indonesia Family Life Survey (IFLS)
|
From page 310... ...
correlates of elderly health outcomes will help to improve knowledge that could be useful in designing health as well as social programs to improve the well-being of the elderly in Indonesia. In this chapter, we document the health and nutrition transition that the elderly population in Indonesia has undergone in the 15 years between 1993 and 2008, using the four full waves of the Indonesia Family Life Survey (IFLS)
|
From page 311... ...
, blood hemoglobin, total and HDL cholesterol levels, hypertension, cognition measured by word recall, and an index of depression (the short CES-D) .7 This is a much broader set of health indicators than is usually analyzed, in large part because such a rich set of health data is not usually available in broad-purposed socioeconomic surveys.
|
From page 312... ...
One potential worry in a study like this over a 15-year period is sample 8 Public-use files from IFLS1 are documented in six volumes under the series title The 1993 Indonesian Family Life Survey, DRU-1195/1–6-NICHD/AID, The RAND Corporation, December 1995. IFLS2 public-use files are documented in seven volumes under the series The Indonesia Family Life Survey, DRU-2238/1-7-NIA/NICHD, RAND, 2000.
|
From page 313... ...
Table 13-1 shows means and standard deviations from the IFLS4 data for our covariates. We create dummy variables for age indicating whether an individual is aged 55 and older, 65 and older, and 75 and older.
|
From page 314... ...
Rural 0.503 0.500 0.492 0.500 Province North Sumatra 0.055 0.229 0.061 0.239 West Sumatra 0.049 0.216 0.055 0.229 South Sumatra 0.047 0.211 0.045 0.208 Lampung 0.044 0.205 0.036 0.187 Jakarta 0.064 0.245 0.060 0.238 West Java 0.166 0.372 0.157 0.364 Central Java 0.138 0.345 0.145 0.352 Yogyakarta 0.072 0.259 0.074 0.261 East Java 0.155 0.362 0.162 0.369 Bali 0.054 0.226 0.054 0.226 West Nusa Tenggara 0.059 0.236 0.058 0.234 South Kalimantan 0.047 0.211 0.040 0.195 South Sulawesi 0.049 0.215 0.052 0.223 SOURCE: Data from IFLS4. way, the coefficients on the dummy variables indicate the marginal change from the next lowest age group (not from the omitted group)
|
From page 315... ...
If it does, then this is more consistent with a causal interpreta 11 The birth-year cohort dummy variables included are as follows: –1928, 1929–1933, 1934–1938, 1939–1943, 1944–1948, 1949–1953, 1954–1958, with 1959–1963 omitted as the base. 12 For health measures that we only have data from 2007, of course, we do not use either year or birth cohort dummies, but we still use the age dummies.
|
From page 316... ...
. RESULTS Physical Measurement: Anthropometry, Hemoglobin Level, and Hypertension BMI We first look at a number of biomarkers: BMI, waist circumference, blood hemoglobin levels, and hypertension.16 BMI, which is weight (in kg)
|
From page 317... ...
Holding BMI constant, greater waist circumference increases the risks of various cardiovascular diseases. For people who are overweight or obese, the risk of future mortality is higher if their waist circumference is greater than 120 cm for men or 88 cm for women.
|
From page 318... ...
Females, aged 45+ Waist (cm) 50 60 70 80 90 100 110 120 50 60 70 80 90 100 110 120 12 14 16 18 20 22 24 26 28 30 32 34 36 12 14 16 18 20 22 24 26 28 30 32 34 36 BMI BMI 2000 2007 2000 2007 FIGURE 13-1 CDF of body mass index and waist circumference by body mass index, adults aged 45 and older.
|
From page 319... ...
% # obs. 45−54 years % undernourished 9.46 1,870 9.39 2,106 % overweight 22.65 1,870 40.18 2,106 % low blood hemoglobin 17.72 1,869 28.08 2,091 % high total cholesterol 12.49 1,832 20.02 2,054 % low HDL 70.59 1,832 40.81 2,054 55−64 years % undernourished 18.22 1,096 16.64 1,211 % overweight 17.26 1,096 30.57 1,211 % low blood hemoglobin 26.01 1,093 32.82 1,212 % high total cholesterol 12.66 1,071 26.00 1,204 % low HDL 64.79 1,071 36.25 1,204 65−74 years % undernourished 27.95 713 29.57 878 % overweight 8.59 713 18.82 878 % low blood hemoglobin 40.86 728 40.25 886 % high total cholesterol 9.48 716 23.07 884 % low HDL 59.49 715 41.31 883 75+ years % undernourished 38.05 338 33.60 438 % overweight 6.31 338 13.96 438 % low blood hemoglobin 52.24 350 50.06 461 % high total cholesterol 8.60 339 21.63 448 % low HDL 65.22 339 34.47 448 All adults 45+ % undernourished 17.54 4,017 17.40 4,633 % overweight 17.31 4,017 31.14 4,633 % low blood hemoglobin 27.12 4,040 33.81 4,650 % high total cholesterol 11.66 3,958 22.33 4,590 % low HDL 66.55 3,957 39.09 4,589 Mean BMI 21.75 4,017 22.90 4,633 Mean blood hemoglobin 13.99 4,040 12.42 4,650 Mean total cholesterol 178.16 3,958 198.46 4,590 Mean HDL 34.94 3,957 44.97 4,589 SOURCE: Data from IFLS4.
|
From page 320... ...
. Education variables are jointly significant for both men and women.
|
From page 321... ...
[18.213] Observations 12,836 14,735 10,305 11,853 R-squared 0.226 0.222 0.123 0.057 Cohort Dummy Variables Yes Yes Yes Yes Province × Rural Dummy Yes Yes Yes Yes Variables + Province × Rural × Year Interactions 321 continued
|
From page 322... ...
age interactions 41.392 0.000 21.542 0.000 8.220 0.000 6.635 0.000 Per capita expenditures 82.506 0.000 78.482 0.000 30.996 0.000 6.474 0.002 Cohort dummy variables 5.143 0.000 11.298 0.000 2.067 0.046 1.649 0.120 Year dummy variables 1.963 0.119 1.292 0.277 5.154 0.006 0.639 0.528 Province × rural dummy variables 4.626 0.000 6.787 0.000 4.182 0.000 4.208 0.000 Year × prov × rural variables 2.233 0.000 2.815 0.000 4.687 0.000 2.728 0.000 interactions NOTES: The dependent variable for BMI regressions is the BMI; for hemoglobin, the hemo globin level (g/dL)
|
From page 323... ...
Given what we know about what blood hemoglobin levels can tell us, this change shows an improvement in one dimension of health in Indonesia over the years. Even so, the 2007 levels are still high compared to what is found in industrial countries, consistent with much evidence that low hemoglobin levels exist in lowincome countries (Tolentino and Friedman, 2007)
|
From page 324... ...
Females, aged 45+ 1 .8 Pr < = Hb .6 .4 .2 0 8 9 10 11 12 13 14 15 16 17 Hb Level 1997 2000 2007 FIGURE 13-2 CDF of hemoglobin levels, adult aged 45 and older, 1997, 2000, and 2007. SOURCE: Data from IFLS, Waves 2-4.
|
From page 325... ...
Table 13-4 reports the regressions. There is very little SES correlation with either the probability of having high total cholesterol or low HDL.
|
From page 326... ...
Low HDL .8 .6 Proportion Male Female .4 .2 40 50 60 70 80 Age FIGURE 13-3 Proportion with high total cholesterol and low HDL, adults aged 45 and older, 2007. SOURCE: Data from IFLS4.
|
From page 327... ...
[5.018] Observations 3,960 4,591 3,958 4,589 R-squared 0.069 0.074 0.047 0.068 Cohort Dummy Variables Yes Yes Yes Yes Province × Rural Yes Yes Yes Yes 327 continued
|
From page 328... ...
age interactions 1.572 0.180 1.868 0.115 1.096 0.358 0.571 0.684 Per capita expenditures 6.692 0.001 1.736 0.177 0.610 0.544 6.908 0.001 Cohort dummy variables 1.649 0.120 1.649 0.120 1.649 0.120 1.649 0.120 Year dummy variables 0.639 0.528 0.639 0.528 0.639 0.528 0.639 0.528 Province × rural dummy variables 6.756 0.000 8.508 0.000 4.386 0.000 8.971 0.000 NOTES: The dependent variable for high total cholesterol is 1 if individual has total cholesterol level ≥ 240 mg/dL, 0 otherwise. The dependent variable for low HDL level is 1 if individual has HDL level < 40 mg/dL, 0 otherwise.
|
From page 329... ...
against age. For both men and women, there is a strong positive relationship between age and being hypertensive.
|
From page 330... ...
330 AGING IN ASIA A Males, aged 45+ .7 Proportion with Hypertension .6 1997 .5 2000 2007 .4 .3 40 50 60 70 80 Age B
|
From page 331... ...
b 43.8 44.4 50.7 52.6 50 63.3 PANEL B Underdiagnosis of hypertension by completed education, adults 45+c Education 2007 2007 no schooling 79.0 69.5 primary schooling 74.4 58.2 junior high 73.2 52.1 senior high + 68.0 62.1 all adults 45+ 73.6 62.1 PANEL C
|
From page 332... ...
[4.411] Observations 10,376 11,994 1,966 2,745 R-squared 0.064 0.088 0.045 0.072 Cohort Dummy Variables Yes Yes No No Province × Rural Dummy Variables + Yes Yes Province × Province × Province × Rural × Year Interactions rural rural
|
From page 333... ...
age interactions 3.953 0.004 2.913 0.021 1.253 0.288 7.305 0.000 Per capita expenditures 2.569 0.078 4.424 0.012 1.891 0.152 8.253 0.000 Cohort dummy variables 7.374 0.000 4.417 0.000 Year dummy variables 4.791 0.009 1.990 0.138 Province x rural dummy variables 1.769 0.019 3.293 0.000 2.653 0.000 6.955 0.000 Year x prov x rural variables 1.979 0.000 2.682 0.000 interactions NOTES: The dependent variable for the hypertension regressions is whether the individual is hypertensive = 1, 0 otherwise; and for the under diagnosis of hypertension, the dependent variable is 1 if the individual has ever been diagnosed with hypertension, 0 otherwise, conditional of being hypertensive. Blood pressure measurement was not collected in 1993, and question about diagnosis was only asked in 2007.
|
From page 334... ...
A strong, positive relationship between education and memory is also evident, with a negative coefficient on the age-schooling interaction term for men, suggesting that education reinforces the negative effects of aging on memory in this case. The pce variables are jointly significant, positively correlated with word recall, with the effect at low levels of pce for men and at high levels for women.
|
From page 335... ...
CES-D score and age 5.5 5 CES-D Score Male 4.5 Female 4 3.5 40 50 60 70 80 Age FIGURE 13-5 Words recalled and CES-D 10 (2007) , adults aged 45 and older.
|
From page 336... ...
years + educ. age 83.455 0.000 137.819 0.000 9.220 0.000 15.859 0.000 interactions Per capita expenditures 13.403 0.000 10.295 0.000 7.236 0.001 2.950 0.053 Province × rural dummy variables 5.347 0.000 5.761 0.000 5.998 0.000 9.180 0.000
|
From page 337... ...
. bKnot point is at the median pce, coefficient represent change in the slope.
|
From page 338... ...
The pce variables, while not individually significant, are jointly significant (at 10% or lower) and show a negative association between pce and CES-D scores.28 CONCLUSIONS Indonesia has undergone major changes in multiple dimensions since the Indonesia Family Life Survey was first fielded in 1993.
|
From page 339... ...
These conditions indicate that the elderly population in Indonesia is increasingly exposed to higher risk factors that are correlated with chronic problems such as cardiovascular diseases and diabetes. This is quite interesting because this period has seen major gyrations in economic activity, including strong growth from 1993 to 1996, a major economic collapse from late 1997 to 1998, and a strong recovery from 2000 to 2007.
|
From page 340... ...
Indonesia Family Life Survey, Wave 4.
|
From page 341... ...
. Socioeconomic Success and Health in Later Life: Evidence from the Indonesia Family Life Survey.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.